{"doc_id": "1912.05372", "revision_depth": 1, "before_revision": "Language models have become a key step to achieve state-of-the-art results in many different Natural Language Processing (NLP) tasks. Leveraging the huge amount of unlabeled texts nowadays available, they provide an efficient way to pre-train continuous word representations that can be fine-tuned for a downstream task, along with their contextualization at the sentence level. This has been widely demonstrated for English using contextualized word representations such as OpenAI GPT (Radford et al., 2018), BERT (Devlin et al., 2019), or XLNet (Yang et al., 2019b). In this paper, we introduce and share FlauBERT, a model learned on a very large and heterogeneous French corpus. Models of different sizes are trained using the new CNRS (French National Centre for Scientific Research) Jean Zay supercomputer. We apply our French language models to complex NLP tasks (natural language inference, parsing, word sense disambiguation) and show that most of the time they outperform other pre-training approaches. Different versions of FlauBERT as well as a unified evaluation protocol for the downstream tasks are shared with the research community for further reproducible experiments in French NLP.", "after_revision": "Language models have become a key step to achieve state-of-the art results in many different Natural Language Processing (NLP) tasks. Leveraging the huge amount of unlabeled texts nowadays available, they provide an efficient way to pre-train continuous word representations that can be fine-tuned for a downstream task, along with their contextualization at the sentence level. This has been widely demonstrated for English using contextualized word representations such as OpenAI GPT (Radford et al., 2018), BERT (Devlin et al., 2019), or XLNet (Yang et al., 2019b). In this paper, we introduce and share FlauBERT, a model learned on a very large and heterogeneous French corpus. Models of different sizes are trained using the new CNRS (French National Centre for Scientific Research) Jean Zay supercomputer. We apply our French language models to complex NLP tasks (natural language inference, parsing, word sense disambiguation) and show that most of the time they outperform other pre-training approaches. Different versions of FlauBERT as well as a unified evaluation protocol for the downstream tasks , called FLUE (French Language Understanding Evaluation), are shared to the research community for further reproducible experiments in French NLP.", "edit_actions": [{"type": "R", "before": "state-of-the-art", "after": "state-of-the art", "start_char_pos": 50, "end_char_pos": 66, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "are shared with", "after": ", called FLUE (French Language Understanding Evaluation), are shared to", "start_char_pos": 1109, "end_char_pos": 1124, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 133, 378, 568, 681, 811, 1011], "domain": "arxiv"}
{"doc_id": "1912.05372", "revision_depth": 2, "before_revision": "Language models have become a key step to achieve state-of-the art results in many different Natural Language Processing (NLP) tasks. Leveraging the huge amount of unlabeled texts nowadays available, they provide an efficient way to pre-train continuous word representations that can be fine-tuned for a downstream task, along with their contextualization at the sentence level. This has been widely demonstrated for English using contextualized word representations such as OpenAI GPT (Radford et al., 2018 ), BERT ( Devlin et al., 2019 ), or XLNet ( Yang et al., 2019b). In this paper, we introduce and share FlauBERT, a model learned on a very large and heterogeneous French corpus. Models of different sizes are trained using the new CNRS (French National Centre for Scientific Research) Jean Zay supercomputer. We apply our French language models to complex NLP tasks (  natural language inference, parsing, word sense disambiguation) and show that most of the time they outperform other pre-training approaches. Different versions of FlauBERT as well as a unified evaluation protocol for the downstream tasks, called FLUE (French Language Understanding Evaluation), are shared to the research community for further reproducible experiments in French NLP.", "after_revision": "Language models have become a key step to achieve state-of-the art results in many different Natural Language Processing (NLP) tasks. Leveraging the huge amount of unlabeled texts nowadays available, they provide an efficient way to pre-train continuous word representations that can be fine-tuned for a downstream task, along with their contextualization at the sentence level. This has been widely demonstrated for English using contextualized representations (Dai and Le, 2015; Peters et al., 2018 ; Howard and Ruder, 2018; Radford et al., 2018; Devlin et al., 2019 ; Yang et al., 2019b). In this paper, we introduce and share FlauBERT, a model learned on a very large and heterogeneous French corpus. Models of different sizes are trained using the new CNRS (French National Centre for Scientific Research) Jean Zay supercomputer. We apply our French language models to diverse NLP tasks ( text classification, paraphrasing, natural language inference, parsing, word sense disambiguation) and show that most of the time they outperform other pre-training approaches. Different versions of FlauBERT as well as a unified evaluation protocol for the downstream tasks, called FLUE (French Language Understanding Evaluation), are shared to the research community for further reproducible experiments in French NLP.", "edit_actions": [{"type": "R", "before": "word representations such as OpenAI GPT (Radford", "after": "representations (Dai and Le, 2015; Peters", "start_char_pos": 446, "end_char_pos": 494, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "), BERT (", "after": "; Howard and Ruder, 2018; Radford et al., 2018;", "start_char_pos": 508, "end_char_pos": 517, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "), or XLNet (", "after": ";", "start_char_pos": 538, "end_char_pos": 551, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "complex", "after": "diverse", "start_char_pos": 855, "end_char_pos": 862, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "text classification, paraphrasing,", "start_char_pos": 875, "end_char_pos": 875, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 133, 378, 572, 685, 815, 1017], "domain": "arxiv"}
{"doc_id": "1912.10514", "revision_depth": 2, "before_revision": "An effective method to generate a large number of parallel sentences for training improved neural machine translation (NMT) systems is the use of  back-translations of the target-side monolingual data. The method was not able to utilize the available huge amount of  monolingual data because of the inability of  models to differentiate between the authentic and synthetic parallel data  . Tagging, or using gates, has been used to enable translation models to distinguish between synthetic and authentic data, improving standard back-translation and also enabling the use of iterative back-translation on language pairs that under-performed using standard back-translation. This workpresents  pre-training and fine-tuning as a simplified but more effective approach of differentiating between the two data. The approach - tag-less back-translation - trains the model on the synthetic data and fine-tunes it on the authentic data. Experiments have shown the approach to outperform the baseline and standard back-translation by 4.0 and 0.7 BLEU respectively on low resource English-Vietnamese NMT. While the need for tagging (noising) the dataset has been removed, the technique outperformed tagged back-translation by 0.4 BLEU. The approach reached the best scores in less training time than the standard and tagged back-translation approaches .", "after_revision": "An effective method to generate a large number of parallel sentences for training improved neural machine translation (NMT) systems is the use of the back-translations of the target-side monolingual data. The standard back-translation method has been shown to be unable to efficiently utilize the available huge amount of existing monolingual data because of the inability of translation models to differentiate between the authentic and synthetic parallel data during training . Tagging, or using gates, has been used to enable translation models to distinguish between synthetic and authentic data, improving standard back-translation and also enabling the use of iterative back-translation on language pairs that underperformed using standard back-translation. In this work, we approach back-translation as a domain adaptation problem, eliminating the need for explicit tagging. In the approach --tag-less back-translation -- the synthetic and authentic parallel data are treated as out-of-domain and in-domain data respectively and, through pre-training and fine-tuning, the translation model is shown to be able to learn more efficiently from them during training. Experimental results have shown that the approach outperforms the standard and tagged back-translation approaches on low resource English-Vietnamese and English-German neural machine translation .", "edit_actions": [{"type": "A", "before": null, "after": "the", "start_char_pos": 146, "end_char_pos": 146, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "method was not able to", "after": "standard back-translation method has been shown to be unable to efficiently", "start_char_pos": 206, "end_char_pos": 228, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "A", "before": null, "after": "existing", "start_char_pos": 266, "end_char_pos": 266, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "translation", "start_char_pos": 312, "end_char_pos": 312, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "during training", "start_char_pos": 387, "end_char_pos": 387, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "under-performed", "after": "underperformed", "start_char_pos": 626, "end_char_pos": 641, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "This workpresents", "after": "In this work, we approach back-translation as a domain adaptation problem, eliminating the need for explicit tagging. In the approach --", "start_char_pos": 675, "end_char_pos": 692, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "A", "before": null, "after": "tag-less back-translation", "start_char_pos": 692, "end_char_pos": 692, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "-- the synthetic and authentic parallel data are treated as out-of-domain and in-domain data respectively and, through", "start_char_pos": 693, "end_char_pos": 693, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "fine-tuning as a simplified but more effective approach of differentiating between the two data. The approach - tag-less", "after": "fine-tuning, the translation model is shown to be able to learn more efficiently from them during training. Experimental results have shown that the approach outperforms the standard and tagged", "start_char_pos": 711, "end_char_pos": 831, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "- trains the model on the synthetic data and fine-tunes it on the authentic data. Experiments have shown the approach to outperform the baseline and standard back-translation by 4.0 and 0.7 BLEU respectively", "after": "approaches", "start_char_pos": 849, "end_char_pos": 1056, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "NMT. While the need for tagging (noising) the dataset has been removed, the technique outperformed tagged back-translation by 0.4 BLEU. The approach reached the best scores in less training time than the standard and tagged back-translation approaches", "after": "and English-German neural machine translation", "start_char_pos": 1092, "end_char_pos": 1343, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 201, 389, 674, 807, 930, 1227], "domain": "arxiv"}
{"doc_id": "1912.10616", "revision_depth": 1, "before_revision": "Authorship attribution is the process of identifying the author of a text. Classification-based approaches work well for small numbers of candidate authors, but only similarity-based methods are applicable for larger numbers of authors or for authors beyond the training set. While deep learning methods have been applied to classification-based approaches, current similarity-based  methods only embody static notions of similarity. Siamese networks have been used to develop learned notions of similarity in one-shot image tasks, and also for tasks of  semantic relatedness in NLP. We examine their application to the stylistic task of authorship attribution , and show that they can substantially outperform both classification- and existing similarity-based approaches on datasets with large numbers of authors .", "after_revision": "Authorship attribution is the process of identifying the author of a text. Classification-based approaches work well for small numbers of candidate authors, but only similarity-based methods are applicable for larger numbers of authors or for authors beyond the training set. While deep learning methods have been applied to classification-based approaches, applications to similarity-based applications have been limited, and most similarity-based methods only embody static notions of similarity. Siamese networks have been used to develop learned notions of similarity in one-shot image tasks, and also for tasks of mostly semantic relatedness in NLP. We examine their application to the stylistic task of authorship attribution on datasets with large numbers of authors, looking at multiple energy functions and neural network architectures, and show that they can substantially outperform both classification- and existing similarity-based approaches . We also find an unexpected relationship between choice of energy function and number of authors, in terms of performance .", "edit_actions": [{"type": "R", "before": "current", "after": "applications to", "start_char_pos": 358, "end_char_pos": 365, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "applications have been limited, and most similarity-based", "start_char_pos": 383, "end_char_pos": 383, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "mostly", "start_char_pos": 554, "end_char_pos": 554, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ", and", "after": "on datasets with large numbers of authors, looking at multiple energy functions and neural network architectures, and", "start_char_pos": 661, "end_char_pos": 666, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "on datasets with large numbers of authors", "after": ". We also find an unexpected relationship between choice of energy function and number of authors, in terms of performance", "start_char_pos": 773, "end_char_pos": 814, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 74, 275, 433, 583], "domain": "arxiv"}
{"doc_id": "1912.10616", "revision_depth": 2, "before_revision": "Authorship attribution is the process of identifying the author of a text. Classification-based approaches work well for small numbers of candidate authors, but only similarity-based methods  are applicable for larger numbers of authors or for authors beyond the training set . While deep learning methodshave been applied to classification-based approaches, applications to similarity-based applications have been limited, and most similarity-based methods only embody static notions of similarity . Siamese networks have been used to develop learned notions of similarity in one-shot image tasks, and also for tasks of mostly semantic relatedness in NLP. We examine their application to the stylistic task of authorship attribution on datasets with large numbers of authors, looking at multiple energy functions and neural network architectures, and show that they can substantially outperform both classification- and existing similarity-based approaches. We also find an unexpected relationship between choice of energy function and number of authors, in terms of performance .", "after_revision": "Authorship attribution is the process of identifying the author of a text. Approaches to tackling it have been conventionally divided into classification-based ones, which work well for small numbers of candidate authors, and similarity-based methods , which are applicable for larger numbers of authors or for authors beyond the training set ; these existing similarity-based methods have only embodied static notions of similarity. Deep learning methods, which blur the boundaries between classification-based and similarity-based approaches, are promising in terms of ability to learn a notion of similarity, but have previously only been used in a conventional small-closed-class classification setup . Siamese networks have been used to develop learned notions of similarity in one-shot image tasks, and also for tasks of mostly semantic relatedness in NLP. We examine their application to the stylistic task of authorship attribution on datasets with large numbers of authors, looking at multiple energy functions and neural network architectures, and show that they can substantially outperform previous approaches .", "edit_actions": [{"type": "R", "before": "Classification-based approaches", "after": "Approaches to tackling it have been conventionally divided into classification-based ones, which", "start_char_pos": 75, "end_char_pos": 106, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "but only", "after": "and", "start_char_pos": 157, "end_char_pos": 165, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": ", which", "start_char_pos": 191, "end_char_pos": 191, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "coherence"]}, {"type": "R", "before": ". While deep learning methodshave been applied to classification-based approaches, applications to", "after": "; these existing", "start_char_pos": 276, "end_char_pos": 374, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "applications have been limited, and most similarity-based methods only embody static notions of similarity", "after": "methods have only embodied static notions of similarity. Deep learning methods, which blur the boundaries between classification-based and similarity-based approaches, are promising in terms of ability to learn a notion of similarity, but have previously only been used in a conventional small-closed-class classification setup", "start_char_pos": 392, "end_char_pos": 498, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "both classification- and existing similarity-based approaches. We also find an unexpected relationship between choice of energy function and number of authors, in terms of performance", "after": "previous approaches", "start_char_pos": 896, "end_char_pos": 1079, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 74, 277, 500, 656, 958], "domain": "arxiv"}
{"doc_id": "1912.11602", "revision_depth": 1, "before_revision": "Lead bias is a common phenomenon in news summarization, where early parts of an article often contain the most salient information. While many algorithms exploit this fact in summary generation, it has a detrimental effect on teaching the model to discriminate and extract important information  . We propose that the lead bias can be leveraged in  a simple and effective way in our favor to pretrain abstractive news summarization models on large-scale unlabeled corpus : predicting the leading sentences using the rest of an article. Via careful data cleaning and filtering , our transformer-based pretrained model without any finetuning achieves remarkable results over various news summarization tasks. With further finetuning, our model outperforms many competitive baseline models. Human evaluations further show the effectiveness of our method .", "after_revision": "Lead bias is a common phenomenon in news summarization, where early parts of an article often contain the most salient information. While many algorithms exploit this fact in summary generation, it has a detrimental effect on teaching the model to discriminate and extract important information in general . We propose that the lead bias can be leveraged in our favor in a simple and effective way to pre-train abstractive news summarization models on large-scale unlabeled news corpora : predicting the leading sentences using the rest of an article. We collect a massive news corpus and conduct data cleaning and filtering via statistical analysis. We then apply the proposed self-supervised pre-training to existing generation models BART and T5 for domain adaptation. Via extensive experiments on six benchmark datasets, we show that this approach can dramatically improve the summarization quality and achieve state-of-the-art results for zero-shot news summarization without any fine-tuning. For example, in the DUC2003 dataset, the ROUGE-1 score of BART increases 13.7\\% after the lead-bias pre-training. We deploy the model in Microsoft News and provide public APIs as well as a demo website for multi-lingual news summarization .", "edit_actions": [{"type": "A", "before": null, "after": "in general", "start_char_pos": 295, "end_char_pos": 295, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "our favor in", "start_char_pos": 348, "end_char_pos": 348, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in our favor to pretrain", "after": "to pre-train", "start_char_pos": 376, "end_char_pos": 400, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "corpus", "after": "news corpora", "start_char_pos": 464, "end_char_pos": 470, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Via careful", "after": "We collect a massive news corpus and conduct", "start_char_pos": 536, "end_char_pos": 547, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ", our transformer-based pretrained model without any finetuning achieves remarkable results over various news summarization tasks. With further finetuning, our model outperforms many competitive baseline models. Human evaluations further show the effectiveness of our method", "after": "via statistical analysis. We then apply the proposed self-supervised pre-training to existing generation models BART and T5 for domain adaptation. Via extensive experiments on six benchmark datasets, we show that this approach can dramatically improve the summarization quality and achieve state-of-the-art results for zero-shot news summarization without any fine-tuning. For example, in the DUC2003 dataset, the ROUGE-1 score of BART increases 13.7\\% after the lead-bias pre-training. We deploy the model in Microsoft News and provide public APIs as well as a demo website for multi-lingual news summarization", "start_char_pos": 576, "end_char_pos": 850, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 131, 297, 535, 706, 787]}
{"doc_id": "1912.11602", "revision_depth": 2, "before_revision": "Lead bias is a common phenomenon in news summarization, where early parts of an article often contain the most salient information . While many algorithms exploit this fact in summary generation , it has a detrimental effect on teaching the model to discriminate and extract important information in general. We propose that the lead bias can be leveraged in our favor in a simple and effective way to pre-train abstractive news summarization models on large-scale unlabeled news corpora: predicting the leading sentences using the rest of an article. We collect a massive news corpus and conduct data cleaning and filtering via statistical analysis. We then apply the proposed self-supervised pre-training  to existing generation models BART and T5 for domain adaptation. Via extensive experiments on six benchmark datasets, we show that this approach can dramatically improve the summarization quality and achieve state-of-the-art results for zero-shot news summarization without any fine-tuning. For example, in the DUC2003 dataset, the ROUGE-1 score of BART increases 13.7\\% after the lead-bias pre-training. We deploy the model in Microsoft News and provide public APIs as well as a demo website for multi-lingual news summarization.", "after_revision": "A typical journalistic convention in news articles is to deliver the most salient information in the beginning, also known as the lead bias. While this phenomenon can be exploited in generating a summary , it has a detrimental effect on teaching a model to discriminate and extract important information in general. We propose that this lead bias can be leveraged in our favor in a simple and effective way to pre-train abstractive news summarization models on large-scale unlabeled news corpora: predicting the leading sentences using the rest of an article. We collect a massive news corpus and conduct data cleaning and filtering via statistical analysis. We then apply self-supervised pre-training on this dataset to existing generation models BART and T5 for domain adaptation. Via extensive experiments on six benchmark datasets, we show that this approach can dramatically improve the summarization quality and achieve state-of-the-art results for zero-shot news summarization without any fine-tuning. For example, in the DUC2003 dataset, the ROUGE-1 score of BART increases 13.7\\% after the lead-bias pre-training. We deploy the model in Microsoft News and provide public APIs as well as a demo website for multi-lingual news summarization.", "edit_actions": [{"type": "R", "before": "Lead bias is a common phenomenon in news summarization, where early parts of an article often contain", "after": "A typical journalistic convention in news articles is to deliver", "start_char_pos": 0, "end_char_pos": 101, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ". While many algorithms exploit this fact in summary generation", "after": "in the beginning, also known as the lead bias. While this phenomenon can be exploited in generating a summary", "start_char_pos": 131, "end_char_pos": 194, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "the", "after": "a", "start_char_pos": 237, "end_char_pos": 240, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "the", "after": "this", "start_char_pos": 325, "end_char_pos": 328, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "D", "before": "the proposed", "after": null, "start_char_pos": 665, "end_char_pos": 677, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "on this dataset", "start_char_pos": 707, "end_char_pos": 707, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 132, 308, 551, 650, 772, 998, 1112], "domain": "arxiv"}
{"doc_id": "1912.13318", "revision_depth": 1, "before_revision": "Pre-training techniques have been verified successfully in a variety of NLP tasks in recent years. Despite the wide spread of pre-training models for NLP applications, they almost focused on text-level manipulation, while neglecting the layout and style information that is vital for document image understanding. In this paper, we propose textbf LayoutLM to jointly model the interaction between text and layout information across scanned document images, which is beneficial for a great number of real-world document image understanding tasks such as information extraction from scanned documents. We also leverage the image features to incorporate the style information of words in LayoutLM. To the best of our knowledge, this is the first time that text and layout are jointly learned in a single framework for document-level pre-training , leading to significant performance improvement in downstream tasks for document image understanding.", "after_revision": "Pre-training techniques have been verified successfully in a variety of NLP tasks in recent years. Despite the widespread of pre-training models for NLP applications, they almost focused on text-level manipulation, while neglecting the layout and style information that is vital for document image understanding. In this paper, we propose the LayoutLM to jointly model the interaction between text and layout information across scanned document images, which is beneficial for a great number of real-world document image understanding tasks such as information extraction from scanned documents. Furthermore, we also leverage the image features to incorporate the visual information of words into LayoutLM. To the best of our knowledge, this is the first time that text and layout are jointly learned in a single framework for document-level pre-training . It achieves new state-of-the-art results in several downstream tasks, including receipt understanding (from 94.02 to 95.24) and document image classification (from 93.07 to 94.42). The code and pre-trained LayoutLM models will be available soon at URL", "edit_actions": [{"type": "R", "before": "wide spread", "after": "widespread", "start_char_pos": 111, "end_char_pos": 122, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "textbf", "after": null, "start_char_pos": 340, "end_char_pos": 346, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": "LayoutLM", "after": "the LayoutLM", "start_char_pos": 347, "end_char_pos": 355, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "We", "after": "Furthermore, we", "start_char_pos": 600, "end_char_pos": 602, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "style", "after": "visual", "start_char_pos": 655, "end_char_pos": 660, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "in", "after": "into", "start_char_pos": 682, "end_char_pos": 684, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", leading to significant performance improvement in downstream tasks for document image understanding.", "after": ". It achieves new state-of-the-art results in several downstream tasks, including receipt understanding (from 94.02 to 95.24) and document image classification (from 93.07 to 94.42). The code and pre-trained LayoutLM models will be available soon at URL", "start_char_pos": 843, "end_char_pos": 945, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 98, 313, 599, 694], "domain": "arxiv"}
{"doc_id": "1912.13318", "revision_depth": 2, "before_revision": "Pre-training techniques have been verified successfully in a variety of NLP tasks in recent years. Despite the widespread of pre-training models for NLP applications, they almost focused on text-level manipulation, while neglecting the layout and style information that is vital for document image understanding. In this paper, we propose the LayoutLM to jointly model the interaction between text and layout information across scanned document images, which is beneficial for a great number of real-world document image understanding tasks such as information extraction from scanned documents. Furthermore, we also leverage the image features to incorporate the visual information of words into LayoutLM. To the best of our knowledge, this is the first time that text and layout are jointly learned in a single framework for document-level pre-training. It achieves new state-of-the-art results in several downstream tasks, including  receipt understanding (from 94.02 to 95.24) and document image classification (from 93.07 to 94.42). The code and pre-trained LayoutLM models will be available soon at URL", "after_revision": "Pre-training techniques have been verified successfully in a variety of NLP tasks in recent years. Despite the widespread of pre-training models for NLP applications, they almost focused on text-level manipulation, while neglecting the layout and style information that is vital for document image understanding. In this paper, we propose the LayoutLM to jointly model the interaction between text and layout information across scanned document images, which is beneficial for a great number of real-world document image understanding tasks such as information extraction from scanned documents. Furthermore, we also leverage the image features to incorporate the visual information of words into LayoutLM. To the best of our knowledge, this is the first time that text and layout are jointly learned in a single framework for document-level pre-training. It achieves new state-of-the-art results in several downstream tasks, including form understanding (from 70.72 to 79.27), receipt understanding (from 94.02 to 95.24) and document image classification (from 93.07 to 94.42). The code and pre-trained LayoutLM models are publicly available at URL", "edit_actions": [{"type": "A", "before": null, "after": "form understanding (from 70.72 to 79.27),", "start_char_pos": 936, "end_char_pos": 936, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "will be available soon", "after": "are publicly available", "start_char_pos": 1079, "end_char_pos": 1101, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 98, 312, 595, 706, 855, 1037], "domain": "arxiv"}
{"doc_id": "2001.00059", "revision_depth": 2, "before_revision": "Recent research has achieved impressive results on understanding and improving source code by building up on machine-learning techniques developed for natural languages. A significant advancement in natural-language understanding has come with the development of pre-trained contextual embeddings, such as BERT, which can be fine-tuned for downstream tasks with less labeled data and training budget, while achieving better accuracies. However, there is no attempt yet to obtain a high-quality contextual embedding of source code, and to evaluate it on multiple program-understanding tasks simultaneously; that is the gap that this paper aims to mitigate. Specifically, first, we curate a massive, deduplicated corpus of 6M Python files from GitHub, which we use to pre-train CuBERT, an open-sourced code-understanding BERT model; and, second, we create an open-sourced benchmark that comprises five classification tasks and one program-repair task, akin to code-understanding tasks proposed in the literature before. We fine-tune CuBERT on our benchmark tasks, and compare the resulting models to different variants of Word2Vec token embeddings, BiLSTM and Transformer models, as well as published state-of-the-art models, showing that CuBERT outperforms them all, even with shorter training, and with fewer labeled examples. Future work on source-code embedding can benefit from reusing our benchmark, and  comparing against CuBERT models as a strong baseline.", "after_revision": "Recent research has achieved impressive results on understanding and improving source code by building up on machine-learning techniques developed for natural languages. A significant advancement in natural-language understanding has come with the development of pre-trained contextual embeddings, such as BERT, which can be fine-tuned for downstream tasks with less labeled data and training budget, while achieving better accuracies. However, there is no attempt yet to obtain a high-quality contextual embedding of source code, and to evaluate it on multiple program-understanding tasks simultaneously; that is the gap that this paper aims to mitigate. Specifically, first, we curate a massive, deduplicated corpus of 7.4M Python files from GitHub, which we use to pre-train CuBERT, an open-sourced code-understanding BERT model; and, second, we create an open-sourced benchmark that comprises five classification tasks and one program-repair task, akin to code-understanding tasks proposed in the literature before. We fine-tune CuBERT on our benchmark tasks, and compare the resulting models to different variants of Word2Vec token embeddings, BiLSTM and Transformer models, as well as published state-of-the-art models, showing that CuBERT outperforms them all, even with shorter training, and with fewer labeled examples. Future work on source-code embedding can benefit from reusing our benchmark, and from comparing against CuBERT models as a strong baseline.", "edit_actions": [{"type": "R", "before": "6M", "after": "7.4M", "start_char_pos": 721, "end_char_pos": 723, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "from", "start_char_pos": 1408, "end_char_pos": 1408, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}], "sents_char_pos": [0, 169, 435, 605, 655, 830, 1017, 1326], "domain": "arxiv"}
{"doc_id": "2001.01037", "revision_depth": 1, "before_revision": "This paper explains predictions of image captioning models with attention mechanisms beyond visualizing the attention itself. In this paper, we develop variants of layer-wise relevance backpropagation (LRP) and gradient backpropagation, tailored to image captioning with attention. The result provides simultaneously pixel-wise image explanation and linguistic explanation for each word in the captions. We show that given a word in the caption to be explained, explanation methods such as LRP reveal supporting and opposing pixels as well as  words. We compare the properties of attention heatmaps systematically against those computed with explanation methods such as LRP, Grad-CAM and Guided Grad-CAM. We show that explanation methods, firstly, correlate to object locations with higher precision than attention, secondly, are able to identify object words that are unsupported by image content, and thirdly, provide guidance to debias and improve the model. Results are reported for image captioning using two different  attention models trained with Flickr30K and MSCOCO2017 datasets. Experimental analyses show the strength of explanation methods for understanding image captioning attention models.", "after_revision": "This paper explains predictions of image captioning models with attention mechanisms beyond visualizing the attention itself. In this paper, we develop variants of layer-wise relevance backpropagation (LRP) and gradient backpropagation, tailored to image captioning models with attention mechanisms. The explanations provide simultaneously pixel-wise image explanation and linguistic explanation for each word in the captions. We show that given a word in the caption to be explained, explanation methods such as LRP reveal supporting and opposing pixels as well as preceding words. We compare the properties of attention heatmaps systematically against those computed with explanation methods such as LRP, Grad-CAM and Guided Grad-CAM. We show that explanation methods, firstly, correlate to object locations with higher precision than attention, secondly, are able to identify object words that are unsupported by image content, and thirdly, provide guidance to improve and de-bias the model. Results are reported using two different image captioning attention models trained with Flickr30K and MSCOCO2017 datasets. Experimental analyses show the strength of explanation methods for understanding image captioning attention models.", "edit_actions": [{"type": "R", "before": "with attention. The result provides", "after": "models with attention mechanisms. The explanations provide", "start_char_pos": 266, "end_char_pos": 301, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "preceding", "start_char_pos": 543, "end_char_pos": 543, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "debias and improve", "after": "improve and de-bias", "start_char_pos": 932, "end_char_pos": 950, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "for image captioning", "after": null, "start_char_pos": 983, "end_char_pos": 1003, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "image captioning", "start_char_pos": 1024, "end_char_pos": 1024, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 125, 281, 403, 550, 704, 961, 1089], "domain": "arxiv"}
{"doc_id": "2001.01037", "revision_depth": 2, "before_revision": "This paper explains predictions of image captioning models with attention mechanisms beyond visualizing the attention itself. In this paper, we develop variants of layer-wise relevance backpropagation (LRP) and gradient backpropagation , tailored to image captioning models with attention mechanisms. The explanations provide simultaneously pixel-wise image explanation and linguistic explanation for each word in the captions. We show that given a word in the caption to be explained, explanation methods such as LRP reveal supporting and opposing pixels as well as preceding words. We compare the properties of attention heatmaps systematically against those computed with explanation methods such as LRP, Grad-CAM  and Guided Grad-CAM. We show that explanation methods , firstly, correlate to object locations with higher precisionthan attention, secondly, are able to identify object wordsthat are unsupported by image content, and thirdly, provide guidance to improve and de-bias the model. Results are reported using two different image captioning attention models trained with Flickr30K and MSCOCO2017 datasets. Experimental analyses show the strength of explanation methods for understanding image captioning attention models .", "after_revision": "This paper interprets the predictions of image captioning models with attention mechanisms beyond visualizing the attention itself. In this paper, we develop variants of layer-wise relevance propagation (LRP) and gradient-based explanation methods , tailored to image captioning models with attention mechanisms. We compare the interpretability of attention heatmaps systematically against the explanations computed with explanation methods such as LRP, Grad-CAM , and Guided Grad-CAM. We show that explanation methods provide simultaneously pixel-wise image explanation (supporting and opposing pixels of the input image) and linguistic explanation (supporting and opposing words of the preceding sequence) for each word in the predicted captions. We demonstrate with extensive experiments that explanation methods can 1) reveal more related evidence used by the model to make decisions than attention; 2) correlate to object locations with high precision; 3) is helpful to `debug' the model such as analyzing the reasons for hallucinated object words. With the observed properties of explanations, we further design an LRP-inference fine-tuning strategy that can alleviate the object hallucination of image captioning models, meanwhile, maintain the sentence fluency. We conduct experiments with two widely used attention mechanisms: the adaptive attention mechanism calculated with the additive attention and the multi-head attention calculated with the scaled dot product .", "edit_actions": [{"type": "R", "before": "explains", "after": "interprets the", "start_char_pos": 11, "end_char_pos": 19, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "backpropagation", "after": "propagation", "start_char_pos": 185, "end_char_pos": 200, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "gradient backpropagation", "after": "gradient-based explanation methods", "start_char_pos": 211, "end_char_pos": 235, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "The explanations provide simultaneously pixel-wise image explanation and linguistic explanation for each word in the captions. We show that given a word in the caption to be explained, explanation methods such as LRP reveal supporting and opposing pixels as well as preceding words. We compare the properties", "after": "We compare the interpretability", "start_char_pos": 301, "end_char_pos": 609, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "those", "after": "the explanations", "start_char_pos": 655, "end_char_pos": 660, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 717, "end_char_pos": 717, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", firstly,", "after": "provide simultaneously pixel-wise image explanation (supporting and opposing pixels of the input image) and linguistic explanation (supporting and opposing words of the preceding sequence) for each word in the predicted captions. We demonstrate with extensive experiments that explanation methods can 1) reveal more related evidence used by the model to make decisions than attention; 2)", "start_char_pos": 772, "end_char_pos": 782, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "higher precisionthan attention, secondly, are able to identify object wordsthat are unsupported by image content, and thirdly, provide guidance to improve and de-bias the model. Results are reported using two different image captioning attention models trained with Flickr30K and MSCOCO2017 datasets. Experimental analyses show the strength of explanation methods for understanding image captioning attention models", "after": "high precision; 3) is helpful to `debug' the model such as analyzing the reasons for hallucinated object words. With the observed properties of explanations, we further design an LRP-inference fine-tuning strategy that can alleviate the object hallucination of image captioning models, meanwhile, maintain the sentence fluency. We conduct experiments with two widely used attention mechanisms: the adaptive attention mechanism calculated with the additive attention and the multi-head attention calculated with the scaled dot product", "start_char_pos": 818, "end_char_pos": 1233, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 125, 300, 427, 583, 738, 995, 1118], "domain": "arxiv"}
{"doc_id": "2001.04063", "revision_depth": 1, "before_revision": "In this paper, we present a new sequence-to-sequence pre-training model called ProphetNet, which introduces a novel self-supervised objective named future n-gram prediction and the proposed n-stream self-attention mechanism. Instead of the optimization of one-step ahead prediction in traditional sequence-to-sequence model, the ProphetNet is optimized by n-step ahead prediction which predicts the next n tokens simultaneously based on previous context tokens at each time step. The future n-gram prediction explicitly encourages the model to plan for the future tokens and prevent overfitting on strong local correlations. We pre-train ProphetNet using a base scale dataset (16GB) and a large scale dataset (160GB) respectively. Experimental results show ProphetNet achieves the best performance on both abstractive summarization and question generation tasks compared to the models using the same base scale pre-training dataset. For the large scale dataset pre-training, ProphetNet achieves new state-of-the-art results on Gigaword and comparable results on CNN/DailyMail using only about 1/5 pre-training epochs of the previous model .", "after_revision": "In this paper, we present a new sequence-to-sequence pre-training model called ProphetNet, which introduces a novel self-supervised objective named future n-gram prediction and the proposed n-stream self-attention mechanism. Instead of the optimization of one-step ahead prediction in traditional sequence-to-sequence model, the ProphetNet is optimized by n-step ahead prediction which predicts the next n tokens simultaneously based on previous context tokens at each time step. The future n-gram prediction explicitly encourages the model to plan for the future tokens and prevent overfitting on strong local correlations. We pre-train ProphetNet using a base scale dataset (16GB) and a large scale dataset (160GB) respectively. Then we conduct experiments on CNN/DailyMail, Gigaword, and SQuAD 1.1 benchmarks for abstractive summarization and question generation tasks . Experimental results show that ProphetNet achieves new state-of-the-art results on all these datasets compared to the models using the same scale pre-training corpus .", "edit_actions": [{"type": "R", "before": "Experimental results show ProphetNet achieves the best performance on both", "after": "Then we conduct experiments on CNN/DailyMail, Gigaword, and SQuAD 1.1 benchmarks for", "start_char_pos": 731, "end_char_pos": 805, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "compared to the models using the same base scale pre-training dataset. For the large scale dataset pre-training,", "after": ". Experimental results show that", "start_char_pos": 862, "end_char_pos": 974, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "Gigaword and comparable results on CNN/DailyMail using only about 1/5", "after": "all these datasets compared to the models using the same scale", "start_char_pos": 1027, "end_char_pos": 1096, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "epochs of the previous model", "after": "corpus", "start_char_pos": 1110, "end_char_pos": 1138, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 224, 479, 624, 730, 932], "domain": "arxiv"}
{"doc_id": "2001.05272", "revision_depth": 1, "before_revision": "Chinese NER is a challenging task. As pictographs, Chinese characters contain latent glyph information , which is often overlooked. We propose the FGN , Fusion Glyph Network for Chinese NER. This method may offer glyph informationfor fusion representation learning with BERT . The major innovations of FGN include: (1) a novel CNN structure called CGS-CNN is proposed to capture glyph information from both character graphs and their neighboring graphs . (2) we provide a method with sliding window and Slice-Attention to extract interactive information between BERT representation and glyph representation  . Experiments are conducted on four NER datasets, showing that FGN with LSTM-CRF as tagger achieves new state-of-the-arts performance for Chinese NER. Further, more experiments are conducted to investigate the influences of various components and settings in FGN.", "after_revision": "Chinese NER is a challenging task. As pictographs, Chinese characters contain latent glyph infor-mation , which is often overlooked. In this paper, we propose the FGN , Fusion Glyph Network for Chinese NER. Except for adding glyph information, this method may also add extra interactive infor-mation with the fusion mechanism . The major in-novations of FGN include: (1) a novel CNN struc-ture called CGS-CNN is proposed to capture both glyph information and interactive information between glyphs from neighboring characters . (2) we provide a method with sliding window and Slice-Attention to fuse the BERT representation and glyph representation for a character, which may capture potential interactive knowledge be-tween context and glyph . Experiments are con-ducted on four NER datasets, showing that FGN with LSTM-CRF as tagger achieves new state-of-the-arts performance for Chinese NER. Further, more experiments are conducted to inves-tigate the influences of various components and settings in FGN.", "edit_actions": [{"type": "R", "before": "information", "after": "infor-mation", "start_char_pos": 91, "end_char_pos": 102, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "We", "after": "In this paper, we", "start_char_pos": 132, "end_char_pos": 134, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "This method may offer glyph informationfor fusion representation learning with BERT", "after": "Except for adding glyph information, this method may also add extra interactive infor-mation with the fusion mechanism", "start_char_pos": 191, "end_char_pos": 274, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "innovations", "after": "in-novations", "start_char_pos": 287, "end_char_pos": 298, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "structure", "after": "struc-ture", "start_char_pos": 331, "end_char_pos": 340, "major_intent": "others", "raw_intents": ["others", "fluency", "others"]}, {"type": "R", "before": "glyph information from both character graphs and their neighboring graphs", "after": "both glyph information and interactive information between glyphs from neighboring characters", "start_char_pos": 379, "end_char_pos": 452, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "extract interactive information between", "after": "fuse the", "start_char_pos": 522, "end_char_pos": 561, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "for a character, which may capture potential interactive knowledge be-tween context and glyph", "start_char_pos": 607, "end_char_pos": 607, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "conducted", "after": "con-ducted", "start_char_pos": 626, "end_char_pos": 635, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "investigate", "after": "inves-tigate", "start_char_pos": 802, "end_char_pos": 813, "major_intent": "others", "raw_intents": ["others", "others", "others"]}], "sents_char_pos": [0, 34, 131, 190, 314, 758], "domain": "arxiv"}
{"doc_id": "2001.05272", "revision_depth": 2, "before_revision": "Chinese NER is a challenging task. As pictographs, Chinese characters contain latent glyph infor-mation , which is often overlooked. In this paper, we propose the FGN, Fusion Glyph Network for Chinese NER. Except for adding glyph information, this method may also add extra interactive infor-mation with the fusion mechanism. The major in-novations of FGN include: (1) a novel CNN struc-ture called CGS-CNN is proposed to capture both glyph information and interactive information between glyphs from neighboring characters. (2) we provide a method with sliding window and Slice-Attention to fuse the BERT representation and glyph representation for a character, which may capture potential interactive knowledge be-tween context and glyph. Experiments are con-ducted on four NER datasets, showing that FGN with LSTM-CRF as tagger achieves new state-of-the-arts performance for Chinese NER. Further, more experiments are conducted to inves-tigate the influences of various components and settings in FGN.", "after_revision": "Chinese NER is a challenging task. As pictographs, Chinese characters contain latent glyph information , which is often overlooked. In this paper, we propose the FGN, Fusion Glyph Network for Chinese NER. Except for adding glyph information, this method may also add extra interactive information with the fusion mechanism. The major innovations of FGN include: (1) a novel CNN structure called CGS-CNN is proposed to capture both glyph information and interactive information between glyphs from neighboring characters. (2) we provide a method with sliding window and Slice-Attention to fuse the BERT representation and glyph representation for a character, which may capture potential interactive knowledge between context and glyph. Experiments are conducted on four NER datasets, showing that FGN with LSTM-CRF as tagger achieves new state-of-the-arts performance for Chinese NER. Further, more experiments are conducted to investigate the influences of various components and settings in FGN.", "edit_actions": [{"type": "R", "before": "infor-mation", "after": "information", "start_char_pos": 91, "end_char_pos": 103, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "infor-mation", "after": "information", "start_char_pos": 286, "end_char_pos": 298, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "in-novations", "after": "innovations", "start_char_pos": 336, "end_char_pos": 348, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "struc-ture", "after": "structure", "start_char_pos": 381, "end_char_pos": 391, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "be-tween", "after": "between", "start_char_pos": 713, "end_char_pos": 721, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "con-ducted", "after": "conducted", "start_char_pos": 757, "end_char_pos": 767, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "inves-tigate", "after": "investigate", "start_char_pos": 934, "end_char_pos": 946, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 34, 132, 205, 325, 524, 740, 890], "domain": "arxiv"}
{"doc_id": "2001.05687", "revision_depth": 3, "before_revision": "Although over 95 million people worldwide speak the Vietnamese language , there are not many research studies on Vietnamese machine reading comprehension (MRC), the task of understanding a text and answering questions about it. One of the reasons is because of the lack of high-quality benchmark datasets for this task. In this work, we construct a dataset which consists of 417 Vietnamese texts and 2,783 pairs of multiple-choice questions and answers . The texts are commonly used for teaching reading comprehension for elementary school pupils. In addition, we propose a lexical-based MRC technique that utilizes semantic similarity measures and external knowledge sources to analyze questions and extract answers from the given text. We compare the performance of the proposed model with several  lexical-based and neural network-based baseline models. Our proposed technique achieves 61.81\\% in accuracy, which is 5.51\\% higher than the best baseline model. We also measure human performance on our dataset and find that there is a big gap between human and model performances. This indicates that significant progress can be made on this task. The dataset is freely available at our website for research purposes.", "after_revision": "Although Vietnamese is the 17th most popular native-speaker language in the world , there are not many research studies on Vietnamese machine reading comprehension (MRC), the task of understanding a text and answering questions about it. One of the reasons is because of the lack of high-quality benchmark datasets for this task. In this work, we construct a dataset which consists of 2,783 pairs of multiple-choice questions and answers based on 417 Vietnamese texts which are commonly used for teaching reading comprehension for elementary school pupils. In addition, we propose a lexical-based MRC method that utilizes semantic similarity measures and external knowledge sources to analyze questions and extract answers from the given text. We compare the performance of the proposed model with several baseline lexical-based and neural network-based models. Our proposed method achieves 61.81\\% by accuracy, which is 5.51\\% higher than the best baseline model. We also measure human performance on our dataset and find that there is a big gap between machine-model and human performances. This indicates that significant progress can be made on this task. The dataset is freely available on our website for research purposes.", "edit_actions": [{"type": "R", "before": "over 95 million people worldwide speak the Vietnamese language", "after": "Vietnamese is the 17th most popular native-speaker language in the world", "start_char_pos": 9, "end_char_pos": 71, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "417 Vietnamese texts and", "after": null, "start_char_pos": 375, "end_char_pos": 399, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ". The texts", "after": "based on 417 Vietnamese texts which", "start_char_pos": 453, "end_char_pos": 464, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "technique", "after": "method", "start_char_pos": 592, "end_char_pos": 601, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "baseline", "start_char_pos": 800, "end_char_pos": 800, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "baseline", "after": null, "start_char_pos": 840, "end_char_pos": 848, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "technique", "after": "method", "start_char_pos": 870, "end_char_pos": 879, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "in", "after": "by", "start_char_pos": 897, "end_char_pos": 899, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "human and model", "after": "machine-model and human", "start_char_pos": 1053, "end_char_pos": 1068, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "at", "after": "on", "start_char_pos": 1182, "end_char_pos": 1184, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 227, 319, 454, 547, 737, 856, 962, 1082, 1149], "domain": "arxiv"}
{"doc_id": "2001.07676", "revision_depth": 2, "before_revision": "Some NLP tasks can be solved in a fully unsupervised fashion by providing a pretrained language model with \"task descriptions\" in natural language (e.g., Radford et al., 2019). While this approach underperforms its supervised counterpart, we show in this work that the two ideas can be combined: We introduce Pattern-Exploiting Training (PET), a semi-supervised training procedure that reformulates input examples as cloze-style phrases to help language models understand a given task. These phrases are then used to assign soft labels to a large set of unlabeled examples. Finally, regular supervised training is performed on the resulting training set. For several tasks and languages, PET outperforms both supervised training and unsupervised approaches in low-resource settings by a large margin.", "after_revision": "Some NLP tasks can be solved in a fully unsupervised fashion by providing a pretrained language model with \"task descriptions\" in natural language (e.g., Radford et al., 2019). While this approach underperforms its supervised counterpart, we show in this work that the two ideas can be combined: We introduce Pattern-Exploiting Training (PET), a semi-supervised training procedure that reformulates input examples as cloze-style phrases to help language models understand a given task. These phrases are then used to assign soft labels to a large set of unlabeled examples. Finally, standard supervised training is performed on the resulting training set. For several tasks and languages, PET outperforms supervised training and strong semi-supervised approaches in low-resource settings by a large margin.", "edit_actions": [{"type": "R", "before": "regular", "after": "standard", "start_char_pos": 583, "end_char_pos": 590, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "both", "after": null, "start_char_pos": 704, "end_char_pos": 708, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "unsupervised", "after": "strong semi-supervised", "start_char_pos": 733, "end_char_pos": 745, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 176, 485, 573, 654], "domain": "arxiv"}
{"doc_id": "2001.08604", "revision_depth": 1, "before_revision": "Recent works have shown that generative data augmentation, where synthetic samples generated from deep generative models are used to augment the training dataset, benefit certain NLP tasks. In this work, we extend this approach to the task of dialogue state tracking for goal-oriented dialogues, in which the data naturally exhibits a hierarchical structure over utterances and related annotations . Deep generative data augmentation for dialogue state tracking requires the generative model to be aware of the hierarchically structured data . We propose  Variational Hierarchical Dialog Autoencoder (VHDA) for modeling various aspects of goal-oriented dialogues , including linguistic and underlying annotation structures. Our experiments show that our  model is able to generate realistic and novel samples that improve the robustness of state-of-the-art dialogue state trackers, ultimately improving their final dialogue state tracking performances on several datasets .", "after_revision": "Recent works have shown that generative data augmentation, where synthetic samples generated from deep generative models are used to augment the training dataset, benefit certain NLP tasks. In this work, we extend this approach to the task of dialog state tracking for goal-oriented dialogs. Since, goal-oriented dialogs naturally exhibit a hierarchical structure over utterances and related annotations , deep generative data augmentation for the task requires the generative model to be aware of the hierarchical nature . We propose the Variational Hierarchical Dialog Autoencoder (VHDA) for modeling complete aspects of goal-oriented dialogs , including linguistic features and underlying structured annotations, namely dialog acts and goals. We also propose two training policies to mitigate issues that arise with training VAE-based models. Experiments show that our hierarchical model is able to generate realistic and novel samples that improve the robustness of state-of-the-art dialog state trackers, ultimately improving the dialog state tracking performances on various dialog domains. Surprisingly, the ability to jointly generate dialog features enables our model to outperform previous state-of-the-arts in related subtasks, such as language generation and user simulation .", "edit_actions": [{"type": "R", "before": "dialogue", "after": "dialog", "start_char_pos": 243, "end_char_pos": 251, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "dialogues, in which the data naturally exhibits", "after": "dialogs. Since, goal-oriented dialogs naturally exhibit", "start_char_pos": 285, "end_char_pos": 332, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": ". Deep", "after": ", deep", "start_char_pos": 398, "end_char_pos": 404, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "dialogue state tracking", "after": "the task", "start_char_pos": 438, "end_char_pos": 461, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "hierarchically structured data", "after": "hierarchical nature", "start_char_pos": 511, "end_char_pos": 541, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 555, "end_char_pos": 555, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "various", "after": "complete", "start_char_pos": 620, "end_char_pos": 627, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "dialogues", "after": "dialogs", "start_char_pos": 653, "end_char_pos": 662, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and underlying annotation structures. Our experiments", "after": "features and underlying structured annotations, namely dialog acts and goals. We also propose two training policies to mitigate issues that arise with training VAE-based models. Experiments", "start_char_pos": 686, "end_char_pos": 739, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "hierarchical", "start_char_pos": 754, "end_char_pos": 754, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "dialogue", "after": "dialog", "start_char_pos": 857, "end_char_pos": 865, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "their final dialogue", "after": "the dialog", "start_char_pos": 903, "end_char_pos": 923, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "several datasets", "after": "various dialog domains. Surprisingly, the ability to jointly generate dialog features enables our model to outperform previous state-of-the-arts in related subtasks, such as language generation and user simulation", "start_char_pos": 955, "end_char_pos": 971, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 189, 399, 543, 723], "domain": "arxiv"}
{"doc_id": "2001.08604", "revision_depth": 2, "before_revision": "Recent works have shown that generative data augmentation, where synthetic samples generated from deep generative models are used to augment the training dataset, benefit certain NLP tasks. In this work, we extend this approach to the task of dialog state tracking for goal-oriented dialogs. Since, goal-oriented dialogs naturally exhibit a hierarchical structure over utterances and related annotations, deep generative data augmentation for the task requires the generative model to be aware of the hierarchical nature . We propose the Variational Hierarchical Dialog Autoencoder (VHDA) for modeling  complete aspects of goal-oriented dialogs, including linguistic features and underlying structured annotations, namely dialog acts and goals. We also propose two training policies to mitigate issues that arise with training VAE-based models. Experiments show that our hierarchical model is able to generate realistic and novel samples that improve the robustness of state-of-the-art dialog state trackers, ultimately improving the dialog state tracking performances on various dialog domains. Surprisingly, the ability to jointly generate dialog features enables our model to outperform previous state-of-the-arts in related subtasks, such as language generation and user simulation  .", "after_revision": "Recent works have shown that generative data augmentation, where synthetic samples generated from deep generative models complement the training dataset, benefit NLP tasks. In this work, we extend this approach to the task of dialog state tracking for goal-oriented dialogs. Due to the inherent hierarchical structure of goal-oriented dialogs over utterances and related annotations, the deep generative model must be capable of capturing the coherence among different hierarchies and types of dialog features . We propose the Variational Hierarchical Dialog Autoencoder (VHDA) for modeling the complete aspects of goal-oriented dialogs, including linguistic features and underlying structured annotations, namely speaker information, dialog acts, and goals. The proposed architecture is designed to model each aspect of goal-oriented dialogs using inter-connected latent variables and learns to generate coherent goal-oriented dialogs from the latent spaces. To overcome training issues that arise from training complex variational models, we propose appropriate training strategies. Experiments on various dialog datasets show that our model improves the downstream dialog trackers' robustness via generative data augmentation. We also discover additional benefits of our unified approach to modeling goal-oriented dialogs: dialog response generation and user simulation , where our model outperforms previous strong baselines .", "edit_actions": [{"type": "R", "before": "are used to augment", "after": "complement", "start_char_pos": 121, "end_char_pos": 140, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "certain", "after": null, "start_char_pos": 171, "end_char_pos": 178, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "Since,", "after": "Due to the inherent hierarchical structure of", "start_char_pos": 292, "end_char_pos": 298, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "dialogs naturally exhibit a hierarchical structure", "after": "dialogs", "start_char_pos": 313, "end_char_pos": 363, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "deep generative data augmentation for the task requires the generative model to be aware of the hierarchical nature", "after": "the deep generative model must be capable of capturing the coherence among different hierarchies and types of dialog features", "start_char_pos": 405, "end_char_pos": 520, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 602, "end_char_pos": 602, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "dialog acts", "after": "speaker information, dialog acts,", "start_char_pos": 722, "end_char_pos": 733, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "We also propose two training policies to mitigate", "after": "The proposed architecture is designed to model each aspect of goal-oriented dialogs using inter-connected latent variables and learns to generate coherent goal-oriented dialogs from the latent spaces. To overcome training", "start_char_pos": 745, "end_char_pos": 794, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "with training VAE-based models. Experiments", "after": "from training complex variational models, we propose appropriate training strategies. Experiments on various dialog datasets", "start_char_pos": 813, "end_char_pos": 856, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "hierarchical model is able to generate realistic and novel samples that improve the robustness of state-of-the-art dialog state trackers, ultimately improving the dialog state tracking performances on various dialog domains. Surprisingly, the ability to jointly generate dialog features enables our model to outperform previous state-of-the-arts in related subtasks, such as language", "after": "model improves the downstream dialog trackers' robustness via generative data augmentation. We also discover additional benefits of our unified approach to modeling goal-oriented dialogs: dialog response", "start_char_pos": 871, "end_char_pos": 1254, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ", where our model outperforms previous strong baselines", "start_char_pos": 1286, "end_char_pos": 1286, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 189, 291, 522, 744, 844, 1095], "domain": "arxiv"}
{"doc_id": "2001.11453", "revision_depth": 1, "before_revision": "Most combinations of NLP tasks and language varieties lack in-domain examples for supervised training because of the paucity of annotated data. How can neural models make sample-efficient generalizations from task-language combinations with available data to low-resource ones? In this work, we propose a Bayesian generative model for the space of neural parameters. We assume that this space can be factorized into latent variables for each language and each task. We infer the posteriors over such latent variables based on data from seen task-language combinations through variational inference. This enables zero-shot classification on unseen combinations at prediction time. For instance, given training data for named entity recognition (NER) in Vietnamese and for part-of-speech (POS) tagging in Wolof, our model can perform accurate predictions for NER in Wolof. In particular, we experiment with a typologically diverse sample of 33 languages from 4 continents and 11 families, and show that our model yields comparable or better results than state-of-the-art, zero-shot cross-lingual transfer methods ; it increases performance by 4.49 points for POS tagging and 7.73 points for NER on average compared to the strongest baseline .", "after_revision": "Most combinations of NLP tasks and language varieties lack in-domain examples for supervised training because of the paucity of annotated data. How can neural models make sample-efficient generalizations from task--language combinations with available data to low-resource ones? In this work, we propose a Bayesian generative model for the space of neural parameters. We assume that this space can be factorized into latent variables for each language and each task. We infer the posteriors over such latent variables based on data from seen task--language combinations through variational inference. This enables zero-shot classification on unseen combinations at prediction time. For instance, given training data for named entity recognition (NER) in Vietnamese and for part-of-speech (POS) tagging in Wolof, our model can perform accurate predictions for NER in Wolof. In particular, we experiment with a typologically diverse sample of 33 languages from 4 continents and 11 families, and show that our model yields comparable or better results than state-of-the-art, zero-shot cross-lingual transfer methods . Moreover, we demonstrate that approximate Bayesian model averaging results in smoother predictive distributions, whose entropy strongly correlates with accuracy. Hence, the proposed framework also offers robust estimates of uncertainty .", "edit_actions": [{"type": "R", "before": "task-language", "after": "task--language", "start_char_pos": 209, "end_char_pos": 222, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "task-language", "after": "task--language", "start_char_pos": 541, "end_char_pos": 554, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "; it increases performance by 4.49 points for POS tagging and 7.73 points for NER on average compared to the strongest baseline", "after": ". Moreover, we demonstrate that approximate Bayesian model averaging results in smoother predictive distributions, whose entropy strongly correlates with accuracy. Hence, the proposed framework also offers robust estimates of uncertainty", "start_char_pos": 1111, "end_char_pos": 1238, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 143, 277, 366, 465, 598, 679, 870, 1112]}
{"doc_id": "2001.11453", "revision_depth": 2, "before_revision": "Most combinations of NLP tasks and language varieties lack in-domain examples for supervised training because of the paucity of annotated data. How can neural models make sample-efficient generalizations from task--language combinations with available data to low-resource ones? In this work, we propose a Bayesian generative model for the space of neural parameters. We assume that this space can be factorized into latent variables for each language and each task. We infer the posteriors over such latent variables based on data from seen task--language combinations through variational inference. This enables zero-shot classification on unseen combinations at prediction time. For instance, given training data for named entity recognition (NER) in Vietnamese and for part-of-speech (POS) tagging in Wolof, our model can perform accurate predictions for NER in Wolof. In particular, we experiment with a typologically diverse sample of 33 languages from 4 continents and 11 families, and show that our model yields comparable or better results than state-of-the-art, zero-shot cross-lingual transfer methods. Moreover, we demonstrate that approximate Bayesian model averaging results in smoother predictive distributions, whose entropy strongly correlates with accuracy. Hence, the proposed framework also offers robust estimates of uncertainty.", "after_revision": "Most combinations of NLP tasks and language varieties lack in-domain examples for supervised training because of the paucity of annotated data. How can neural models make sample-efficient generalizations from task-language combinations with available data to low-resource ones? In this work, we propose a Bayesian generative model for the space of neural parameters. We assume that this space can be factorized into latent variables for each language and each task. We infer the posteriors over such latent variables based on data from seen task-language combinations through variational inference. This enables zero-shot classification on unseen combinations at prediction time. For instance, given training data for named entity recognition (NER) in Vietnamese and for part-of-speech (POS) tagging in Wolof, our model can perform accurate predictions for NER in Wolof. In particular, we experiment with a typologically diverse sample of 33 languages from 4 continents and 11 families, and show that our model yields comparable or better results than state-of-the-art, zero-shot cross-lingual transfer methods. Moreover, we demonstrate that approximate Bayesian model averaging results in smoother predictive distributions, whose entropy inversely correlates with accuracy. Hence, the proposed framework also offers robust estimates of prediction uncertainty. Our code is located at github.com/cambridgeltl/parameter-factorization", "edit_actions": [{"type": "R", "before": "task--language", "after": "task-language", "start_char_pos": 209, "end_char_pos": 223, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "task--language", "after": "task-language", "start_char_pos": 542, "end_char_pos": 556, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "strongly", "after": "inversely", "start_char_pos": 1241, "end_char_pos": 1249, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "uncertainty.", "after": "prediction uncertainty. Our code is located at github.com/cambridgeltl/parameter-factorization", "start_char_pos": 1338, "end_char_pos": 1350, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 143, 278, 367, 466, 600, 681, 872, 1113, 1275]}
{"doc_id": "2002.06353", "revision_depth": 1, "before_revision": "We propose UniViLM: a Unified Video and Language pre-training Model for multimodal understanding and generation. Motivated by the recent success of BERT based pre-training technique for NLP and image-language tasks, VideoBERT and CBT are proposed to exploit BERT model for video and language pre-training using narrated instructional videos. Different from their works which only pre-train understanding task, we propose a unified video-language pre-training model for both understanding and generation tasks . Our model comprises of 4 components including two single-modal encoders, a cross encoder and a decoder with the Transformer backbone. We first pre-train our model to learn the universal representation for both video and language on a large instructional video dataset. Then we fine-tune the model on two multimodal tasks including understanding task (text-based video retrieval) and generation task (multimodal video captioning). Our extensive experiments show that our method can improve the performance of both understanding and generation tasks and achieves the state-of-the art results.", "after_revision": "With the recent success of pre-training technique for NLP and image-linguistic tasks, there are still few works on video-linguistic pre-training . Besides, most of the existing multimodal models are pre-trained for understanding task, which leads to a pretrain-finetune discrepency for generation tasks. In this paper, we propose UniViLM: a Unified Video and Language pre-training Model for both multimodal understanding and generation . Our model comprises of 4 components including two single-modal encoders, a cross encoder and a decoder with the Transformer backbone. We first pre-train our model to learn the universal representation for both video and language on a large instructional video dataset. Then we fine-tune the model on two multimodal tasks including understanding task (text-based video retrieval) and generation task (multimodal video captioning). Our extensive experiments show that our method can improve the performance of both understanding and generation tasks and achieves the state-of-the art results.", "edit_actions": [{"type": "R", "before": "We propose UniViLM: a Unified Video and Language pre-training Model for multimodal understanding and generation. Motivated by", "after": "With", "start_char_pos": 0, "end_char_pos": 125, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "BERT based", "after": null, "start_char_pos": 148, "end_char_pos": 158, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "R", "before": "image-language tasks, VideoBERT and CBT are proposed to exploit BERT model for video and language", "after": "image-linguistic tasks, there are still few works on video-linguistic", "start_char_pos": 194, "end_char_pos": 291, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "using narrated instructional videos. Different from their works which only pre-train", "after": ". Besides, most of the existing multimodal models are pre-trained for", "start_char_pos": 305, "end_char_pos": 389, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "we propose a unified video-language", "after": "which leads to a pretrain-finetune discrepency for generation tasks. In this paper, we propose UniViLM: a Unified Video and Language", "start_char_pos": 410, "end_char_pos": 445, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "model for both", "after": "Model for both multimodal", "start_char_pos": 459, "end_char_pos": 473, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "tasks", "after": null, "start_char_pos": 503, "end_char_pos": 508, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 112, 341, 510, 644, 779, 940], "domain": "arxiv"}
{"doc_id": "2002.09253", "revision_depth": 1, "before_revision": "Autonomous reinforcement learning agents must be intrinsically motivated to explore their environment, discover potential goals, represent them and learn how to achieve them. As children do the same, they benefit from exposure to language, using it to formulate goals and imagine new ones as they learn their meaning. In our proposed learning architecture (IMAGINE), the agent freely explores its environment and turns natural language descriptions of interesting interactions from a social partner into potential goals. IMAGINE learns to represent goals by jointly learning a language model and a goal-conditioned reward function. Just like humans, our agent uses language compositionality to generate new goals by composing known ones  . Leveraging modular model architectures based on Deep Sets and gated-attention mechanisms, IMAGINE autonomously builds a repertoire of behaviors and shows good zero-shot generalization properties for various types of generalization. When imagining its own goals, the agent leverages zero-shot generalization of the reward function to further train on imagined goals and refine its behavior. We present experiments in a simulated domain where the agent interacts with procedurally generated scenes containing objects of various types and colors, discovers goals, imagines others and learns to achieve them.", "after_revision": "Autonomous reinforcement learning agents must be intrinsically motivated to explore their environment, discover potential goals, represent them and learn how to achieve them. As children do the same, they benefit from exposure to language, using it to formulate goals and imagine new ones as they learn their meaning. In our proposed learning architecture (IMAGINE), the agent freely explores its environment and turns natural language descriptions of interesting interactions from a social partner into potential goals. IMAGINE learns to represent goals by jointly learning a language encoder and a goal-conditioned reward function. Just like humans, our agent uses language compositionality to generate new goals by composing known ones , using an algorithm grounded in construction grammar models of child language acquisition . Leveraging modular model architectures based on deepsets and gated attention mechanisms, IMAGINE autonomously builds a repertoire of behaviors and shows good zero-shot generalization properties for various types of generalization. When imagining its own goals, the agent leverages zero-shot generalization of the reward function to further train on imagined goals and refine its behavior. We present experiments in a simulated domain where the agent interacts with procedurally generated scenes containing objects of various types and colors, discovers goals, imagines others and learns to achieve them.", "edit_actions": [{"type": "R", "before": "model", "after": "encoder", "start_char_pos": 586, "end_char_pos": 591, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ", using an algorithm grounded in construction grammar models of child language acquisition", "start_char_pos": 737, "end_char_pos": 737, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "Deep Sets and gated-attention", "after": "deepsets and gated attention", "start_char_pos": 788, "end_char_pos": 817, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 174, 317, 520, 631, 971, 1129], "domain": "arxiv"}
{"doc_id": "2002.09253", "revision_depth": 2, "before_revision": "Autonomous reinforcement learning agents must be intrinsically motivated to explore their environment, discover potential goals, represent them and learn how to achieve them. As children do the same, they benefit from exposure to language, using it to formulate goals and imagine new ones as they learn their meaning. In our proposed learning architecture (IMAGINE), the agent freely explores its environment and turns natural language descriptions of interesting interactions from a social partner into potential goals. IMAGINE learns to represent goalsby jointly learning a language encoder and a goal-conditioned reward function . Just like humans, our agent uses language compositionality to generate new goals by composing known ones, using an algorithm grounded in construction grammar models of child language acquisition. Leveraging modular model architectures based on deepsets and gated attention mechanisms, IMAGINE autonomously builds a repertoire of behaviors and shows good zero-shot generalization properties for various types of generalization. When imagining its own goals, the agent leverages zero-shot generalization of the reward function to further train on imagined goals and refine its behavior. We present experiments in a simulated domain where the agent interacts with procedurally generated scenes containing objects of various types and colors, discovers goals, imagines others and learns to achieve them .", "after_revision": "Developmental machine learning studies how artificial agents can model the way children learn open-ended repertoires of skills. Such agents need to create and represent goals, select which ones to pursue and learn to achieve them. Recent approaches have considered goal spaces that were either fixed and hand-defined or learned using generative models of states. This limited agents to sample goals within the distribution of known effects. We argue that the ability to imagine out-of-distribution goals is key to enable creative discoveries and open-ended learning. Children do so by leveraging the compositionality of language as a tool to imagine descriptions of outcomes they never experienced before, targeting them as goals during play. We introduce Imagine, an intrinsically motivated deep reinforcement learning architecture that models this ability. Such imaginative agents, like children, benefit from the guidance of a social peer who provides language descriptions. To take advantage of goal imagination, agents must be able to leverage these descriptions to interpret their imagined out-of-distribution goals. This generalization is made possible by modularity: a decomposition between learned goal-achievement reward function and policy relying on deep sets, gated attention and object-centered representations. We introduce the Playground environment and study how this form of goal imagination improves generalization and exploration over agents lacking this capacity. In addition, we identify the properties of goal imagination that enable these results and study the impacts of modularity and social interactions .", "edit_actions": [{"type": "R", "before": "Autonomous reinforcement learning agents must be intrinsically motivated to explore their environment, discover potential goals, represent them and learn how", "after": "Developmental machine learning studies how artificial agents can model the way children learn open-ended repertoires of skills. Such agents need to create and represent goals, select which ones to pursue and learn", "start_char_pos": 0, "end_char_pos": 157, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "As children do the same, they benefit from exposure to language, using it to formulate goals and imagine new ones as they learn their meaning. In our proposed learning architecture (IMAGINE), the agent freely explores its environment and turns natural language descriptions of interesting interactions from a social partner into potential goals. IMAGINE learns to represent goalsby jointly learning a language encoder and a goal-conditioned reward function . Just like humans, our agent uses language compositionality to generate new goals by composing known ones, using an algorithm grounded in construction grammar models of child language acquisition. Leveraging modular model architectures based on deepsets and gated attention mechanisms, IMAGINE autonomously builds a repertoire of behaviors and shows good zero-shot generalization properties for various types of generalization. When imagining its own goals, the agent leverages zero-shot generalization of the reward function to further train on imagined goals and refine its behavior. We present experiments in a simulated domain where the agent interacts with procedurally generated scenes containing objects of various types and colors, discovers goals, imagines others and learns to achieve them", "after": "Recent approaches have considered goal spaces that were either fixed and hand-defined or learned using generative models of states. This limited agents to sample goals within the distribution of known effects. We argue that the ability to imagine out-of-distribution goals is key to enable creative discoveries and open-ended learning. Children do so by leveraging the compositionality of language as a tool to imagine descriptions of outcomes they never experienced before, targeting them as goals during play. We introduce Imagine, an intrinsically motivated deep reinforcement learning architecture that models this ability. Such imaginative agents, like children, benefit from the guidance of a social peer who provides language descriptions. To take advantage of goal imagination, agents must be able to leverage these descriptions to interpret their imagined out-of-distribution goals. This generalization is made possible by modularity: a decomposition between learned goal-achievement reward function and policy relying on deep sets, gated attention and object-centered representations. We introduce the Playground environment and study how this form of goal imagination improves generalization and exploration over agents lacking this capacity. In addition, we identify the properties of goal imagination that enable these results and study the impacts of modularity and social interactions", "start_char_pos": 175, "end_char_pos": 1432, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 174, 317, 520, 829, 1060, 1218], "domain": "arxiv"}
{"doc_id": "2002.09616", "revision_depth": 1, "before_revision": "Producing natural and accurate responses like human beings is the ultimate goal of intelligent dialogue agents. So far, most of the past works concentrate on selecting or generating one pertinent and fluent response according to current query and its context. These models work on a one-to-one environment, making one response to one utterance each round.However, in real human-human conversations, human often sequentially sends several short messages for readability instead of a long message in one turn. Thus messages will not end with an explicit ending signal, which is crucial for agents to decide when to reply. So the first step for an intelligent dialogue agent is not replying but deciding if it should reply at the moment. To address this issue, in this paper , we propose a novel Imagine-then-Arbitrate (ITA) neural dialogue model to help the agent decide whether to wait or to make a response directly. Our method has two imaginator modules and an arbitrator module. The two imaginators will learn the agent's and user's speaking style respectively, generate possible utterances as the input of the arbitrator, combining with dialogue history. And the arbitrator decides whether to wait or to make a response to the user directly. To verify the performance and effectiveness of our method, we prepared two dialogue datasets and compared our approach with several popular models. Experimental results show that our model performs well on addressing ending prediction issue and outperforms baseline models .", "after_revision": "Different people have different habits of describing their intents in conversations. Some people may tend to deliberate their full intents in several successive utterances, i.e., they use several consistent messages for readability instead of a long sentence to express their question. This creates a predicament faced by dialogue systems' application, especially in real-world industrial scenarios, in which the dialogue system is unsure that whether it should answer the user's query immediately or wait for users' further supplementary input. Motivated by such interesting quandary, we define a novel task: Wait-or-Answer to better tackle this dilemma faced by dialogue systems. We shed light on a new research topic about how the dialogue system can be more competent to behave in this Wait-or-Answer quandary. Further , we propose a predictive approach dubbed Imagine-then-Arbitrate (ITA) to resolve this Wait-or-Answer task. More specifically, we take advantage of an arbitrator model to help the dialogue system decide to wait or answer. The arbitrator's decision is made with the assistance of two ancillary imaginator models: a wait imaginator and an answer imaginator. The wait imaginator tries to predict what the user would supplement and use its prediction to persuade the arbitrator that the user has some information to add, so the dialogue system should wait. The answer imaginator, nevertheless, struggles to predict the answer of the dialogue system and convince the arbitrator that it's a superior choice to answer the users' query immediately. To our best knowledge, our paper is the first work to explicitly define the Wait-or-Answer task in the dialogue system. Additionally, our proposed ITA approach significantly outperforms the existing models in solving this Wait-or-Answer problem .", "edit_actions": [{"type": "R", "before": "Producing natural and accurate responses like human beings is the ultimate goal of intelligent dialogue agents. So far, most of the past works concentrate on selecting or generating one pertinent and fluent response according to current query and its context. These models work on a one-to-one environment, making one response to one utterance each round.However, in real human-human conversations, human often sequentially sends several short", "after": "Different people have different habits of describing their intents in conversations. Some people may tend to deliberate their full intents in several successive utterances, i.e., they use several consistent", "start_char_pos": 0, "end_char_pos": 443, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "message in one turn. Thus messages will not end with an explicit ending signal, which is crucial for agents to decide when to reply. So the first step for an intelligent dialogue agent is not replying but deciding if it should reply at the moment. To address this issue, in this paper", "after": "sentence to express their question. This creates a predicament faced by dialogue systems' application, especially in real-world industrial scenarios, in which the dialogue system is unsure that whether it should answer the user's query immediately or wait for users' further supplementary input. Motivated by such interesting quandary, we define a novel task: Wait-or-Answer to better tackle this dilemma faced by dialogue systems. We shed light on a new research topic about how the dialogue system can be more competent to behave in this Wait-or-Answer quandary. Further", "start_char_pos": 487, "end_char_pos": 771, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "novel", "after": "predictive approach dubbed", "start_char_pos": 787, "end_char_pos": 792, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "neural dialogue", "after": "to resolve this Wait-or-Answer task. More specifically, we take advantage of an arbitrator", "start_char_pos": 822, "end_char_pos": 837, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "agent decide whether", "after": "dialogue system decide", "start_char_pos": 856, "end_char_pos": 876, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "to make a response directly. Our method has two imaginator modules and an arbitrator module. The two imaginators will learn the agent's and user's speaking style respectively, generate possible utterances as the input of the arbitrator, combining with dialogue history. And the arbitrator decides whether to wait or to make a response to the user directly. To verify the performance and effectiveness of our method, we prepared two dialogue datasets and compared our approach with several popular models. Experimental results show that our model performs well on addressing ending prediction issue and outperforms baseline models", "after": "answer. The arbitrator's decision is made with the assistance of two ancillary imaginator models: a wait imaginator and an answer imaginator. The wait imaginator tries to predict what the user would supplement and use its prediction to persuade the arbitrator that the user has some information to add, so the dialogue system should wait. The answer imaginator, nevertheless, struggles to predict the answer of the dialogue system and convince the arbitrator that it's a superior choice to answer the users' query immediately. To our best knowledge, our paper is the first work to explicitly define the Wait-or-Answer task in the dialogue system. Additionally, our proposed ITA approach significantly outperforms the existing models in solving this Wait-or-Answer problem", "start_char_pos": 888, "end_char_pos": 1517, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 111, 259, 355, 507, 619, 734, 916, 980, 1157, 1244, 1392], "domain": "arxiv"}
{"doc_id": "2002.09616", "revision_depth": 2, "before_revision": "Different people have different habits of describing their intents in conversations. Some people may tend to deliberate their full intents in several successive utterances, i. e., they use several consistent messages for readability instead of a long sentence to express their question. This creates a predicament faced by dialogue systems' application, especially in real-world industrial scenarios, in which the dialogue system is unsure that whether it should answer the user's query immediately or wait for users' further supplementary input. Motivated by such interesting quandary, we define a novel task: Wait-or-Answer to better tackle this dilemma faced by dialogue systems. We shed light on a new research topic about how the dialogue system can be more competent to behave in this Wait-or-Answer quandary. Further , we propose a predictive approach dubbed Imagine-then-Arbitrate (ITA) to resolve this Wait-or-Answer task. More specifically, we take advantage of an arbitrator model to help the dialogue system decide to wait or answer. The arbitrator's decision is made with the assistance of two ancillary imaginator models: a wait imaginator and an answer imaginator. The wait imaginator tries to predict what the user would supplement and use its prediction to persuade the arbitrator that the userhas some information to add, so the dialogue system should wait. The answer imaginator, nevertheless, struggles to predict the answer of the dialogue system and convince the arbitrator that it's a superior choice to answer the users' query immediately. To our best knowledge, our paper is the first work to explicitly define the Wait-or-Answer task in the dialogue system. Additionally, our proposed ITA approach significantly outperforms the existing modelsin solving this Wait-or-Answer problem .", "after_revision": "Producing natural and accurate responses like human beings is the ultimate goal of intelligent dialogue agents. So far, most of the past works concentrate on selecting or generating one pertinent and fluent response according to current query and its context. These models work on a one-to-one environment, making one response to one utterance each round. However, in real human-human conversations, human often sequentially sends several short messages for readability instead of a long message in one turn. Thus messages will not end with an explicit ending signal, which is crucial for agents to decide when to reply. So the first step for an intelligent dialogue agent is not replying but deciding if it should reply at the moment. To address this issue, in this paper , we propose a novel Imagine-then-Arbitrate (ITA) neural dialogue model to help the agent decide whether to wait or to make a response directly. Our method has two imaginator modules and an arbitrator module. The two imaginators will learn the agent's and user's speaking style respectively, generate possible utterances as the input of the arbitrator, combining with dialogue history. And the arbitrator decides whether to wait or to make a response to the user directly. To verify the performance and effectiveness of our method, we prepared two dialogue datasets and compared our approach with several popular models. Experimental results show that our model performs well on addressing ending prediction issue and outperforms baseline models .", "edit_actions": [{"type": "R", "before": "Different people have different habits of describing their intents in conversations. Some people may tend to deliberate their full intents in several successive utterances, i. e., they use several consistent", "after": "Producing natural and accurate responses like human beings is the ultimate goal of intelligent dialogue agents. So far, most of the past works concentrate on selecting or generating one pertinent and fluent response according to current query and its context. These models work on a one-to-one environment, making one response to one utterance each round. However, in real human-human conversations, human often sequentially sends several short", "start_char_pos": 0, "end_char_pos": 207, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "sentence to express their question. This creates a predicament faced by dialogue systems' application, especially in real-world industrial scenarios, in which the dialogue system is unsure that whether it should answer the user's query immediately or wait for users' further supplementary input. Motivated by such interesting quandary, we define a novel task: Wait-or-Answer to better tackle this dilemma faced by dialogue systems. We shed light on a new research topic about how the dialogue system can be more competent to behave in this Wait-or-Answer quandary. Further", "after": "message in one turn. Thus messages will not end with an explicit ending signal, which is crucial for agents to decide when to reply. So the first step for an intelligent dialogue agent is not replying but deciding if it should reply at the moment. To address this issue, in this paper", "start_char_pos": 251, "end_char_pos": 823, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "predictive approach dubbed", "after": "novel", "start_char_pos": 839, "end_char_pos": 865, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "to resolve this Wait-or-Answer task. More specifically, we take advantage of an arbitrator", "after": "neural dialogue", "start_char_pos": 895, "end_char_pos": 985, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "dialogue system decide", "after": "agent decide whether", "start_char_pos": 1004, "end_char_pos": 1026, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "answer. The arbitrator's decision is made with the assistance of two ancillary imaginator models: a wait imaginator and an answer imaginator. The wait imaginator tries to predict what the user would supplement and use its prediction to persuade the arbitrator that the userhas some information to add, so the dialogue system should wait. The answer imaginator, nevertheless, struggles to predict the answer of the dialogue system and convince the arbitrator that it's a superior choice to answer the users' query immediately. To our best knowledge, our paper is the first work to explicitly define the Wait-or-Answer task in the dialogue system. Additionally, our proposed ITA approach significantly outperforms the existing modelsin solving this Wait-or-Answer problem", "after": "to make a response directly. Our method has two imaginator modules and an arbitrator module. The two imaginators will learn the agent's and user's speaking style respectively, generate possible utterances as the input of the arbitrator, combining with dialogue history. And the arbitrator decides whether to wait or to make a response to the user directly. To verify the performance and effectiveness of our method, we prepared two dialogue datasets and compared our approach with several popular models. Experimental results show that our model performs well on addressing ending prediction issue and outperforms baseline models", "start_char_pos": 1038, "end_char_pos": 1807, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 84, 286, 546, 682, 815, 931, 1045, 1179, 1375, 1563, 1683], "domain": "arxiv"}
{"doc_id": "2002.10107", "revision_depth": 2, "before_revision": "Community Question-Answering websites, such as StackOverflow and Quora, expect users to follow specific guidelines in order to maintain content quality. These systems mainly rely on community reports for assessing contents, which has serious problems such as the slow handling of violations, the loss of normal and experienced users' time, the low quality of some reports, and discouraging feedback to new users. Therefore, with the overall goal of providing solutions for automating moderation actions in Q&A websites, we aim to provide a model to predict 20 quality or subjective aspects of questions in QA websites. To this end, we used data gathered by the CrowdSource team at Google Research in 2019 and  fine-tuned pre-trained BERT model on our problem. Based on  evaluation by Mean-Squared-Error (MSE), model achieved the value of 0.046 after 2 epochs of training, which did not improve substantially in the next ones. Results confirm that by simple fine-tuning, we can achieve accurate models in little time and on less amount of data.", "after_revision": "Community Question-Answering websites, such as StackOverflow and Quora, expect users to follow specific guidelines in order to maintain content quality. These systems mainly rely on community reports for assessing contents, which has serious problems such as the slow handling of violations, the loss of normal and experienced users' time, the low quality of some reports, and discouraging feedback to new users. Therefore, with the overall goal of providing solutions for automating moderation actions in Q&A websites, we aim to provide a model to predict 20 quality or subjective aspects of questions in QA websites. To this end, we used data gathered by the CrowdSource team at Google Research in 2019 and a fine-tuned pre-trained BERT model on our problem. Based on the evaluation by Mean-Squared-Error (MSE), the model achieved a value of 0.046 after 2 epochs of training, which did not improve substantially in the next ones. Results confirm that by simple fine-tuning, we can achieve accurate models in little time and on less amount of data.", "edit_actions": [{"type": "A", "before": null, "after": "a", "start_char_pos": 709, "end_char_pos": 709, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 769, "end_char_pos": 769, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "model achieved the", "after": "the model achieved a", "start_char_pos": 810, "end_char_pos": 828, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 152, 412, 618, 759, 925], "domain": "arxiv"}
{"doc_id": "2003.02645", "revision_depth": 1, "before_revision": "We introduce sentenceMIM, a probabilistic auto-encoder for language modelling, trained with Mutual Information Machine (MIM) learning. Previous attempts to learn variational auto-encoders for language data ? have had mixed success, with empirical performance well below state-of-the-art auto-regressive models, a key barrier being the ? occurrence of posterior collapse with VAEs. The recently proposed MIM framework encourages high mutual information between observations and latent variables, and is more robust against posterior collapse. This paper formulates a MIM model for text data, along with a corresponding learning algorithm. We demonstrate excellent perplexity (PPL) results on several datasets, and show that the framework learns a rich latent space, allowing for interpolation between sentences of different lengths with a fixed-dimensional latent representation. We also demonstrate the versatility of sentenceMIM by utilizing a trained model for question-answering, a transfer learning task, without fine-tuning. To the best of our knowledge, this is the first latent variable model (LVM) for text modelling that achieves competitive performance with non-LVM models.", "after_revision": "We introduce sentenceMIM, a probabilistic auto-encoder for language modelling, trained with Mutual Information Machine (MIM) learning. Previous attempts to learn variational auto-encoders for language data have had mixed success, with empirical performance well below state-of-the-art auto-regressive models, a key barrier being the occurrence of posterior collapse with VAEs. The recently proposed MIM framework encourages high mutual information between observations and latent variables, and is more robust against posterior collapse. This paper formulates a MIM model for text data, along with a corresponding learning algorithm. We demonstrate excellent perplexity (PPL) results on several datasets, and show that the framework learns a rich latent space, allowing for interpolation between sentences of different lengths with a fixed-dimensional latent representation. We also demonstrate the versatility of sentenceMIM by utilizing a trained model for question-answering, a transfer learning task, without fine-tuning. To the best of our knowledge, this is the first latent variable model (LVM) for text modelling that achieves competitive performance with non-LVM models.", "edit_actions": [{"type": "D", "before": "?", "after": null, "start_char_pos": 206, "end_char_pos": 207, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "?", "after": null, "start_char_pos": 335, "end_char_pos": 336, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 134, 380, 541, 637, 878, 1029], "domain": "arxiv"}
{"doc_id": "2003.02645", "revision_depth": 2, "before_revision": "We introduce sentenceMIM, a probabilistic auto-encoder for language modelling , trained with Mutual Information Machine (MIM) learning  . Previous attempts to learn variational auto-encoders for language data have had mixed success, with empirical performance well below state-of-the-art auto-regressive models, a key barrier being the occurrence of posterior collapsewith VAEs. The recently proposed MIM framework encourages high mutual information between observations and latent variables, and is more robust against posterior collapse. This paper formulates a MIM model for text data, along with a corresponding learning algorithm. We demonstrate excellent perplexity (PPL) results on several datasets, and show that the framework learns a rich latent space, allowing for interpolation between sentences of different lengths with a fixed-dimensional latent representation. We also demonstrate the versatility of sentenceMIM by utilizing a trained model for question-answering , a transfer learningtask , without fine-tuning . To the best of our knowledge, this is the first latent variable model (LVM) for text modelling that achieves competitive performance with non-LVM models .", "after_revision": "SentenceMIM is a probabilistic auto-encoder for language data , trained with Mutual Information Machine (MIM) learning to provide a fixed length representation of variable length language observations (ie, similar to VAE) . Previous attempts to learn VAEs for language data faced challenges due to posterior collapse. MIM learning encourages high mutual information between observations and latent variables, and is robust against posterior collapse. As such, it learns informative representations whose dimension can be an order of magnitude higher than existing language VAEs. Importantly, the SentenceMIM loss has no hyper-parameters, simplifying optimization. We compare sentenceMIM with VAE, and AE on multiple datasets. SentenceMIM yields excellent reconstruction, comparable to AEs, with a rich structured latent space, comparable to VAEs. The structured latent representation is demonstrated with interpolation between sentences of different lengths . We demonstrate the versatility of sentenceMIM by utilizing a trained model for question-answering and transfer learning , without fine-tuning , outperforming VAE and AE with similar architectures .", "edit_actions": [{"type": "R", "before": "We introduce sentenceMIM,", "after": "SentenceMIM is", "start_char_pos": 0, "end_char_pos": 25, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "modelling", "after": "data", "start_char_pos": 68, "end_char_pos": 77, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "to provide a fixed length representation of variable length language observations (ie, similar to VAE)", "start_char_pos": 135, "end_char_pos": 135, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "variational auto-encoders", "after": "VAEs", "start_char_pos": 165, "end_char_pos": 190, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "have had mixed success, with empirical performance well below state-of-the-art auto-regressive models, a key barrier being the occurrence of posterior collapsewith VAEs. The recently proposed MIM framework", "after": "faced challenges due to posterior collapse. MIM learning", "start_char_pos": 209, "end_char_pos": 414, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "more", "after": null, "start_char_pos": 500, "end_char_pos": 504, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "This paper formulates a MIM model for text data, along with a corresponding learning algorithm. We demonstrate excellent perplexity (PPL) results on several datasets, and show that the framework learns a rich", "after": "As such, it learns informative representations whose dimension can be an order of magnitude higher than existing language VAEs. Importantly, the SentenceMIM loss has no hyper-parameters, simplifying optimization. We compare sentenceMIM with VAE, and AE on multiple datasets. SentenceMIM yields excellent reconstruction, comparable to AEs, with a rich structured", "start_char_pos": 540, "end_char_pos": 748, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "allowing for", "after": "comparable to VAEs. The structured latent representation is demonstrated with", "start_char_pos": 763, "end_char_pos": 775, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "with a fixed-dimensional latent representation. We also", "after": ". We", "start_char_pos": 829, "end_char_pos": 884, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ", a transfer learningtask", "after": "and transfer learning", "start_char_pos": 980, "end_char_pos": 1005, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": ". To the best of our knowledge, this is the first latent variable model (LVM) for text modelling that achieves competitive performance with non-LVM models", "after": ", outperforming VAE and AE with similar architectures", "start_char_pos": 1028, "end_char_pos": 1182, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 378, 539, 635, 876, 1029], "domain": "arxiv"}
{"doc_id": "2004.12316", "revision_depth": 1, "before_revision": "Empathetic conversational models have been shown to improve user satisfaction and task outcomes in numerous domains. In Psychology, persona has been shown to be highly correlated to personality, which in turn influences empathy. In addition, our empirical analysis also suggests that persona plays an important role in empathetic conversations . To this end, we propose a new task towards persona-based empathetic conversations and present the first empirical study on the impacts of persona on empathetic responding. Specifically, we first present a novel large-scale multi-domain dataset for persona-based empathetic conversations . We then propose CoBERT, an efficient BERT-based response selection model that obtains the state-of-the-art performance on our dataset. Finally, we conduct extensive experiments to investigate the impacts of persona on empathetic responding. Notably, our results show that persona improves empathetic responding more when CoBERT is trained on empathetic conversations than non-empathetic ones, establishing an empirical link between persona and empathy in human conversations .", "after_revision": "Empathetic dialogue systems have been shown to improve user satisfaction and task outcomes in numerous domains. In Psychology, persona has been shown to be highly correlated to personality, which in turn influences empathy. In addition, our empirical analysis also suggests that persona plays an important role in empathetic dialogues . To this end, we propose a new task to endow empathetic dialogue systems with personas and present the first empirical study on the impacts of persona on empathetic responding. Specifically, we first present a novel large-scale multi-domain dataset for empathetic dialogues with personas . We then propose CoBERT, an efficient BERT-based response selection model that obtains the state-of-the-art performance on our dataset. Finally, we conduct extensive experiments to investigate the impacts of persona on empathetic responding. Notably, our results show that persona improves empathetic responding more when CoBERT is trained on empathetic dialogues than non-empathetic ones, establishing an empirical link between persona and empathy in human dialogues .", "edit_actions": [{"type": "R", "before": "conversational models", "after": "dialogue systems", "start_char_pos": 11, "end_char_pos": 32, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "conversations", "after": "dialogues", "start_char_pos": 330, "end_char_pos": 343, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "towards persona-based empathetic conversations", "after": "to endow empathetic dialogue systems with personas", "start_char_pos": 381, "end_char_pos": 427, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "persona-based empathetic conversations", "after": "empathetic dialogues with personas", "start_char_pos": 594, "end_char_pos": 632, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "conversations", "after": "dialogues", "start_char_pos": 988, "end_char_pos": 1001, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "conversations", "after": "dialogues", "start_char_pos": 1096, "end_char_pos": 1109, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 116, 228, 345, 517, 634, 769, 875], "domain": "arxiv"}
{"doc_id": "2004.12316", "revision_depth": 2, "before_revision": "Empathetic dialogue systems have been shown to improve user satisfaction and task outcomes in numerous domains. In Psychology, persona has been shown to be highly correlated to personality, which in turn influences empathy. In addition, our empirical analysis also suggests that persona plays an important role in empathetic dialogues . To this end, we propose a new task to endow empathetic dialogue systems with personas and present the first empirical study on the impacts of persona on empathetic responding. Specifically, we first present a novel large-scale multi-domain dataset for empathetic dialogues with personas . We then propose CoBERT, an efficient BERT-based response selection model that obtains the state-of-the-art performance on our dataset. Finally, we conduct extensive experiments to investigate the impacts of persona on empathetic responding. Notably, our results show that persona improves empathetic responding more when CoBERT is trained on empathetic dialogues than non-empathetic ones, establishing an empirical link between persona and empathy in human dialogues .", "after_revision": "Empathetic conversational models have been shown to improve user satisfaction and task outcomes in numerous domains. In Psychology, persona has been shown to be highly correlated to personality, which in turn influences empathy. In addition, our empirical analysis also suggests that persona plays an important role in empathetic conversations . To this end, we propose a new task towards persona-based empathetic conversations and present the first empirical study on the impact of persona on empathetic responding. Specifically, we first present a novel large-scale multi-domain dataset for persona-based empathetic conversations . We then propose CoBERT, an efficient BERT-based response selection model that obtains the state-of-the-art performance on our dataset. Finally, we conduct extensive experiments to investigate the impact of persona on empathetic responding. Notably, our results show that persona improves empathetic responding more when CoBERT is trained on empathetic conversations than non-empathetic ones, establishing an empirical link between persona and empathy in human conversations .", "edit_actions": [{"type": "R", "before": "dialogue systems", "after": "conversational models", "start_char_pos": 11, "end_char_pos": 27, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "dialogues", "after": "conversations", "start_char_pos": 325, "end_char_pos": 334, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "to endow empathetic dialogue systems with personas", "after": "towards persona-based empathetic conversations", "start_char_pos": 372, "end_char_pos": 422, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "impacts", "after": "impact", "start_char_pos": 468, "end_char_pos": 475, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "empathetic dialogues with personas", "after": "persona-based empathetic conversations", "start_char_pos": 589, "end_char_pos": 623, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "impacts", "after": "impact", "start_char_pos": 822, "end_char_pos": 829, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "dialogues", "after": "conversations", "start_char_pos": 979, "end_char_pos": 988, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "dialogues", "after": "conversations", "start_char_pos": 1083, "end_char_pos": 1092, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 111, 223, 336, 512, 625, 760, 866], "domain": "arxiv"}
{"doc_id": "2004.12765", "revision_depth": 1, "before_revision": "Automatic humor detection has interesting use cases in modern technologies, such as chatbots and personal assistants. In this paper, we describe a novel approach for detecting humor in short texts  using BERT sentence embedding. Our proposed model uses BERT to generate tokens and sentence embedding for texts. It sends embedding outputs as input to a two-layered neural networkthat predicts the target value. For evaluation  , we created a new dataset for humor detection consisting of 200k formal short texts (100k positive , 100k negative). Experimental results show an accuracy of 98.1 percent for the proposed method , 2.1 percent improvement compared to the best CNN and RNN models and 1.1 percentbetter than a fine-tuned BERT model. In addition, the combination of RNN-CNN was not successful in this task compared to the CNN model .", "after_revision": "Automatic humor detection has interesting use cases in modern technologies, such as chatbots and virtual assistants. Based on the general linguistic structure of humor, in this paper, we propose a novel approach for detecting humor in short texts by using BERT sentence embedding. Our proposed method uses BERT to generate embeddings for sentences of a given text and uses these embeddings as inputs for parallel lines of hidden layers in a neural network. These lines are finally concatenated to predict the target value. For evaluation purposes , we created a new dataset for humor detection consisting of 200k formal short texts (100k positive and 100k negative). Experimental results show that our proposed method can determine humor in short texts with accuracy and an F1-score of 98.2 percent. Our 8-layer model with 110M parameters outperforms all baseline models with a large margin, showing the importance of utilizing linguistic structure in machine learning models .", "edit_actions": [{"type": "R", "before": "personal assistants. In", "after": "virtual assistants. Based on the general linguistic structure of humor, in", "start_char_pos": 97, "end_char_pos": 120, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "describe", "after": "propose", "start_char_pos": 136, "end_char_pos": 144, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "by", "start_char_pos": 197, "end_char_pos": 197, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "model", "after": "method", "start_char_pos": 242, "end_char_pos": 247, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "tokens and sentence embedding for texts. It sends embedding outputs as input to a two-layered neural networkthat predicts", "after": "embeddings for sentences of a given text and uses these embeddings as inputs for parallel lines of hidden layers in a neural network. These lines are finally concatenated to predict", "start_char_pos": 270, "end_char_pos": 391, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "purposes", "start_char_pos": 425, "end_char_pos": 425, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ",", "after": "and", "start_char_pos": 526, "end_char_pos": 527, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "an accuracy of 98.1 percent for the proposed method , 2.1 percent improvement compared to the best CNN and RNN models and 1.1 percentbetter than a fine-tuned BERT model. In addition, the combination of RNN-CNN was not successful in this task compared to the CNN model", "after": "that our proposed method can determine humor in short texts with accuracy and an F1-score of 98.2 percent. Our 8-layer model with 110M parameters outperforms all baseline models with a large margin, showing the importance of utilizing linguistic structure in machine learning models", "start_char_pos": 570, "end_char_pos": 837, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 117, 228, 310, 409, 543, 739], "domain": "arxiv"}
{"doc_id": "2004.14519", "revision_depth": 2, "before_revision": "Arabic is a morphological rich language, posing many challenges for information extraction (IE) tasks, including Named Entity Recognition (NER) , Part-of-Speech tagging (POS), Argument Role Labeling (ARL), and Relation Extraction (RE). A few multilingual pre-trained models have been proposed and show good performance for Arabic, however, most experiment results are reported on language understanding tasks, such as natural language inference, question answering and sentiment analysis. Their performance on the IEtasks is less known, in particular, the cross-lingual transfer capability from English to Arabic . In this work , we pre-train a Gigaword-based bilingual language model (GigaBERT) to study these two distant languages as well as zero-short transfer learningon various IE tasks. Our GigaBERT outperforms multilingual BERT and and monolingual AraBERT on these tasks, in both supervised and zero-shot learning settings. footnote We have made our pre-trained models publicly available at URL", "after_revision": "Multilingual pre-trained Transformers, such as mBERT (Devlin et al., 2019) and XLM-RoBERTa (Conneau et al., 2020a), have been shown to enable the effective cross-lingual zero-shot transfer. However, their performance on Arabic information extraction (IE) tasks is not very well studied . In this paper , we pre-train a customized bilingual BERT, dubbed GigaBERT, that is designed specifically for Arabic NLP and English-to-Arabic zero-shot transfer learning. We study GigaBERT's effectiveness on zero-short transfer across four IE tasks: named entity recognition, part-of-speech tagging, argument role labeling, and relation extraction. Our best model significantly outperforms mBERT, XLM-RoBERTa, and AraBERT (Antoun et al., 2020) in both the supervised and zero-shot transfer settings. We have made our pre-trained models publicly available at URL", "edit_actions": [{"type": "R", "before": "Arabic is a morphological rich language, posing many challenges for information extraction (IE) tasks, including Named Entity Recognition (NER) , Part-of-Speech tagging (POS), Argument Role Labeling (ARL), and Relation Extraction (RE). A few multilingual", "after": "Multilingual", "start_char_pos": 0, "end_char_pos": 254, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "models have been proposed and show good performance for Arabic, however, most experiment results are reported on language understanding tasks, such as natural language inference, question answering and sentiment analysis. Their performance on the IEtasks is less known, in particular, the", "after": "Transformers, such as mBERT (Devlin et al., 2019) and XLM-RoBERTa (Conneau et al., 2020a), have been shown to enable the effective", "start_char_pos": 267, "end_char_pos": 555, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "transfer capability from English to Arabic", "after": "zero-shot transfer. However, their performance on Arabic information extraction (IE) tasks is not very well studied", "start_char_pos": 570, "end_char_pos": 612, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "work", "after": "paper", "start_char_pos": 623, "end_char_pos": 627, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Gigaword-based bilingual language model (GigaBERT) to study these two distant languages as well as zero-short transfer learningon various IE tasks. Our GigaBERT outperforms multilingual BERT and and monolingual AraBERT on these tasks, in both", "after": "customized bilingual BERT, dubbed GigaBERT, that is designed specifically for Arabic NLP and English-to-Arabic zero-shot transfer learning. We study GigaBERT's effectiveness on zero-short transfer across four IE tasks: named entity recognition, part-of-speech tagging, argument role labeling, and relation extraction. Our best model significantly outperforms mBERT, XLM-RoBERTa, and AraBERT (Antoun et al., 2020) in both the", "start_char_pos": 645, "end_char_pos": 887, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "learning settings. footnote", "after": "transfer settings.", "start_char_pos": 913, "end_char_pos": 940, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 235, 488, 614, 792], "domain": "arxiv"}
{"doc_id": "2004.14601", "revision_depth": 1, "before_revision": "We propose a novel methodology for analyzing the encoding of grammatical structure in neural language models through transfer learning. We test how a language model can leverage its internal representations to transfer knowledge across languages and symbol systems. We train LSTMs on non-linguistic , structured data and test their performance on human language to assess which kinds of data induce generalizable encodings that LSTMs can use for natural language. We find that models trained on structured data such as music and Java codehave internal representations that help in modelling human language, and that, surprisingly, adding minimal amounts of structure to the training data makes a large difference in transfer to natural language . Further experiments on transfer between human languages show that zero-shot performance on a test language is highly correlated with  syntactic similarity to the training language, even after removing any vocabulary overlap. This suggests that the internal representations induced from natural languages are typologically coherent: they encode the features and differences outlined in typological studies . Our results provide insights into how neural networks represent linguistic structure, and also about the kinds of structural biases that give learners the ability to model language.", "after_revision": "We propose transfer learning as a method for analyzing the encoding of grammatical structure in neural language models . We train LSTMs on non-linguistic data and evaluate their performance on natural language to assess which kinds of data induce generalizable structural features that LSTMs can use for natural language. We find that training on non-linguistic data with latent structure (MIDI music or Java code) improves test performance on natural language, despite no overlap in surface form or vocabulary. Training on artificial languages containing recursion (hierarchical structure) also improves performance on natural language, again with no vocabulary overlap. Surprisingly, training on artificial languages consisting of sets of separated pairs of words, but with no recursion, improves performance on natural language as well as recursive languages do. Experiments on transfer between natural languages show that zero-shot performance on a test language is highly correlated with typological syntactic similarity to the training language, suggesting that representations induced from natural languages correspond to the cross-linguistic syntactic properties studied in linguistic typology . Our results provide insights into the ways that neural models represent abstract syntactic structure, and also about the kind of structural inductive biases which a learner needs to model language.", "edit_actions": [{"type": "R", "before": "a novel methodology", "after": "transfer learning as a method", "start_char_pos": 11, "end_char_pos": 30, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "through transfer learning. We test how a language model can leverage its internal representations to transfer knowledge across languages and symbol systems. We", "after": ". We", "start_char_pos": 109, "end_char_pos": 268, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": ", structured data and test", "after": "data and evaluate", "start_char_pos": 299, "end_char_pos": 325, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "human", "after": "natural", "start_char_pos": 347, "end_char_pos": 352, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "encodings", "after": "structural features", "start_char_pos": 413, "end_char_pos": 422, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "models trained on structured data such as music and Java codehave internal representations that help in modelling human language, and that, surprisingly, adding minimal amounts of structure to the training data makes a large difference in transfer to natural language . Further experiments", "after": "training on non-linguistic data with latent structure (MIDI music or Java code) improves test performance on natural language, despite no overlap in surface form or vocabulary. Training on artificial languages containing recursion (hierarchical structure) also improves performance on natural language, again with no vocabulary overlap. Surprisingly, training on artificial languages consisting of sets of separated pairs of words, but with no recursion, improves performance on natural language as well as recursive languages do. Experiments", "start_char_pos": 477, "end_char_pos": 766, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "human", "after": "natural", "start_char_pos": 787, "end_char_pos": 792, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "typological", "start_char_pos": 880, "end_char_pos": 880, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "even after removing any vocabulary overlap. This suggests that the internal", "after": "suggesting that", "start_char_pos": 928, "end_char_pos": 1003, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "are typologically coherent: they encode the features and differences outlined in typological studies", "after": "correspond to the cross-linguistic syntactic properties studied in linguistic typology", "start_char_pos": 1051, "end_char_pos": 1151, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "how neural networks represent linguistic", "after": "the ways that neural models represent abstract syntactic", "start_char_pos": 1188, "end_char_pos": 1228, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "kinds of structural biases that give learners the ability", "after": "kind of structural inductive biases which a learner needs", "start_char_pos": 1259, "end_char_pos": 1316, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 135, 265, 463, 746, 971, 1153], "domain": "arxiv"}
{"doc_id": "2004.14601", "revision_depth": 2, "before_revision": "We propose transfer learning as a method for analyzing the encoding of grammatical structure in neural language models. We train LSTMs on non-linguistic data and evaluate their performance on natural language to assess which kinds of data induce generalizable structural features that LSTMs can use for natural language. We find that training on non-linguistic data with latent structure (MIDI music or Java code) improves test performance on natural language, despite no overlap in surface form or vocabulary. Training on artificial languages containing recursion (hierarchical structure) also improves performance on natural language, again with no vocabulary overlap . Surprisingly, training  on artificial languages consisting of sets of separated pairs of words, but with no recursion, improves performance on natural language as well as recursive languages do. Experiments on transfer between natural languages  show that zero-shot performance on a test language is highly correlated with typological syntactic similarity to the training language, suggesting that representations induced from natural languages correspond to the cross-linguistic syntactic properties studied in linguistic typology . Our results provide insights into the ways that neural models represent abstract syntactic structure, and also about the kind of structural inductive biases which a learner needs to model language .", "after_revision": "We propose transfer learning as a method for analyzing the encoding of grammatical structure in neural language models. We train LSTMs on non-linguistic data and evaluate their performance on natural language to assess which kinds of data induce generalizable structural features that LSTMs can use for natural language. We find that training on non-linguistic data with latent structure (MIDI music or Java code) improves test performance on natural language, despite no overlap in surface form or vocabulary. To pinpoint the kinds of abstract structure that models may be encoding to lead to this improvement, we run similar experiments with two artificial parentheses languages: one which has a hierarchical recursive structure, and a control which has paired tokens but no recursion . Surprisingly, training a model on either of these artificial languages leads to the same substantial gains when testing on natural language . Further experiments on transfer between natural languages controlling for vocabulary overlap show that zero-shot performance on a test language is highly correlated with typological syntactic similarity to the training language, suggesting that representations induced by pre-training correspond to the cross-linguistic syntactic properties . Our results provide insights into the ways that neural models represent abstract syntactic structure, and also about the kind of structural inductive biases which allow for natural language acquisition .", "edit_actions": [{"type": "R", "before": "Training on artificial languages containing recursion (hierarchical structure) also improves performance on natural language, again with no vocabulary overlap", "after": "To pinpoint the kinds of abstract structure that models may be encoding to lead to this improvement, we run similar experiments with two artificial parentheses languages: one which has a hierarchical recursive structure, and a control which has paired tokens but no recursion", "start_char_pos": 511, "end_char_pos": 669, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "a model on either of these artificial languages leads to the same substantial gains when testing", "start_char_pos": 695, "end_char_pos": 695, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "artificial languages consisting of sets of separated pairs of words, but with no recursion, improves performance on", "after": null, "start_char_pos": 699, "end_char_pos": 814, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "as well as recursive languages do. Experiments", "after": ". Further experiments", "start_char_pos": 832, "end_char_pos": 878, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "controlling for vocabulary overlap", "start_char_pos": 917, "end_char_pos": 917, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "from natural languages", "after": "by pre-training", "start_char_pos": 1094, "end_char_pos": 1116, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "studied in linguistic typology", "after": null, "start_char_pos": 1173, "end_char_pos": 1203, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "a learner needs to model language", "after": "allow for natural language acquisition", "start_char_pos": 1369, "end_char_pos": 1402, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}], "sents_char_pos": [0, 119, 320, 510, 866, 1205], "domain": "arxiv"}
{"doc_id": "2004.14623", "revision_depth": 2, "before_revision": "In adversarial testing, we pose hard generalization tasks in order to gain insights into the solutions found by our models. What properties must a system have in order to succeed at these hard behavioral tasks? We argue that an essential factor is modular internal structure. Our central contribution is a new experimental method called 'interchange interventions', in which systematic manipulations of model-internal states are related to causal effects on their outputs, thereby allowing us to identify modular structure. Our work is grounded empirically in a new challenge Natural Language Inference dataset designed to assess systems on their ability to reason about entailment and negation. We find that a BERT model is strikingly successful at the systematic generalization task we pose using this dataset , and our active manipulations of model-internal vectors help us understand why: despite the densely interconnected nature of the BERT architecture, the learned model embeds modular, general theories of lexical entailment relations .", "after_revision": "We address whether neural models for Natural Language Inference (NLI) can learn the compositional interactions between lexical entailment and negation, using four methods: the behavioral evaluation methods of (1) challenge test sets and (2) systematic generalization tasks, and the structural evaluation methods of (3) probes and (4) interventions. To facilitate this holistic evaluation, we present Monotonicity NLI (MoNLI), a new naturalistic dataset focused on lexical entailment and negation. In our behavioral evaluations, we find that models trained on general-purpose NLI datasets fail systematically on MoNLI examples containing negation, but that MoNLI fine-tuning addresses this failure. In our structural evaluations, we look for evidence that our top-performing BERT-based model has learned to implement the monotonicity algorithm behind MoNLI. Probes yield evidence consistent with this conclusion , and our intervention experiments bolster this, showing that the causal dynamics of the model mirror the causal dynamics of this algorithm on subsets of MoNLI. This suggests that the BERT model at least partially embeds a theory of lexical entailment and negation at an algorithmic level .", "edit_actions": [{"type": "R", "before": "In adversarial testing, we pose hard generalization tasks in order to gain insights into the solutions found by our models. What properties must a system have in order to succeed at these hard behavioral tasks? We argue that an essential factor is modular internal structure. Our central contribution is a new experimental method called 'interchange interventions', in which systematic manipulations of model-internal states are related to causal effects on their outputs, thereby allowing us to identify modular structure. Our work is grounded empirically in a new challenge Natural Language Inference dataset designed to assess systems on their ability to reason about", "after": "We address whether neural models for Natural Language Inference (NLI) can learn the compositional interactions between lexical entailment and negation, using four methods: the behavioral evaluation methods of (1) challenge test sets and (2) systematic generalization tasks, and the structural evaluation methods of (3) probes and (4) interventions. To facilitate this holistic evaluation, we present Monotonicity NLI (MoNLI), a new naturalistic dataset focused on lexical", "start_char_pos": 0, "end_char_pos": 670, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "We find that a BERT model is strikingly successful at the systematic generalization task we pose using this dataset", "after": "In our behavioral evaluations, we find that models trained on general-purpose NLI datasets fail systematically on MoNLI examples containing negation, but that MoNLI fine-tuning addresses this failure. In our structural evaluations, we look for evidence that our top-performing BERT-based model has learned to implement the monotonicity algorithm behind MoNLI. Probes yield evidence consistent with this conclusion", "start_char_pos": 696, "end_char_pos": 811, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "active manipulations of model-internal vectors help us understand why: despite the densely interconnected nature of", "after": "intervention experiments bolster this, showing that the causal dynamics of", "start_char_pos": 822, "end_char_pos": 937, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "BERT architecture, the learned model embeds modular, general theories", "after": "model mirror the causal dynamics of this algorithm on subsets of MoNLI. This suggests that the BERT model at least partially embeds a theory", "start_char_pos": 942, "end_char_pos": 1011, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "relations", "after": "and negation at an algorithmic level", "start_char_pos": 1034, "end_char_pos": 1043, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}], "sents_char_pos": [0, 123, 210, 275, 523, 695], "domain": "arxiv"}
{"doc_id": "2004.14974", "revision_depth": 1, "before_revision": "We introduce the task of scientific fact-checking. Given a corpus of scientific articles and a claim about a scientific finding, a fact-checking model must identify abstracts that support or refute the claim. In addition, it must provide rationales for its predictions in the form of evidentiary sentences from the retrieved abstracts. For this task, we introduce SciFact, a dataset of 1.4K expert-written scientific claims paired with evidence-containing abstracts , and annotated with labels and rationales. We present a baseline model and assess its performance on SciFact. We observe that, while fact-checking models trained on Wikipedia articles or political news have difficulty generalizing to our task, simple domain adaptation techniques represent a promising avenue for improvement. Finally, we provide initial results showing how our model can be used to verify claims relevant to COVID-19 on the CORD-19 corpus. Our dataset will be made publicly available at URL", "after_revision": "We introduce scientific claim verification, a new task to select abstracts from the research literature containing evidence that supports or refutes a given scientific claim, and to identify rationales justifying each decision. To study this task, we construct SciFact, a dataset of 1.4K expert-written scientific claims paired with evidence-containing abstracts annotated with labels and rationales. We develop baseline models for SciFact, and demonstrate that these models benefit from combined training on a large dataset of claims about Wikipedia articles, together with the new SciFact data. We show that our claim verification system is able to identify plausible evidence for 23 / 36 claims relevant to COVID-19 on the CORD-19 corpus. Our results and experiments strongly suggest that our new task and data will support significant future research efforts.", "edit_actions": [{"type": "R", "before": "the task of scientific fact-checking. Given a corpus of scientific articles and a claim about a scientific finding, a fact-checking model must identify abstracts that support or refute the claim. In addition, it must provide rationales for its predictions in the form of evidentiary sentences from the retrieved abstracts. For", "after": "scientific claim verification, a new task to select abstracts from the research literature containing evidence that supports or refutes a given scientific claim, and to identify rationales justifying each decision. To study", "start_char_pos": 13, "end_char_pos": 339, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "introduce", "after": "construct", "start_char_pos": 354, "end_char_pos": 363, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "D", "before": ", and", "after": null, "start_char_pos": 466, "end_char_pos": 471, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "present a baseline model and assess its performance on SciFact. We observe that, while fact-checking models trained on Wikipedia articles or political news have difficulty generalizing to our task, simple domain adaptation techniques represent a promising avenue for improvement. Finally, we provide initial results showing how our model can be used to verify", "after": "develop baseline models for SciFact, and demonstrate that these models benefit from combined training on a large dataset of claims about Wikipedia articles, together with the new SciFact data. We show that our claim verification system is able to identify plausible evidence for 23 / 36", "start_char_pos": 513, "end_char_pos": 872, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "dataset will be made publicly available at URL", "after": "results and experiments strongly suggest that our new task and data will support significant future research efforts.", "start_char_pos": 928, "end_char_pos": 974, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 50, 208, 335, 509, 576, 792, 923], "domain": "arxiv"}
{"doc_id": "2004.15003", "revision_depth": 1, "before_revision": "One key principle for assessing semantic similarity between texts is to measure the degree of semantic overlap of them by considering word-by-word alignment. However, alignment-based approaches are inferior to the generic sentence vectorsin terms of performance. We hypothesize that the reason for the inferiority of alignment-based methods is due to the fact that they do not distinguish word importance and word meaning. To solve this , we propose to separate word importance and word meaning by decomposing word vectors into their norm and direction , then compute the alignment-based similarity with the help of earth mover's distance . We call the method word rotator's distance (WRD) because direction vectors are aligned by rotation on the unit hypersphere. In addition, to incorporate the advance of cutting edge additive sentence encoders, we propose to re-decompose such sentence vectors into word vectors and use them as inputs to WRD. Empirically, the proposed method outperforms current methods considering the word-by-word alignment including word mover's distance with a big difference; moreover, our method outperforms state-of-the-art additive sentence encoders on the most competitive dataset, STS-benchmark .", "after_revision": "One key principle for assessing textual similarity is measuring the degree of semantic overlap between two texts by considering the word alignment. Such alignment-based approaches are both intuitive and interpretable; however, they are empirically inferior to the simple cosine similarity between general-purpose sentence vectors. To remedy this, we focus on the fact that the norm of word vectors is a good proxy for word importance, and the angle of them is a good proxy for word similarity. Alignment-based approaches do not distinguish the norm and direction, whereas sentence-vector approaches automatically use the norm as the word importance. Accordingly , we propose to decouple word vectors into their norm and direction then computing the alignment-based similarity using earth mover's distance (optimal transport cost), which we refer to as word rotator's distance . Furthermore, we demonstrate how to grow the norm and direction of word vectors (vector converter); this is a new systematic approach derived from the sentence-vector estimation methods, which can significantly improve the performance of the proposed method. On several STS benchmarks, our simple proposed methods outperformed not only alignment-based approaches but also strong baselines .", "edit_actions": [{"type": "R", "before": "semantic similarity between texts is to measure", "after": "textual similarity is measuring", "start_char_pos": 32, "end_char_pos": 79, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "of them by considering word-by-word alignment. However,", "after": "between two texts by considering the word alignment. Such", "start_char_pos": 111, "end_char_pos": 166, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "are", "after": "are both intuitive and interpretable; however, they are empirically", "start_char_pos": 194, "end_char_pos": 197, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "generic sentence vectorsin terms of performance. We hypothesize that the reason for the inferiority of alignment-based methods is due to the fact that they", "after": "simple cosine similarity between general-purpose sentence vectors. To remedy this, we focus on the fact that the norm of word vectors is a good proxy for word importance, and the angle of them is a good proxy for word similarity. Alignment-based approaches", "start_char_pos": 214, "end_char_pos": 369, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "word importance and word meaning. To solve this", "after": "the norm and direction, whereas sentence-vector approaches automatically use the norm as the word importance. Accordingly", "start_char_pos": 389, "end_char_pos": 436, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "separate word importance and word meaning by decomposing word", "after": "decouple word", "start_char_pos": 453, "end_char_pos": 514, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": ", then compute", "after": "then computing", "start_char_pos": 553, "end_char_pos": 567, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "with the help of", "after": "using", "start_char_pos": 599, "end_char_pos": 615, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": ". We call the method", "after": "(optimal transport cost), which we refer to as", "start_char_pos": 639, "end_char_pos": 659, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "(WRD) because direction vectors are aligned by rotation on the unit hypersphere. In addition, to incorporate the advance of cutting edge additive sentence encoders, we propose to re-decompose such sentence vectors into word vectors and use them as inputs to WRD. Empirically, the proposed method outperforms current methods considering the word-by-word alignment including word mover's distance with a big difference; moreover, our method outperforms state-of-the-art additive sentence encoders on the most competitive dataset, STS-benchmark", "after": ". Furthermore, we demonstrate how to grow the norm and direction of word vectors (vector converter); this is a new systematic approach derived from the sentence-vector estimation methods, which can significantly improve the performance of the proposed method. On several STS benchmarks, our simple proposed methods outperformed not only alignment-based approaches but also strong baselines", "start_char_pos": 684, "end_char_pos": 1225, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 157, 262, 422, 640, 764, 946, 1101], "domain": "arxiv"}
{"doc_id": "2004.15003", "revision_depth": 2, "before_revision": "One key principle for assessing textual similarity is measuring the degree of semantic overlap between two texts by considering the word alignment. Such alignment-based approaches are both intuitive and interpretable; however, they are empirically inferior to the simple cosine similarity between general-purpose sentence vectors. To remedy this , we focus on  the fact that the norm of word vectors is a good proxy for word importance, and the angle of them is a good proxy for word similarity. Alignment-based approaches do not distinguish the norm and direction , whereas sentence-vector approaches automatically use the norm as the word importance. Accordingly, we propose to decouple word vectors into their norm and direction then computing the alignment-based similarity using earth mover's distance (  optimal transport cost), which we refer to as word rotator's distance. Furthermore, we demonstrate how to grow the norm and direction of word vectors (vector converter) ; this is a new systematic approach derived from the sentence-vector estimation methods , which can significantly improve the performance of the proposed method . On several STS benchmarks, our simple proposed methods outperformed not only alignment-based approaches but also strong baselines. ", "after_revision": "A key principle in assessing textual similarity is measuring the degree of semantic overlap between two texts by considering the word alignment. Such alignment-based approaches are intuitive and interpretable; however, they are empirically inferior to the simple cosine similarity between general-purpose sentence vectors. To address this issue , we focus on and demonstrate the fact that the norm of word vectors is a good proxy for word importance, and their angle is a good proxy for word similarity. Alignment-based approaches do not distinguish them , whereas sentence-vector approaches automatically use the norm as the word importance. Accordingly, we propose a method that first decouples word vectors into their norm and direction , and then computes alignment-based similarity using earth mover's distance ( i.e., optimal transport cost), which we refer to as word rotator's distance. Besides, we find how to grow the norm and direction of word vectors (vector converter) , which is a new systematic approach derived from sentence-vector estimation methods . On several textual similarity datasets, the combination of these simple proposed methods outperformed not only alignment-based approaches but also strong baselines. The source code is available at URL", "edit_actions": [{"type": "R", "before": "One key principle for", "after": "A key principle in", "start_char_pos": 0, "end_char_pos": 21, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "D", "before": "both", "after": null, "start_char_pos": 184, "end_char_pos": 188, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "remedy this", "after": "address this issue", "start_char_pos": 334, "end_char_pos": 345, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "and demonstrate", "start_char_pos": 360, "end_char_pos": 360, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "the angle of them", "after": "their angle", "start_char_pos": 441, "end_char_pos": 458, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the norm and direction", "after": "them", "start_char_pos": 542, "end_char_pos": 564, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "to decouple", "after": "a method that first decouples", "start_char_pos": 677, "end_char_pos": 688, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "then computing the", "after": ", and then computes", "start_char_pos": 732, "end_char_pos": 750, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "i.e.,", "start_char_pos": 809, "end_char_pos": 809, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "coherence"]}, {"type": "R", "before": "Furthermore, we demonstrate", "after": "Besides, we find", "start_char_pos": 881, "end_char_pos": 908, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "; this", "after": ", which", "start_char_pos": 979, "end_char_pos": 985, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 1028, "end_char_pos": 1031, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "D", "before": ", which can significantly improve the performance of the proposed method", "after": null, "start_char_pos": 1067, "end_char_pos": 1139, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "STS benchmarks, our", "after": "textual similarity datasets, the combination of these", "start_char_pos": 1153, "end_char_pos": 1172, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "The source code is available at URL", "start_char_pos": 1273, "end_char_pos": 1273, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 147, 217, 330, 495, 652, 880, 980, 1141], "domain": "arxiv"}
{"doc_id": "2004.15011", "revision_depth": 1, "before_revision": "We introduce TLDR generation for scientific papers, a new automatic summarization task with high source compression  requiring expert background knowledge and complex language understanding. To facilitate research on this task, we introduce SciTLDR, a dataset of 3.9K TLDRs. Furthermore, we introduce a novel annotation protocol for scalably curating additional gold summaries by rewriting peer review comments. We use this protocol to augment our test set, yielding multiple gold TLDRs for evaluation, which is unlike most recent summarization datasets that assume only one valid gold summary. We present a training strategy for adapting pretrained language models that exploits similarities between TLDR generation and the related tasks of extreme summarization and title generation, which outperforms strong extractive and abstractive summarization baselines.", "after_revision": "We introduce TLDR generation for scientific papers, a new automatic summarization task with high source compression , requiring expert background knowledge and complex language understanding. To facilitate research on this task, we introduce SciTLDR, a dataset of 3.9K TLDRs. Furthermore, we introduce a novel annotation protocol for scalably curating additional gold summaries by rewriting peer review comments. We use this protocol to augment our test set, yielding multiple gold TLDRs for evaluation, which is unlike most recent summarization datasets that assume only one valid gold summary. We present a training strategy for adapting pretrained language models that exploits similarities between TLDR generation and the related task of title generation, which outperforms strong extractive and abstractive summarization baselines.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 116, "end_char_pos": 116, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "tasks of extreme summarization and", "after": "task of", "start_char_pos": 733, "end_char_pos": 767, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 190, 274, 411, 594], "domain": "arxiv"}
{"doc_id": "2004.15011", "revision_depth": 2, "before_revision": "We introduce TLDR generation for scientific papers , a new automatic summarizationtask with high source compression , requiring expert background knowledge and complex language understanding . To facilitate research on this task, we introduce SciTLDR, a dataset of 3.9K TLDRs . Furthermore, we introduce a novel annotation protocol for scalably curating additional gold summaries by rewriting peer review comments. We use this protocol to augment our test set, yielding multiple gold TLDRs for evaluation, which is unlike most recent summarization datasets that assume only one valid gold summary. We present a training strategy for adapting pretrained language models that exploits similarities between TLDR generation and the related task of title generation, which outperforms strong extractive and abstractive summarization baselines .", "after_revision": "We introduce TLDR generation , a new form of extreme summarization, for scientific papers. TLDR generation involves high source compression and requires expert background knowledge and understanding of complex domain-specific language . To facilitate study on this task, we introduce SciTLDR, a new multi-target dataset of 5.4K TLDRs over 3.2K papers. SciTLDR contains both author-written and expert-derived TLDRs, where the latter are collected using a novel annotation protocol that produces high-quality summaries while minimizing annotation burden. We propose CATTS, a simple yet effective learning strategy for generating TLDRs that exploits titles as an auxiliary training signal. CATTS improves upon strong baselines under both automated metrics and human evaluations. Data and code are publicly available at URL", "edit_actions": [{"type": "D", "before": "for scientific papers", "after": null, "start_char_pos": 29, "end_char_pos": 50, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "automatic summarizationtask with", "after": "form of extreme summarization, for scientific papers. TLDR generation involves", "start_char_pos": 59, "end_char_pos": 91, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", requiring", "after": "and requires", "start_char_pos": 116, "end_char_pos": 127, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "complex language understanding", "after": "understanding of complex domain-specific language", "start_char_pos": 160, "end_char_pos": 190, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "research", "after": "study", "start_char_pos": 207, "end_char_pos": 215, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "dataset of 3.9K TLDRs . Furthermore, we introduce", "after": "new multi-target dataset of 5.4K TLDRs over 3.2K papers. SciTLDR contains both author-written and expert-derived TLDRs, where the latter are collected using", "start_char_pos": 254, "end_char_pos": 303, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "for scalably curating additional gold summaries by rewriting peer review comments. We use this protocol to augment our test set, yielding multiple gold TLDRs for evaluation, which is unlike most recent summarization datasets that assume only one valid gold summary. We present a training strategy for adapting pretrained language models that exploits similarities between TLDR generation and the related task of title generation, which outperforms strong extractive and abstractive summarization baselines .", "after": "that produces high-quality summaries while minimizing annotation burden. We propose CATTS, a simple yet effective learning strategy for generating TLDRs that exploits titles as an auxiliary training signal. CATTS improves upon strong baselines under both automated metrics and human evaluations. Data and code are publicly available at URL", "start_char_pos": 332, "end_char_pos": 839, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 192, 414, 597], "domain": "arxiv"}
{"doc_id": "2005.00192", "revision_depth": 2, "before_revision": "In the automatic evaluation of generative question answering (GenQA) systems, it is difficult to assess the correctness of generated answers due to the free-form of the answer. Moreover, there is a lack of benchmark datasets to evaluate the suitability of existing metrics in terms of correctness. To study a better metric for GenQA, we first create high-quality human judgments of correctness on two standard GenQA datasets. Using our human-evaluation datasets, we show that widely used n-gram similarity metrics do not correlate with human judgments . To alleviate this problem, we propose  a new metric for evaluating the correctness of GenQA. Specifically, our new metric assigns different weights to each token via keyphrase prediction, thereby judging whether a generated answer sentence captures the key meaning of the reference answer. Our proposed metric shows a significantly higher correlation with human judgments than existing metrics in various datasets.", "after_revision": "In the automatic evaluation of generative question answering (GenQA) systems, it is difficult to assess the correctness of generated answers due to the free-form of the answer. Especially, widely used n-gram similarity metrics often fail to discriminate the incorrect answers since they equally consider all of the tokens . To alleviate this problem, we propose KPQA-metric, a new metric for evaluating the correctness of GenQA. Specifically, our new metric assigns different weights to each token via keyphrase prediction, thereby judging whether a generated answer sentence captures the key meaning of the reference answer. To evaluate our metric, we create high-quality human judgments of correctness on two GenQA datasets. Using our human-evaluation datasets, we show that our proposed metric has a significantly higher correlation with human judgments than existing metrics . The code is available at URL", "edit_actions": [{"type": "R", "before": "Moreover, there is a lack of benchmark datasets to evaluate the suitability of existing metrics in terms of correctness. To study a better metric for GenQA, we first create high-quality human judgments of correctness on two standard GenQA datasets. Using our human-evaluation datasets, we show that", "after": "Especially,", "start_char_pos": 177, "end_char_pos": 475, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "do not correlate with human judgments", "after": "often fail to discriminate the incorrect answers since they equally consider all of the tokens", "start_char_pos": 514, "end_char_pos": 551, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "KPQA-metric,", "start_char_pos": 592, "end_char_pos": 592, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Our proposed metric shows", "after": "To evaluate our metric, we create high-quality human judgments of correctness on two GenQA datasets. Using our human-evaluation datasets, we show that our proposed metric has", "start_char_pos": 844, "end_char_pos": 869, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in various datasets.", "after": ". The code is available at URL", "start_char_pos": 948, "end_char_pos": 968, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 176, 297, 425, 553, 646, 843], "domain": "arxiv"}
{"doc_id": "2005.00782", "revision_depth": 1, "before_revision": "Pre-trained language models (PTLM) have greatly improved performance on commonsense inference benchmarks, however, it remains unclear whether they share a human's ability to consistently make correct inferences under perturbations. Prior studies of PTLMs have found inference deficits, but have failed to provide a systematic means of understanding whether these deficits are due to low inference abilities or poor inference robustness . In this work, we address this gap by developing a procedure that allows for the systematized probing of both PTLMs' inference abilities and robustness. Our procedure centers around the methodical creation of logically-equivalent, but syntactically-different sets of probes, of which we create a corpus of 14,400 probes coming from 60 logically-equivalent sets that can be used to probe PTLMs in three task settings. We find that despite the recent success of large PTLMs on commonsense benchmarks, their performances on our probes are no better than random guessing (even with fine-tuning) and are heavily dependent on biases--the poor overall performance, unfortunately, inhibits us from studying robustness. We hope our approach and initial probe set will assist future work in improving PTLMs' inference abilities , while also providing a probing set to test robustness under several linguistic variations--code and data will be released .", "after_revision": "Pre-trained language models (PTLM) have impressive performance on commonsense inference benchmarks, but their ability to practically employ commonsense to communicate with humans is fiercely debated. Prior evaluations of PTLMs have focused on factual world knowledge or the ability to reason when the necessary knowledge is provided explicitly. However, effective communication with humans requires inferences based on implicit commonsense relationships, and robustness despite paraphrasing. In the pursuit of advancing fluid human-AI communication, we propose a new challenge, RICA, that evaluates the capabilities of making commonsense inferences and the robustness of these inferences to language variations. In our work, we develop a systematic procedure to probe PTLMs across three different evaluation settings. Extensive experiments on our generated probe sets show that PTLMs perform no better than random guessing (even with fine-tuning) , are heavily impacted by statistical biases, and are not robust to perturbation attacks. Our framework and probe sets can help future work improve PTLMs' inference abilities and robustness to linguistic variations--bringing us closer to more fluid communication .", "edit_actions": [{"type": "R", "before": "greatly improved", "after": "impressive", "start_char_pos": 40, "end_char_pos": 56, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "however, it remains unclear whether they share a human's ability to consistently make correct inferences under perturbations. Prior studies", "after": "but their ability to practically employ commonsense to communicate with humans is fiercely debated. Prior evaluations", "start_char_pos": 106, "end_char_pos": 245, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "found inference deficits, but have failed to provide a systematic means of understanding whether these deficits are due to low inference abilities or poor inference robustness . In this work, we address this gap by developing a procedure that allows for the systematized probing of both PTLMs' inference abilities and robustness. Our procedure centers around the methodical creation of logically-equivalent, but syntactically-different sets of probes, of which we create a corpus of 14,400 probes coming from 60 logically-equivalent sets that can be used", "after": "focused on factual world knowledge or the ability to reason when the necessary knowledge is provided explicitly. However, effective communication with humans requires inferences based on implicit commonsense relationships, and robustness despite paraphrasing. In the pursuit of advancing fluid human-AI communication, we propose a new challenge, RICA, that evaluates the capabilities of making commonsense inferences and the robustness of these inferences to language variations. In our work, we develop a systematic procedure", "start_char_pos": 260, "end_char_pos": 814, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in three task settings. We find that despite the recent success of large PTLMs on commonsense benchmarks, their performances on our probes are", "after": "across three different evaluation settings. Extensive experiments on our generated probe sets show that PTLMs perform", "start_char_pos": 830, "end_char_pos": 972, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "and are heavily dependent on biases--the poor overall performance, unfortunately, inhibits us from studying robustness. We hope our approach and initial probe set will assist future work in improving", "after": ", are heavily impacted by statistical biases, and are not robust to perturbation attacks. Our framework and probe sets can help future work improve", "start_char_pos": 1028, "end_char_pos": 1227, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": ", while also providing a probing set to test robustness under several linguistic variations--code and data will be released", "after": "and robustness to linguistic variations--bringing us closer to more fluid communication", "start_char_pos": 1255, "end_char_pos": 1378, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 231, 437, 589, 853, 1147]}
{"doc_id": "2005.00782", "revision_depth": 2, "before_revision": "Pre-trained language models ( PTLM) have impressive performance on commonsense inference benchmarks, but their ability to practically employ commonsense to communicate with humans is fiercely debated. Prior evaluations of PTLMs have focused on factual world knowledge or the ability to reason when the necessary knowledge is provided explicitly. However, effective communication with humansrequires inferences based on implicit commonsense relationships, and robustness despite paraphrasing . In the pursuit of advancing fluid human-AI communication, we propose a new challenge, RICA  , that evaluates the capabilities of making commonsense inferences and the robustness of these inferences to language variations. In our work , we develop a systematic procedure to probe PTLMs across three different evaluation settings. Extensive experiments on our generated probe sets  show that PTLMs perform no better than random guessing (even with fine-tuning) , are heavily impacted by statistical biases, and are not robust to perturbation attacks. Our framework and probe sets can help future work improve PTLMs ' inference abilities and robustness to linguistic variations--bringing us closer to more fluid communication .", "after_revision": "Pre-trained language models ( PTLMs) have achieved impressive performance on commonsense inference benchmarks, but their ability to employ commonsense to make robust inferences, which is crucial for effective communications with humans, is debated . In the pursuit of advancing fluid human-AI communication, we propose a new challenge, RICA : Robust Inference capability based on Commonsense Axioms , that evaluates robust commonsense inference despite textual perturbations. To generate data for this challenge , we develop a systematic and scalable procedure using commonsense knowledge bases and probe PTLMs across two different evaluation settings. Extensive experiments on our generated probe sets with more than 10k statements show that PTLMs perform no better than random guessing on the zero-shot setting , are heavily impacted by statistical biases, and are not robust to perturbation attacks. We also find that fine-tuning on similar statements offer limited gains, as PTLMs still fail to generalize to unseen inferences. Our new large-scale benchmark exposes a significant gap between PTLMs and human-level language understanding and offers a new challenge for PTLMs to demonstrate commonsense .", "edit_actions": [{"type": "R", "before": "PTLM) have", "after": "PTLMs) have achieved", "start_char_pos": 30, "end_char_pos": 40, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "D", "before": "practically", "after": null, "start_char_pos": 122, "end_char_pos": 133, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "communicate with humans is fiercely debated. Prior evaluations of PTLMs have focused on factual world knowledge or the ability to reason when the necessary knowledge is provided explicitly. However, effective communication with humansrequires inferences based on implicit commonsense relationships, and robustness despite paraphrasing", "after": "make robust inferences, which is crucial for effective communications with humans, is debated", "start_char_pos": 156, "end_char_pos": 490, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ": Robust Inference capability based on Commonsense Axioms", "start_char_pos": 584, "end_char_pos": 584, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the capabilities of making commonsense inferences and the robustness of these inferences to language variations. In our work", "after": "robust commonsense inference despite textual perturbations. To generate data for this challenge", "start_char_pos": 602, "end_char_pos": 726, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "procedure to", "after": "and scalable procedure using commonsense knowledge bases and", "start_char_pos": 753, "end_char_pos": 765, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "three", "after": "two", "start_char_pos": 785, "end_char_pos": 790, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "with more than 10k statements", "start_char_pos": 872, "end_char_pos": 872, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "(even with fine-tuning)", "after": "on the zero-shot setting", "start_char_pos": 928, "end_char_pos": 951, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "Our framework and probe sets can help future work improve PTLMs ' inference abilities and robustness to linguistic variations--bringing us closer to more fluid communication", "after": "We also find that fine-tuning on similar statements offer limited gains, as PTLMs still fail to generalize to unseen inferences. Our new large-scale benchmark exposes a significant gap between PTLMs and human-level language understanding and offers a new challenge for PTLMs to demonstrate commonsense", "start_char_pos": 1042, "end_char_pos": 1215, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 200, 345, 492, 714, 821, 1041]}
{"doc_id": "2005.01795", "revision_depth": 2, "before_revision": "Following each patient visit, physicians must draft a detailed clinical summary called a SOAP note. Moreover, with electronic health records, these notes must be digitized. Despite the benefits of this documentation, their creation remains an onerous process, contributing to increasing physician burnout. In this paper, we present the first study to evaluate complete pipelines to train summarization models to generate these notes from conversations between physicians and patients. We benefit from a dataset that, along with transcripts and paired SOAP notes, consists of annotations marking noteworthy utterances that support each summary sentence. We decompose the problem into extractive and abstractive subtasks, exploring a spectrum of approaches according to how much they demand from each component. We observe that the performance improves constantly as the extractive subtask is made more complex - an observation that we also replicate on the well-known AMI meeting summarization dataset. Our best performing method first (i) extracts noteworthy utterances via multi-label classification, assigning each to summary section(s) ; (ii) clusters noteworthy utteranceson a per-section basis; and (iii) generates the summary sentences by conditioning on the corresponding cluster and the subsection of the SOAP sentence to be generated. Compared to an end-to-end approach that generates the full SOAP note from the full conversation, our approach improves by around 8 ROUGE-1 points .", "after_revision": "Following each patient visit, physicians draft long semi-structured clinical summaries called SOAP notes. While invaluable to clinicians and researchers, creating digital SOAP notes is burdensome, contributing to physician burnout. In this paper, we introduce the first complete pipelines to leverage deep summarization models to generate these notes based on transcripts of conversations between physicians and patients. After exploring a spectrum of methods across the extractive-abstractive spectrum, we propose Cluster2Sent, an algorithm that (i) extracts important utterances relevant to each summary section ; (ii) clusters together related utterances; and then (iii) generates one summary sentence per cluster. Cluster2Sent outperforms its purely abstractive counterpart by 8 ROUGE-1 points, and produces significantly more factual and coherent sentences as assessed by expert human evaluators. For reproducibility, we demonstrate similar benefits on the publicly available AMI dataset. Our results speak to the benefits of structuring summaries into sections and annotating supporting evidence when constructing summarization corpora .", "edit_actions": [{"type": "R", "before": "must draft a detailed clinical summary called a SOAP note. Moreover, with electronic health records, these notes must be digitized. Despite the benefits of this documentation, their creation remains an onerous process, contributing to increasing", "after": "draft long semi-structured clinical summaries called SOAP notes. While invaluable to clinicians and researchers, creating digital SOAP notes is burdensome, contributing to", "start_char_pos": 41, "end_char_pos": 286, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "present the first study to evaluate", "after": "introduce the first", "start_char_pos": 324, "end_char_pos": 359, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "train", "after": "leverage deep", "start_char_pos": 382, "end_char_pos": 387, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "from", "after": "based on transcripts of", "start_char_pos": 433, "end_char_pos": 437, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "We benefit from a dataset that, along with transcripts and paired SOAP notes, consists of annotations marking noteworthy utterances that support each summary sentence. We decompose the problem into extractive and abstractive subtasks,", "after": "After", "start_char_pos": 485, "end_char_pos": 719, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "approaches according to how much they demand from each component. We observe that the performance improves constantly as the extractive subtask is made more complex - an observation that we also replicate on the well-known AMI meeting summarization dataset. Our best performing method first", "after": "methods across the extractive-abstractive spectrum, we propose Cluster2Sent, an algorithm that", "start_char_pos": 744, "end_char_pos": 1034, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "noteworthy utterances via multi-label classification, assigning each to summary section(s)", "after": "important utterances relevant to each summary section", "start_char_pos": 1048, "end_char_pos": 1138, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "noteworthy utteranceson a per-section basis; and", "after": "together related utterances; and then", "start_char_pos": 1155, "end_char_pos": 1203, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the summary sentences by conditioning on the corresponding cluster and the subsection of the SOAP sentence to be generated. Compared to an end-to-end approach that generates the full SOAP note from the full conversation, our approach improves by around", "after": "one summary sentence per cluster. Cluster2Sent outperforms its purely abstractive counterpart by", "start_char_pos": 1220, "end_char_pos": 1472, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "points", "after": "points, and produces significantly more factual and coherent sentences as assessed by expert human evaluators. For reproducibility, we demonstrate similar benefits on the publicly available AMI dataset. Our results speak to the benefits of structuring summaries into sections and annotating supporting evidence when constructing summarization corpora", "start_char_pos": 1483, "end_char_pos": 1489, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 99, 172, 305, 484, 652, 809, 1001, 1140, 1199, 1343], "domain": "arxiv"}
{"doc_id": "2005.03954", "revision_depth": 1, "before_revision": "We focus on the study of conversational recommendation in the context of multi-type dialogs, where the bots can proactively and naturally lead a conversation from a non-recommendation dialog (e.g., QA) to a recommendation dialog, taking into account user's interests and feedback. To facilitate the study of this task, we create a human-to-human Chinese dialog dataset DuRecDial (about 10k dialogs, 156k utterances), where there are multiple sequential dialogs for a pair of a recommendation seeker (user) and a recommender (bot). In each dialog, the recommender proactively leads a multi-type dialog to approach recommendation targets and then makes multiple recommendations with rich interaction behavior. This dataset allows us to systematically investigate different parts of the overall problem, e.g., how to naturally lead a dialog, how to interact with users for recommendation. Finally we establish baseline results on DuRecDial for future studies. Dataset and codes are publicly available at URL", "after_revision": "We focus on the study of conversational recommendation in the context of multi-type dialogs, where the bots can proactively and naturally lead a conversation from a non-recommendation dialog (e.g., QA) to a recommendation dialog, taking into account user's interests and feedback. To facilitate the study of this task, we create a human-to-human Chinese dialog dataset DuRecDial (about 10k dialogs, 156k utterances), which contains multiple sequential dialogs for every pair of a recommendation seeker (user) and a recommender (bot). In each dialog, the recommender proactively leads a multi-type dialog to approach recommendation targets and then makes multiple recommendations with rich interaction behavior. This dataset allows us to systematically investigate different parts of the overall problem, e.g., how to naturally lead a dialog, how to interact with users for recommendation. Finally we establish baseline results on DuRecDial for future studies. Dataset and codes are publicly available at URL", "edit_actions": [{"type": "R", "before": "where there are", "after": "which contains", "start_char_pos": 417, "end_char_pos": 432, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "a", "after": "every", "start_char_pos": 465, "end_char_pos": 466, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 280, 530, 707, 885, 956], "domain": "arxiv"}
{"doc_id": "2005.03975", "revision_depth": 1, "before_revision": "To address the need for refined information in COVID-19 pandemic, we propose a deep learning-based system that uses state-of-the-art natural language processing (NLP) question answering (QA) techniques combined with summarization  for mining the available scientific literature. Our system leverages the Information Retrieval (IR) system and QA models to extract relevant snippets from the existing literature given a query. Fluent summaries are also provided to help understand the content in a more efficient way. In this paper, we describe our CAiRE-COVID system architecture and methodology for building the system. To bootstrap the further study, the code for our system is available at URL", "after_revision": "The outbreak of COVID-19 raises attention from the researchers from various communities. While many scientific articles have been published, a system that can provide reliable information to COVID-19 related questions from the latest academic resources is crucial, especially for the medical community in the current time-critical race to treat patients and to find a cure for the virus. To address the requests, we propose our CAiRE-COVID, a neural-based system that uses open-domain question answering (QA) techniques combined with summarization techniques for mining the available scientific literature. It leverages the Information Retrieval (IR) system and QA models to extract relevant snippets from existing literature given a query. Fluent summaries are also provided to help understand the content in a more efficient way. Our system has been awarded as winner for one of the tasks in CORD-19 Kaggle Challenge. We also launched our CAiRE-COVID website for broader use. The code for our system is also open-sourced to bootstrap further study.", "edit_actions": [{"type": "R", "before": "To address the need for refined information in", "after": "The outbreak of", "start_char_pos": 0, "end_char_pos": 46, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "pandemic, we propose a deep learning-based", "after": "raises attention from the researchers from various communities. While many scientific articles have been published, a system that can provide reliable information to COVID-19 related questions from the latest academic resources is crucial, especially for the medical community in the current time-critical race to treat patients and to find a cure for the virus. To address the requests, we propose our CAiRE-COVID, a neural-based", "start_char_pos": 56, "end_char_pos": 98, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "state-of-the-art natural language processing (NLP)", "after": "open-domain", "start_char_pos": 116, "end_char_pos": 166, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "techniques", "start_char_pos": 230, "end_char_pos": 230, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "Our system", "after": "It", "start_char_pos": 279, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 386, "end_char_pos": 389, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "In this paper, we describe our CAiRE-COVID system architecture and methodology for building the system. To bootstrap the further study, the", "after": "Our system has been awarded as winner for one of the tasks in CORD-19 Kaggle Challenge. We also launched our CAiRE-COVID website for broader use. The", "start_char_pos": 516, "end_char_pos": 655, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "available at URL", "after": "also open-sourced to bootstrap further study.", "start_char_pos": 679, "end_char_pos": 695, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 278, 424, 515, 619], "domain": "arxiv"}
{"doc_id": "2005.03975", "revision_depth": 2, "before_revision": "The outbreak of COVID-19 raises attention from the researchers from various communities. While many scientific articles have been published , a system that can provide reliable information to COVID-19 related questions from the latest academic resources is crucial, especially for the medical community in the current time-critical race to treat patients and to find a cure for the virus. To address the requests, we propose our CAiRE-COVID, a neural-based system that uses open-domain question answering (QA) techniques combined with summarization techniques for mining the available scientific literature. It leverages the Information Retrieval (IR) system and QA models to extract relevant snippets from existing literature given a query. Fluent summaries are also provided to help understand the content in a more efficient way. Our system has been awarded as winner for one of the tasks in CORD-19 Kaggle Challenge. We also launched our CAiRE-COVID website for broader use . The code for our system is also open-sourced to bootstrap further study  .", "after_revision": "We present CAiRE-COVID, a real-time question answering (QA) and multi-document summarization system, which won one of the 10 tasks in the Kaggle COVID-19 Open Research Dataset Challenge, judged by medical experts. Our system aims to tackle the recent challenge of mining the numerous scientific articles being published on COVID-19 by answering high priority questions from the community and summarizing salient question-related information. It combines information extraction with state-of-the-art QA and query-focused multi-document summarization techniques, selecting and highlighting evidence snippets from existing literature given a query. We also propose query-focused abstractive and extractive multi-document summarization methods, to provide more relevant information related to the question. We further conduct quantitative experiments that show consistent improvements on various metrics for each module. We have launched our website CAiRE-COVID for broader use by the medical community, and have open-sourced the code for our system , to bootstrap further study by other researches .", "edit_actions": [{"type": "R", "before": "The outbreak of", "after": "We present CAiRE-COVID, a real-time question answering (QA) and multi-document summarization system, which won one of the 10 tasks in the Kaggle", "start_char_pos": 0, "end_char_pos": 15, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "raises attention from the researchers from various communities. While many scientific articles have been published , a system that can provide reliable information to", "after": "Open Research Dataset Challenge, judged by medical experts. Our system aims to tackle the recent challenge of mining the numerous scientific articles being published on", "start_char_pos": 25, "end_char_pos": 191, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "related", "after": "by answering high priority", "start_char_pos": 201, "end_char_pos": 208, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "latest academic resources is crucial, especially for the medical community in the current time-critical race to treat patients and to find a cure for the virus. To address the requests, we propose our CAiRE-COVID, a neural-based system that uses open-domain question answering (QA) techniques combined with summarization techniques for mining the available scientific literature. It leverages the Information Retrieval (IR) system and QA models to extract relevant", "after": "community and summarizing salient question-related information. It combines information extraction with state-of-the-art QA and query-focused multi-document summarization techniques, selecting and highlighting evidence", "start_char_pos": 228, "end_char_pos": 692, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "Fluent summaries are also provided to help understand the content in a more efficient way. Our system has been awarded as winner for one of the tasks in CORD-19 Kaggle Challenge. We also launched our CAiRE-COVID website", "after": "We also propose query-focused abstractive and extractive multi-document summarization methods, to provide more relevant information related to the question. We further conduct quantitative experiments that show consistent improvements on various metrics for each module. We have launched our website CAiRE-COVID", "start_char_pos": 742, "end_char_pos": 961, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". The", "after": "by the medical community, and have open-sourced the", "start_char_pos": 978, "end_char_pos": 983, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "is also open-sourced", "after": ",", "start_char_pos": 1004, "end_char_pos": 1024, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "by other researches", "start_char_pos": 1052, "end_char_pos": 1052, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 88, 388, 607, 741, 832, 920, 979], "domain": "arxiv"}
{"doc_id": "2005.05298", "revision_depth": 1, "before_revision": "This paper presents a new method SOLOIST, which uses transfer learning to efficiently build task-oriented dialog systems at scale. We parameterize a dialog system using a Transformer-based auto-regressive language model, which subsumes different dialog mod-ules (e.g., state tracker, dialog policy, responsegenerator ) into a single neural model. We pre-train, on large heterogeneous dialog corpora, a large-scale Transformer model which can generate dialog responses grounded in user goals and real-world knowledge for task completion. The pre-trained model can be efficiently adapted to accomplish a new dialog task with a handful of task-specific dialogs via machine teaching. Our experiments demonstrate that (i) SOLOIST creates new state-of-the-art results on two well-known benchmarks, CamRest and MultiWOZ, (ii) in the few-shot learning setting, the dialog systems developed by SOLOIST significantly outperform those  by existing methods, and (iii) the use of machine teaching substantially reduces the labeling cost. We will release our code and pre-trained models for reproducible research.", "after_revision": "This paper presents a new method SOLOIST, which uses transfer learning to efficiently build task-oriented dialog systems at scale. We parameterize a dialog system using a Transformer-based auto-regressive language model, which subsumes different dialog modules (e.g., state tracker, dialog policy, response generator ) into a single neural model. We pre-train, on large heterogeneous dialog corpora, a large-scale Transformer model which can generate dialog responses grounded in user goals and real-world knowledge for task completion. The pre-trained model can be efficiently adapted to accomplish a new dialog task with a handful of task-specific dialogs via machine teaching. Our experiments demonstrate that (i) SOLOIST creates new state-of-the-art results on two well-known benchmarks, CamRest and MultiWOZ, (ii) in the few-shot learning setting, the dialog systems developed by SOLOIST significantly outperform those developed by existing methods, and (iii) the use of machine teaching substantially reduces the labeling cost. We will release our code and pre-trained models for reproducible research.", "edit_actions": [{"type": "R", "before": "mod-ules", "after": "modules", "start_char_pos": 253, "end_char_pos": 261, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "responsegenerator", "after": "response generator", "start_char_pos": 299, "end_char_pos": 316, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "developed", "start_char_pos": 924, "end_char_pos": 924, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 130, 346, 536, 679, 1024], "domain": "arxiv"}
{"doc_id": "2005.06012", "revision_depth": 1, "before_revision": "We describe Mega-COV, a billion-scale dataset from Twitter for studying COVID-19. The dataset is diverse (covers 234 countries), longitudinal (goes as back as 2007), multilingual (comes in 65 languages), and has a significant number of location-tagged tweets ( ~32M tweets). We release tweet IDs from the dataset , hoping it will be useful for studying various phenomena related to the ongoing pandemic and accelerating viable solutions to associated problems.", "after_revision": "We describe Mega-COV, a billion-scale dataset from Twitter for studying COVID-19. The dataset is diverse (covers 268 countries), longitudinal (goes as back as 2007), multilingual (comes in 100+ languages), and has a significant number of location-tagged tweets ( 169M tweets). We release tweet IDs from the dataset . We also develop and release a powerful model (acc=94\\%)", "edit_actions": [{"type": "R", "before": "234", "after": "268", "start_char_pos": 113, "end_char_pos": 116, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "65", "after": "100+", "start_char_pos": 189, "end_char_pos": 191, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "~32M", "after": "169M", "start_char_pos": 261, "end_char_pos": 265, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", hoping it will be useful for studying various phenomena related to the ongoing pandemic and accelerating viable solutions to associated problems.", "after": ". We also develop and release a powerful model (acc=94\\%)", "start_char_pos": 313, "end_char_pos": 460, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}], "sents_char_pos": [0, 81, 274], "domain": "arxiv"}
{"doc_id": "2005.06377", "revision_depth": 1, "before_revision": "ROUGE is the de facto criterion for summarization research. However, its two major drawbackslimit the research and application of automated summarization systems . First, ROUGE favors lexical similarity instead of semantic similarity , making it especially unfit for abstractive summarization . Second, ROUGE cannot function without a reference summary, which is expensive or impossible to obtain in many cases . Therefore , we introduce a new end-to-end metric system for summary quality assessment by leveraging the semantic similarities of words and/or sentences in deep learning. Models trained in our framework can evaluate a summary directly against the input document , without the need of a reference summary . The proposed approach exhibits very promising results on gold-standard datasets and suggests its great potential to future summarization research. The scores from our models have correlation coefficients up to 0.54 with human evaluations on machine generated summaries in TAC2010. Its performance is also very close to ROUGE metrics' .", "after_revision": "Canonical automatic summary evaluation metrics, such as ROUGE, suffer from two drawbacks . First, semantic similarity and linguistic quality are not captured well . Second, a reference summary, which is expensive or impossible to obtain in many cases , is needed. Existing efforts to address the two drawbacks are done separately and have limitations. To holistically address them , we introduce an end-to-end approach for summary quality assessment by leveraging sentence or document embedding and introducing two negative sampling approaches to create training data for this supervised approach . The proposed approach exhibits promising results on several summarization datasets of various domains including news, legislative bills, scientific papers, and patents. When rating machine-generated summaries in TAC2010, our approach outperforms ROUGE in terms of linguistic quality, and achieves a correlation coefficient of up to 0.5702 with human evaluations in terms of modified pyramid scores. We hope our approach can facilitate summarization research or applications when reference summaries are infeasible or costly to obtain, or when linguistic quality is a focus .", "edit_actions": [{"type": "R", "before": "ROUGE is the de facto criterion for summarization research. However, its two major drawbackslimit the research and application of automated summarization systems", "after": "Canonical automatic summary evaluation metrics, such as ROUGE, suffer from two drawbacks", "start_char_pos": 0, "end_char_pos": 161, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "ROUGE favors lexical similarity instead of semantic similarity , making it especially unfit for abstractive summarization", "after": "semantic similarity and linguistic quality are not captured well", "start_char_pos": 171, "end_char_pos": 292, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "ROUGE cannot function without", "after": null, "start_char_pos": 303, "end_char_pos": 332, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": ". Therefore", "after": ", is needed. Existing efforts to address the two drawbacks are done separately and have limitations. To holistically address them", "start_char_pos": 411, "end_char_pos": 422, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "a new", "after": "an", "start_char_pos": 438, "end_char_pos": 443, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "metric system", "after": "approach", "start_char_pos": 455, "end_char_pos": 468, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the semantic similarities of words and/or sentences in deep learning. Models trained in our framework can evaluate a summary directly against the input document , without the need of a reference summary", "after": "sentence or document embedding and introducing two negative sampling approaches to create training data for this supervised approach", "start_char_pos": 514, "end_char_pos": 716, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "very", "after": null, "start_char_pos": 750, "end_char_pos": 754, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "gold-standard datasets and suggests its great potential to future summarization research. The scores from our models have correlation coefficients up to 0.54", "after": "several summarization datasets of various domains including news, legislative bills, scientific papers, and patents. When rating machine-generated summaries in TAC2010, our approach outperforms ROUGE in terms of linguistic quality, and achieves a correlation coefficient of up to 0.5702", "start_char_pos": 776, "end_char_pos": 933, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "on machine generated summaries in TAC2010. Its performance is also very close to ROUGE metrics'", "after": "in terms of modified pyramid scores. We hope our approach can facilitate summarization research or applications when reference summaries are infeasible or costly to obtain, or when linguistic quality is a focus", "start_char_pos": 957, "end_char_pos": 1052, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 59, 163, 294, 412, 583, 718, 865, 999], "domain": "arxiv"}
{"doc_id": "2005.07202", "revision_depth": 1, "before_revision": "Bidirectional Encoder Representations from Transformers (BERT) models for biomedical specialties such as BioBERT and clinicalBERT  have significantly improved in biomedical text-mining tasks and enabled us to extract valuable information from biomedical literature . However, we benefitted only in English because of the significant scarcity of high-quality medical documents, such as PubMed, in each language. Therefore, we propose a method that realizes a high-performance BERT model by using a small corpus. We introduce the method to train a BERT model on a small medical corpus both in English and Japanese, respectively, and then we evaluate each of them in terms of the biomedical language understanding evaluation (BLUE) benchmark and the medical-document-classification task in Japanese, respectively. After confirming their satisfactory performances, we apply our method to develop a model that outperforms the pre-existing models. Bidirectional Encoder Representations from Transformers for Biomedical Text Mining by Osaka University (ouBioBERT) achieves the best scores on 7 of the 10 datasets in terms of the BLUE benchmark. The total score is 1.0 points above that of BioBERT  .", "after_revision": "Bidirectional Encoder Representations from Transformers (BERT) models for medical specialties, such as BioBERT and clinicalBERT , have significantly improved in performing biomedical text mining tasks and have enabled extracting valuable information from biomedical literature ; however, only English speakers benefit due to the significant scarcity of high-quality medical documents, such as PubMed, in each language. Therefore, we propose a method to train a high-performance BERT model using a small corpus. We introduce the method to train a BERT model on a small medical corpus both in English and in Japanese, and we present the evaluation of each model in terms of the biomedical language understanding evaluation (BLUE) benchmark and the medical document classification task in Japanese, respectively. After confirming their satisfactory performances, we applied our method to develop a model comparable to the publicly available models. OuBioBERT, short for Bidirectional Encoder Representations from Transformers for Biomedical Text Mining by Osaka University , achieved the best score in terms of the BLUE benchmark. The total score is 1.1 points above that of BioBERT and 0.3 points above that of the ablated model trained without our proposed method. This proposed technique is an effective approach to develop localized medical BERT models and to enhance domain-specific models in the biomedical domain .", "edit_actions": [{"type": "R", "before": "biomedical specialties", "after": "medical specialties,", "start_char_pos": 74, "end_char_pos": 96, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 130, "end_char_pos": 130, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "biomedical text-mining tasks and enabled us to extract", "after": "performing biomedical text mining tasks and have enabled extracting", "start_char_pos": 162, "end_char_pos": 216, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": ". However, we benefitted only in English because of", "after": "; however, only English speakers benefit due to", "start_char_pos": 265, "end_char_pos": 316, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "that realizes", "after": "to train", "start_char_pos": 442, "end_char_pos": 455, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": "by", "after": null, "start_char_pos": 486, "end_char_pos": 488, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Japanese, respectively, and then we evaluate each of them", "after": "in Japanese, and we present the evaluation of each model", "start_char_pos": 603, "end_char_pos": 660, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "medical-document-classification", "after": "medical document classification", "start_char_pos": 747, "end_char_pos": 778, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "apply", "after": "applied", "start_char_pos": 864, "end_char_pos": 869, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "that outperforms the pre-existing models.", "after": "comparable to the publicly available models. OuBioBERT, short for", "start_char_pos": 900, "end_char_pos": 941, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "(ouBioBERT) achieves the best scores on 7 of the 10 datasets", "after": ", achieved the best score", "start_char_pos": 1045, "end_char_pos": 1105, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "1.0", "after": "1.1", "start_char_pos": 1157, "end_char_pos": 1160, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "and 0.3 points above that of the ablated model trained without our proposed method. This proposed technique is an effective approach to develop localized medical BERT models and to enhance domain-specific models in the biomedical domain", "start_char_pos": 1190, "end_char_pos": 1190, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 266, 410, 510, 810, 941, 1137], "domain": "arxiv"}
{"doc_id": "2005.07202", "revision_depth": 2, "before_revision": "Bidirectional Encoder Representations from Transformers (BERT)models for medical specialties , such as BioBERT and clinicalBERT, have significantly improved in performing biomedical text mining tasks and have enabled extracting valuable information from biomedical literature ; however, only English speakers benefit due to the significant scarcity of high-quality medical documents, such as PubMed, in each language. Therefore, we propose a method to train a high-performance BERT model using a small corpus . We introduce the method to train a BERT model on a small medical corpus both in English and in Japanese , and we present the evaluation of each model in terms of the biomedical language understanding evaluation (BLUE) benchmark and the medical document classification task in Japanese, respectively. After confirming their satisfactory performances, we applied our method to develop a modelcomparable to the publicly available models. OuBioBERT, short for Bidirectional Encoder Representations from Transformers for Biomedical Text Mining by Osaka University, achieved the best score in terms of the BLUE benchmark . The total score is 1.1 points above that of BioBERT and 0.3 points above that of the ablated model trained without our proposed method. This proposed technique is an effective approach to develop localized medical BERT models and to enhance domain-specific models in the biomedical domain .", "after_revision": "Pre-training large-scale neural language models on raw texts has made a significant contribution to improving transfer learning in natural language processing (NLP). With the introduction of transformer-based language models , such as bidirectional encoder representations from transformers (BERT), the performance of information extraction from a free text by NLP has significantly improved for both the general domain and medical domain ; however, it is difficult to train specific BERT models that perform well for domains in which there are few publicly available databases of high quality and large size. We hypothesized that this problem can be addressed by up-sampling a domain-specific corpus and using it for pre-training with a larger corpus in a balanced manner. Our proposed method consists of a single intervention with one option: simultaneous pre-training after up-sampling and amplified vocabulary. We conducted three experiments and evaluated the resulting products. We confirmed that our Japanese medical BERT outperformed conventional baselines and the other BERT models in terms of the medical document classification task and that our English BERT pre-trained using both the general and medical-domain corpora performed sufficiently well for practical use in terms of the biomedical language understanding evaluation (BLUE) benchmark . Moreover, our enhanced biomedical BERT model, in which clinical notes were not used during pre-training, showed that both the clinical and biomedical scores of the BLUE benchmark were 0.3 points above that of the ablation model trained without our proposed method. Well-balanced pre-training by up-sampling instances derived from a corpus appropriate for the target task allows us to construct a high-performance BERT model .", "edit_actions": [{"type": "R", "before": "Bidirectional Encoder Representations from Transformers (BERT)models for medical specialties", "after": "Pre-training large-scale neural language models on raw texts has made a significant contribution to improving transfer learning in natural language processing (NLP). With the introduction of transformer-based language models", "start_char_pos": 0, "end_char_pos": 92, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "BioBERT and clinicalBERT, have significantly improved in performing biomedical text mining tasks and have enabled extracting valuable information from biomedical literature", "after": "bidirectional encoder representations from transformers (BERT), the performance of information extraction from a free text by NLP has significantly improved for both the general domain and medical domain", "start_char_pos": 103, "end_char_pos": 275, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "only English speakers benefit due to the significant scarcity of high-quality medical documents, such as PubMed, in each language. Therefore, we propose a method to train a high-performance BERT model using a small corpus . We introduce the method to train a BERT model on a small medical corpus both in English and in Japanese , and we present the evaluation of each model", "after": "it is difficult to train specific BERT models that perform well for domains in which there are few publicly available databases of high quality and large size. We hypothesized that this problem can be addressed by up-sampling a domain-specific corpus and using it for pre-training with a larger corpus in a balanced manner. Our proposed method consists of a single intervention with one option: simultaneous pre-training after up-sampling and amplified vocabulary. We conducted three experiments and evaluated the resulting products. We confirmed that our Japanese medical BERT outperformed conventional baselines and the other BERT models in terms of the medical document classification task and that our English BERT pre-trained using both the general and medical-domain corpora performed sufficiently well for practical use", "start_char_pos": 287, "end_char_pos": 660, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "and the medical document classification task in Japanese, respectively. After confirming their satisfactory performances, we applied our method to develop a modelcomparable to the publicly available models. OuBioBERT, short for Bidirectional Encoder Representations from Transformers for Biomedical Text Mining by Osaka University, achieved the best score in terms", "after": ". Moreover, our enhanced biomedical BERT model, in which clinical notes were not used during pre-training, showed that both the clinical and biomedical scores", "start_char_pos": 739, "end_char_pos": 1103, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": ". The total score is 1.1 points above that of BioBERT and", "after": "were", "start_char_pos": 1126, "end_char_pos": 1183, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "ablated", "after": "ablation", "start_char_pos": 1213, "end_char_pos": 1220, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "This proposed technique is an effective approach to develop localized medical BERT models and to enhance domain-specific models in the biomedical domain", "after": "Well-balanced pre-training by up-sampling instances derived from a corpus appropriate for the target task allows us to construct a high-performance BERT model", "start_char_pos": 1264, "end_char_pos": 1416, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 277, 417, 510, 810, 945, 1127, 1263], "domain": "arxiv"}
{"doc_id": "2005.07456", "revision_depth": 1, "before_revision": "Word embeddings represent words in a numeric space in such a way that semantic relations between words are encoded as distances and directions in the vector space. Cross-lingual word embeddings map words from one language to the vector space of another language, or words from multiple languages to the same vector space where similar words are aligned.  Cross-lingual embeddings can be used to transfer machine learning models between languages and thereby compensate for insufficient data in less-resourced languages. We use cross-lingual word embeddings to transfer machine learning prediction models for Twitter sentiment between 13 languages. We focus on two transfer mechanisms using the joint numerical space for many languages as implemented in the LASER library : the transfer of trained models, and expansion of training sets with instances from other languages . Our experiments show that the transfer of models between similar languages is sensible, while dataset expansion did not increase the predictive performance .", "after_revision": "Word embeddings represent words in a numeric space so that semantic relations between words are represented as distances and directions in the vector space. Cross-lingual word embeddings transform vector spaces of different languages so that similar words are aligned. This is done by constructing a mapping between vector spaces of two languages or learning a joint vector space for multiple languages. Cross-lingual embeddings can be used to transfer machine learning models between languages , thereby compensating for insufficient data in less-resourced languages. We use cross-lingual word embeddings to transfer machine learning prediction models for Twitter sentiment between 13 languages. We focus on two transfer mechanisms that recently show superior transfer performance. The first mechanism uses the trained models whose input is the joint numerical space for many languages as implemented in the LASER library . The second mechanism uses large pretrained multilingual BERT language models . Our experiments show that the transfer of models between similar languages is sensible, even with no target language data. The performance of cross-lingual models obtained with the multilingual BERT and LASER library is comparable, and the differences are language-dependent. The transfer with CroSloEngual BERT, pretrained on only three languages, is superior on these and some closely related languages .", "edit_actions": [{"type": "R", "before": "in such a way", "after": "so", "start_char_pos": 51, "end_char_pos": 64, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "encoded", "after": "represented", "start_char_pos": 107, "end_char_pos": 114, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "map words from one language to the vector space of another language, or words from multiple languages to the same vector space where", "after": "transform vector spaces of different languages so that", "start_char_pos": 194, "end_char_pos": 326, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "This is done by constructing a mapping between vector spaces of two languages or learning a joint vector space for multiple languages.", "start_char_pos": 354, "end_char_pos": 354, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "and thereby compensate", "after": ", thereby compensating", "start_char_pos": 446, "end_char_pos": 468, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "using the", "after": "that recently show superior transfer performance. The first mechanism uses the trained models whose input is the", "start_char_pos": 684, "end_char_pos": 693, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": ": the transfer of trained models, and expansion of training sets with instances from other languages", "after": ". The second mechanism uses large pretrained multilingual BERT language models", "start_char_pos": 771, "end_char_pos": 871, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "while dataset expansion did not increase the predictive performance", "after": "even with no target language data. The performance of cross-lingual models obtained with the multilingual BERT and LASER library is comparable, and the differences are language-dependent. The transfer with CroSloEngual BERT, pretrained on only three languages, is superior on these and some closely related languages", "start_char_pos": 962, "end_char_pos": 1029, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}], "sents_char_pos": [0, 163, 353, 519, 647, 873], "domain": "arxiv"}
{"doc_id": "2005.12086", "revision_depth": 1, "before_revision": "Text style transfer is the task that generates a sentence by preserving the content of the input sentence and transferring the style. Most existing studies are progressing on non-parallel datasets because parallel datasets are limited and hard to construct. In this work, we introduce a method that follows two stages in non-parallel datasets. The first stage is to delete attribute markers of a sentence directly through the classifier. The second stage is to generate the transferred sentence by combining the content tokens and the target style. We evaluate systems on two benchmark datasets . Transferred sentences are evaluated in terms of context, style, fluency, and semantic. These evaluation metricsare used to determine a stable system. Only robust systems in all evaluation metrics are suitable for use in real applications. Many previous systems are difficult to use in certain situations because they are unstable in some evaluation metrics. However, our system is stable in all  evaluation metrics and has results comparable to other models. ", "after_revision": "Text style transfer is the task that generates a sentence by preserving the content of the input sentence and transferring the style. Most existing studies are progressing on non-parallel datasets because parallel datasets are limited and hard to construct. In this work, we introduce a method that follows two stages in non-parallel datasets. The first stage is to delete attribute markers of a sentence directly through a classifier. The second stage is to generate a transferred sentence by combining the content tokens and the target style. We experiment on two benchmark datasets and evaluate context, style, fluency, and semantic. It is difficult to select the best system using only these automatic metrics, but it is possible to select stable systems. We consider only robust systems in all automatic evaluation metrics to be the minimum conditions that can be used in real applications. Many previous systems are difficult to use in certain situations because performance is significantly lower in several evaluation metrics. However, our system is stable in all automatic evaluation metrics and has results comparable to other models. Also, we compare the performance results of our system and the unstable system through human evaluation. Our code and data are available at the URL", "edit_actions": [{"type": "R", "before": "the", "after": "a", "start_char_pos": 422, "end_char_pos": 425, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "the", "after": "a", "start_char_pos": 470, "end_char_pos": 473, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "evaluate systems", "after": "experiment", "start_char_pos": 552, "end_char_pos": 568, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ". Transferred sentences are evaluated in terms of", "after": "and evaluate", "start_char_pos": 595, "end_char_pos": 644, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "These evaluation metricsare used to determine a stable system. Only", "after": "It is difficult to select the best system using only these automatic metrics, but it is possible to select stable systems. We consider only", "start_char_pos": 684, "end_char_pos": 751, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "evaluation metrics are suitable for use", "after": "automatic evaluation metrics to be the minimum conditions that can be used", "start_char_pos": 774, "end_char_pos": 813, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "style"]}, {"type": "R", "before": "they are unstable in some", "after": "performance is significantly lower in several", "start_char_pos": 909, "end_char_pos": 934, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "automatic", "start_char_pos": 992, "end_char_pos": 992, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "A", "before": null, "after": "Also, we compare the performance results of our system and the unstable system through human evaluation. Our code and data are available at the URL", "start_char_pos": 1056, "end_char_pos": 1056, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 133, 257, 343, 437, 548, 596, 683, 746, 835, 954], "domain": "arxiv"}
{"doc_id": "2005.12766", "revision_depth": 1, "before_revision": "Pretrained language models such as BERT, GPT have shown great effectiveness in language understanding. The auxiliary predictive tasks in existing pretraining approaches are mostly defined on tokens, thus may not be able to capture sentence-level semantics very well. To address this issue, we propose CERT: Contrastive self-supervised Encoder Representations from Transformers, which pretrains language representation models using contrastive self-supervised learning at the sentence level. CERT creates augmentations of original sentences using back-translation. Then it finetunes a pretrained language encoder (e.g., BERT) by predicting whether two augmented sentences originate from the same sentence. CERT is simple to use and can be flexibly plugged into any pretraining-finetuning NLP pipeline. We evaluate CERT on three language understanding tasks : CoLA, RTE, and QNLI. CERT outperforms BERT significantly.", "after_revision": "Pretrained language models such as BERT, GPT have shown great effectiveness in language understanding. The auxiliary predictive tasks in existing pretraining approaches are mostly defined on tokens, thus may not be able to capture sentence-level semantics very well. To address this issue, we propose CERT: Contrastive self-supervised Encoder Representations from Transformers, which pretrains language representation models using contrastive self-supervised learning at the sentence level. CERT creates augmentations of original sentences using back-translation. Then it finetunes a pretrained language encoder (e.g., BERT) by predicting whether two augmented sentences originate from the same sentence. CERT is simple to use and can be flexibly plugged into any pretraining-finetuning NLP pipeline. We evaluate CERT on 11 natural language understanding tasks in the GLUE benchmark where CERT outperforms BERT on 7 tasks, achieves the same performance as BERT on 2 tasks, and performs worse than BERT on 2 tasks. On the averaged score of the 11 tasks, CERT outperforms BERT . The data and code are available at URL", "edit_actions": [{"type": "R", "before": "three", "after": "11 natural", "start_char_pos": 821, "end_char_pos": 826, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ": CoLA, RTE, and QNLI.", "after": "in the GLUE benchmark where CERT outperforms BERT on 7 tasks, achieves the same performance as BERT on 2 tasks, and performs worse than BERT on 2 tasks. On the averaged score of the 11 tasks,", "start_char_pos": 856, "end_char_pos": 878, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "significantly.", "after": ". The data and code are available at URL", "start_char_pos": 901, "end_char_pos": 915, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 102, 266, 490, 563, 704, 800], "domain": "arxiv"}
{"doc_id": "2005.12889", "revision_depth": 1, "before_revision": "Few resources represent implicit roles for natural language understanding , and existing studies in NLP only make coarse distinctions between categories of arguments omitted from linguistic form. In this paper , we design a typology for fine-grained implicit argument annotation on top of Universal Conceptual Cognitive Annotation's foundational layer (Abend and Rappoport, 2013). Our design aligns with O'Gorman (2019)'s implicit role interpretation in a linguistic and computational model. The proposed implicit argument categorisation set consists of six types: Deictic, Generic, Genre-based, Type-identifiable, Non-specific, and Iterated-set. We corroborate the theory by reviewing and refining part of the UCCA EWT corpus and providing a new dataset alongside comparative analysis with other schemes . It is anticipated that our study will inspire tailored design of implicit role annotation in other meaning representation frameworks, and stimulate research in relevant fields, such as coreference resolution and question answering .", "after_revision": "Predicate-argument structure analysis is a central component in meaning representations of text. The fact that some arguments are not explicitly mentioned in a sentence gives rise to ambiguity in language understanding, and renders it difficult for machines to interpret text correctly. However, only few resources represent implicit roles for NLU , and existing studies in NLP only make coarse distinctions between categories of arguments omitted from linguistic form. This paper proposes a typology for fine-grained implicit argument annotation on top of Universal Conceptual Cognitive Annotation's foundational layer . The proposed implicit argument categorisation is driven by theories of implicit role interpretation and consists of six types: Deictic, Generic, Genre-based, Type-identifiable, Non-specific, and Iterated-set. We exemplify our design by revisiting part of the UCCA EWT corpus , providing a new dataset annotated with the refinement layer, and making a comparative analysis with other schemes .", "edit_actions": [{"type": "R", "before": "Few", "after": "Predicate-argument structure analysis is a central component in meaning representations of text. The fact that some arguments are not explicitly mentioned in a sentence gives rise to ambiguity in language understanding, and renders it difficult for machines to interpret text correctly. However, only few", "start_char_pos": 0, "end_char_pos": 3, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "natural language understanding", "after": "NLU", "start_char_pos": 43, "end_char_pos": 73, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": "In this paper , we design", "after": "This paper proposes", "start_char_pos": 196, "end_char_pos": 221, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "(Abend and Rappoport, 2013). Our design aligns with O'Gorman (2019)'s implicit role interpretation in a linguistic and computational model.", "after": ".", "start_char_pos": 352, "end_char_pos": 491, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "set", "after": "is driven by theories of implicit role interpretation and", "start_char_pos": 538, "end_char_pos": 541, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "corroborate the theory by reviewing and refining", "after": "exemplify our design by revisiting", "start_char_pos": 650, "end_char_pos": 698, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "and", "after": ",", "start_char_pos": 727, "end_char_pos": 730, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "alongside", "after": "annotated with the refinement layer, and making a", "start_char_pos": 755, "end_char_pos": 764, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ". It is anticipated that our study will inspire tailored design of implicit role annotation in other meaning representation frameworks, and stimulate research in relevant fields, such as coreference resolution and question answering", "after": null, "start_char_pos": 805, "end_char_pos": 1037, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 195, 380, 491, 646, 806], "domain": "arxiv"}
{"doc_id": "2005.13837", "revision_depth": 1, "before_revision": "One of the most crucial challenges in questionanswering (QA) is the scarcity of labeled data, since it is costly to obtain question-answer(QA) pairs for a target text domain with human annotation. An alternative approach totackle the problem is to use automatically generated QA pairs from either the problem context or from large amount of unstructured texts(e.g. Wikipedia). In this work, we propose a hierarchical conditional variational autoencoder(HCVAE) for generating QA pairs given unstructured texts as contexts, while maximizingthe mutual information between generated QApairs to ensure their consistency. We validateourInformation MaximizingHierarchicalConditionalVariationalAutoEncoder (Info-HCVAE) on several benchmark datasets byevaluating the performance of the QA model(BERT-base) using only the generated QApairs (QA-based evaluation) or by using boththe generated and human-labeled pairs (semi-supervised learning) for training, against state-of-the-art baseline models. The results showthat our model obtains impressive performance gains over all baselines on both tasks, using only a fraction of data for training ", "after_revision": "One of the most crucial challenges in question answering (QA) is the scarcity of labeled data, since it is costly to obtain question-answer(QA) pairs for a target text domain with human annotation. An alternative approach to tackle the problem is to use automatically generated QA pairs from either the problem context or from large amount of unstructured texts(e.g. Wikipedia). In this work, we propose a hierarchical conditional variational autoencoder(HCVAE) for generating QA pairs given unstructured texts as contexts, while maximizing the mutual information between generated QA pairs to ensure their consistency. We validate our Information Maximizing Hierarchical Conditional Variational AutoEncoder (Info-HCVAE) on several benchmark datasets by evaluating the performance of the QA model(BERT-base) using only the generated QA pairs (QA-based evaluation) or by using both the generated and human-labeled pairs (semi-supervised learning) for training, against state-of-the-art baseline models. The results show that our model obtains impressive performance gains over all baselines on both tasks, using only a fraction of data for training .", "edit_actions": [{"type": "R", "before": "questionanswering", "after": "question answering", "start_char_pos": 38, "end_char_pos": 55, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "totackle", "after": "to tackle", "start_char_pos": 221, "end_char_pos": 229, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "maximizingthe", "after": "maximizing the", "start_char_pos": 528, "end_char_pos": 541, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "QApairs", "after": "QA pairs", "start_char_pos": 579, "end_char_pos": 586, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "validateourInformation MaximizingHierarchicalConditionalVariationalAutoEncoder", "after": "validate our Information Maximizing Hierarchical Conditional Variational AutoEncoder", "start_char_pos": 619, "end_char_pos": 697, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "byevaluating", "after": "by evaluating", "start_char_pos": 741, "end_char_pos": 753, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "QApairs", "after": "QA pairs", "start_char_pos": 822, "end_char_pos": 829, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "boththe", "after": "both the", "start_char_pos": 864, "end_char_pos": 871, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "showthat", "after": "show that", "start_char_pos": 1001, "end_char_pos": 1009, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 1134, "end_char_pos": 1134, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 196, 376, 615, 988], "domain": "arxiv"}
{"doc_id": "2005.14672", "revision_depth": 1, "before_revision": "Transfer learning, particularly approaches that combine multi-task learning with pre-trained contextualized embeddings and fine-tuning, have advanced the field of Natural Language Processing tremendously in recent years. In this paper we present MaChAmp, a toolkit for easy use of fine-tuning BERT-like models in multi-task settings. The benefits of MaChAmp are its flexible configuration options, and the support of a variety of NLP tasks in a uniform toolkit, from text classification to sequence labeling and dependency parsing .", "after_revision": "Transfer learning, particularly approaches that combine multi-task learning with pre-trained contextualized embeddings and fine-tuning, have advanced the field of Natural Language Processing tremendously in recent years. In this paper we present MaChAmp, a toolkit for easy fine-tuning of contextualized embeddings in multi-task settings. The benefits of MaChAmp are its flexible configuration options, and the support of a variety of natural language processing tasks in a uniform toolkit, from text classification and sequence labeling to dependency parsing, masked language modeling, and text generation .", "edit_actions": [{"type": "D", "before": "use of", "after": null, "start_char_pos": 274, "end_char_pos": 280, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "BERT-like models", "after": "of contextualized embeddings", "start_char_pos": 293, "end_char_pos": 309, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "NLP", "after": "natural language processing", "start_char_pos": 430, "end_char_pos": 433, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "to sequence labeling and dependency parsing", "after": "and sequence labeling to dependency parsing, masked language modeling, and text generation", "start_char_pos": 487, "end_char_pos": 530, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}], "sents_char_pos": [0, 220, 333], "domain": "arxiv"}
{"doc_id": "2005.14716", "revision_depth": 1, "before_revision": "The average predictability (aka informativity) of a word in context has been shown to condition word duration (Seyfarth, 2014). All else being equal, words that tend to occur in more predictable environments are shorter than words that tend to occur in less predictable environments. One account of the informativity effect on duration is that the acoustic details of word reduction are stored as part of a word's  representation. Other research has argued that predictability effects are tied to prosodic structure in integral ways. With the aim of assessing a potential prosodic basis for informativity effects in speech production, this study extends past work in two directions; it investigated informativity effects in another large language, Mandarin Chinese, and broadened the study beyond word duration to additional acoustic dimensions, pitch and intensity, known to index prosodic prominence. The acoustic information of content words was extracted from a large telephone conversation speech corpus with over 400,000 tokens and 6,000 word types spoken by 1,655 individuals and analyzed for the effect of informativity using frequency statistics estimated from a 431 million word subtitle corpus. Results indicated that words with low informativity have shorter durations, replicating the effect found in English. In addition, informativity had significant effects on maximum pitch and intensity, two phonetic dimensions related to prosodic prominence. Extending this interpretation, these results suggest that informativity is closely linked to prosodic prominence, and that  lexical representation of a word includes phonetic details associated with its prosodic prominence . In other words, the lexicon absorbs prosodic influences on speech production.", "after_revision": "The average predictability (aka informativity) of a word in context has been shown to condition word duration (Seyfarth, 2014). All else being equal, words that tend to occur in more predictable environments are shorter than words that tend to occur in less predictable environments. One account of the informativity effect on duration is that the acoustic details of probabilistic reduction are stored as part of a word's mental representation. Other research has argued that predictability effects are tied to prosodic structure in integral ways. With the aim of assessing a potential prosodic basis for informativity effects in speech production, this study extends past work in two directions; it investigated informativity effects in another large language, Mandarin Chinese, and broadened the study beyond word duration to additional acoustic dimensions, pitch and intensity, known to index prosodic prominence. The acoustic information of content words was extracted from a large telephone conversation speech corpus with over 400,000 tokens and 6,000 word types spoken by 1,655 individuals and analyzed for the effect of informativity using frequency statistics estimated from a 431 million word subtitle corpus. Results indicated that words with low informativity have shorter durations, replicating the effect found in English. In addition, informativity had significant effects on maximum pitch and intensity, two phonetic dimensions related to prosodic prominence. Extending this interpretation, these results suggest that predictability is closely linked to prosodic prominence, and that the lexical representation of a word includes phonetic details associated with its average prosodic prominence in discourse . In other words, the lexicon absorbs prosodic influences on speech production.", "edit_actions": [{"type": "R", "before": "word", "after": "probabilistic", "start_char_pos": 368, "end_char_pos": 372, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "mental", "start_char_pos": 414, "end_char_pos": 414, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "informativity", "after": "predictability", "start_char_pos": 1520, "end_char_pos": 1533, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 1585, "end_char_pos": 1585, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "prosodic prominence", "after": "average prosodic prominence in discourse", "start_char_pos": 1665, "end_char_pos": 1684, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 127, 283, 430, 533, 682, 902, 1205, 1322, 1461, 1686], "domain": "arxiv"}
{"doc_id": "2006.00119", "revision_depth": 1, "before_revision": " We present ExplainIt, a review summarization system centered around opinion explainability: the simple notion of high-level opinions (e.g. \"noisy room\") being explainable by lower-level ones (e.g., \"loud fridge\"). ExplainIt utilizes a combination of supervised and unsupervised components to mine the opinion phrasesfrom reviews URLanize them in an Opinion Causality Graph (OCG), a novel semi-structured representation which summarizes causal relations. To construct an OCG, we cluster semantically similar opinions in single nodes, thus canonicalizing opinion paraphrases, and draw directed edges between node pairs that are likely connected by a causal relation. OCGs can be used to generate structured summaries at different levels of granularity and for certain aspects of interest, while simultaneously providing explanations . In this paper, we present the system's individual components and evaluate their effectiveness on their respective sub-tasks, where we report substantial improvements over baselines across two domains. Finally, we validate these results with a user study, showing that ExplainIt produces reasonable opinion explanations according to human judges .", "after_revision": "The Web is a major resource of both factual and subjective information. While there are significant efforts URLanize factual information into knowledge bases, there is much less work URLanizing opinions, which are abundant in subjective data, into a structured format. We present ExplainIt, a system that extracts URLanizes opinions into an opinion graph, which are useful for downstream applications such as generating explainable review summaries and facilitating search over opinion phrases. In such graphs, a node represents a set of semantically similar opinions extracted from reviews and an edge between two nodes signifies that one node explains the other. ExplainIt mines explanations in a supervised method and groups similar opinions together in a weakly supervised way before combining the clusters of opinions together with their explanation relationships into an opinion graph. We experimentally demonstrate that the explanation relationships generated in the opinion graph are of good quality and our labeled datasets for explanation mining and grouping opinions are publicly available .", "edit_actions": [{"type": "A", "before": null, "after": "The Web is a major resource of both factual and subjective information. While there are significant efforts URLanize factual information into knowledge bases, there is much less work URLanizing opinions, which are abundant in subjective data, into a structured format.", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "review summarization system centered around opinion explainability: the simple notion of high-level opinions (e.g. \"noisy room\") being explainable by lower-level ones (e.g., \"loud fridge\"). ExplainIt utilizes a combination of supervised and unsupervised components to mine the opinion phrasesfrom reviews URLanize them in an Opinion Causality Graph (OCG), a novel semi-structured representation which summarizes causal relations. To construct an OCG, we cluster", "after": "system that extracts URLanizes opinions into an opinion graph, which are useful for downstream applications such as generating explainable review summaries and facilitating search over opinion phrases. In such graphs, a node represents a set of", "start_char_pos": 25, "end_char_pos": 486, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "in single nodes, thus canonicalizing opinion paraphrases, and draw directed edges between node pairs that are likely connected by a causal relation. OCGs can be used to generate structured summaries at different levels of granularity and for certain aspects of interest, while simultaneously providing explanations . In this paper, we present the system's individual components and evaluate their effectiveness on their respective sub-tasks, where we report substantial improvements over baselines across two domains. Finally, we validate these results with a user study, showing that ExplainIt produces reasonable opinion explanations according to human judges", "after": "extracted from reviews and an edge between two nodes signifies that one node explains the other. ExplainIt mines explanations in a supervised method and groups similar opinions together in a weakly supervised way before combining the clusters of opinions together with their explanation relationships into an opinion graph. We experimentally demonstrate that the explanation relationships generated in the opinion graph are of good quality and our labeled datasets for explanation mining and grouping opinions are publicly available", "start_char_pos": 517, "end_char_pos": 1178, "major_intent": "style", "raw_intents": ["style", "coherence", "style"]}], "sents_char_pos": [0, 214, 454, 665, 833, 1034], "domain": "arxiv"}
{"doc_id": "2006.00575", "revision_depth": 1, "before_revision": "In this survey, we provide a comprehensive description of recent neural entity linking (EL) systems . We distill their generic architecture that includes candidate generation , entity ranking , and unlinkable mention prediction components. For each of them, we summarize the prominent methods and models, including approaches to mention encoding based on the self-attention architecture. Since many EL models take advantage of entity embeddings to improve their generalization capabilities, we provide an overview of the widely-used entity embedding techniques. We group the variety of EL approaches by several common research directions : joint entity recognition and linking, models for global EL , domain-independent techniques including zero-shot and distant supervision methods, and cross-lingual approaches. We also discuss the novel application of EL for enhancing word representation models like BERT. We systemize the critical design features of EL systems and provide their reported evaluation results .", "after_revision": "In this survey, we provide a comprehensive description of recent neural entity linking (EL) systems developed since 2015 as a result of the \"deep learning revolution\" in NLP. Our goal is to systemize design features of neural entity linking systems and compare their performances to the best classic methods on the common benchmarks. We distill generic architectural components of a neural EL system, like candidate generation and entity ranking summarizing the prominent methods for each of them, such as approaches to mention encoding based on the self-attention architecture. The vast variety of modifications of this general neural entity linking architecture are grouped by several common themes : joint entity recognition and linking, models for global linking , domain-independent techniques including zero-shot and distant supervision methods, and cross-lingual approaches. Since many neural models take advantage of pre-trained entity embeddings to improve their generalization capabilities, we provide an overview of popular entity embedding techniques. Finally, we briefly discuss applications of entity linking, focusing on the recently emerged use-case of enhancing deep pre-trained masked language models such as BERT .", "edit_actions": [{"type": "R", "before": ". We distill their generic architecture that includes candidate generation , entity ranking , and unlinkable mention prediction components. For", "after": "developed since 2015 as a result of the \"deep learning revolution\" in NLP. Our goal is to systemize design features of neural entity linking systems and compare their performances to the best classic methods on the common benchmarks. We distill generic architectural components of a neural EL system, like candidate generation and entity ranking summarizing the prominent methods for", "start_char_pos": 100, "end_char_pos": 243, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "we summarize the prominent methods and models, including", "after": "such as", "start_char_pos": 258, "end_char_pos": 314, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Since many EL models take advantage of entity embeddings to improve their generalization capabilities, we provide an overview of the widely-used entity embedding techniques. We group the variety of EL approaches", "after": "The vast variety of modifications of this general neural entity linking architecture are grouped", "start_char_pos": 388, "end_char_pos": 599, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "research directions", "after": "themes", "start_char_pos": 618, "end_char_pos": 637, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "EL", "after": "linking", "start_char_pos": 696, "end_char_pos": 698, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "We also discuss the novel application of EL for enhancing word representation models like BERT. We systemize the critical design features of EL systems and provide their reported evaluation results", "after": "Since many neural models take advantage of pre-trained entity embeddings to improve their generalization capabilities, we provide an overview of popular entity embedding techniques. Finally, we briefly discuss applications of entity linking, focusing on the recently emerged use-case of enhancing deep pre-trained masked language models such as BERT", "start_char_pos": 814, "end_char_pos": 1011, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 101, 239, 387, 561, 813, 909], "domain": "arxiv"}
{"doc_id": "2006.00885", "revision_depth": 1, "before_revision": "As the COVID-19 virus quickly spreads around the world, unfortunately, misinformation related to COVID-19 also gets created and spreads like wild fire. Such misinformation has caused confusion among people, disruptions in society, and even deadly consequences in health problems. To be able to understand, detect, and mitigate such COVID-19 misinformation, therefore, has not only deep intellectual values but also huge societal impacts. To help researchers combat COVID-19 health misinformation, therefore, we present CoAID (Covid-19 heAlthcare mIsinformation Dataset), with diverse COVID-19 healthcare misinformation, including fake news on websites and social platforms, along with users' social engagement about such news. CoAID includes 1,896 news, 183,564 related user engagements, 516 social platform posts about COVID-19, and ground truth labels. The dataset is available at: URL", "after_revision": "As the COVID-19 virus quickly spreads around the world, unfortunately, misinformation related to COVID-19 also gets created and spreads like wild fire. Such misinformation has caused confusion among people, disruptions in society, and even deadly consequences in health problems. To be able to understand, detect, and mitigate such COVID-19 misinformation, therefore, has not only deep intellectual values but also huge societal impacts. To help researchers combat COVID-19 health misinformation, therefore, we present CoAID (Covid-19 heAlthcare mIsinformation Dataset), with diverse COVID-19 healthcare misinformation, including fake news on websites and social platforms, along with users' social engagement about such news. CoAID includes 3,235 news, 294,692 related user engagements, 851 social platform posts about COVID-19, and ground truth labels. The dataset is available at: URL", "edit_actions": [{"type": "R", "before": "1,896 news, 183,564", "after": "3,235 news, 294,692", "start_char_pos": 742, "end_char_pos": 761, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "516", "after": "851", "start_char_pos": 788, "end_char_pos": 791, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 151, 279, 437, 726, 854], "domain": "arxiv"}
{"doc_id": "2006.00885", "revision_depth": 2, "before_revision": "As the COVID-19 virus quickly spreads around the world, unfortunately, misinformation related to COVID-19 also gets created and spreads like wild fire. Such misinformation has caused confusion among people, disruptions in society, and even deadly consequences in health problems. To be able to understand, detect, and mitigate such COVID-19 misinformation, therefore, has not only deep intellectual values but also huge societal impacts. To help researchers combat COVID-19 health misinformation, therefore, we present CoAID (Covid-19 heAlthcare mIsinformation Dataset), with diverse COVID-19 healthcare misinformation, including fake news on websites and social platforms, along with users' social engagement about such news. CoAID includes 3,235 news, 294,692 related user engagements, 851 social platform posts about COVID-19, and ground truth labels. The dataset is available at: URL", "after_revision": "As the COVID-19 virus quickly spreads around the world, unfortunately, misinformation related to COVID-19 also gets created and spreads like wild fire. Such misinformation has caused confusion among people, disruptions in society, and even deadly consequences in health problems. To be able to understand, detect, and mitigate such COVID-19 misinformation, therefore, has not only deep intellectual values but also huge societal impacts. To help researchers combat COVID-19 health misinformation, therefore, we present CoAID (Covid-19 heAlthcare mIsinformation Dataset), with diverse COVID-19 healthcare misinformation, including fake news on websites and social platforms, along with users' social engagement about such news. CoAID includes 4,251 news, 296,000 related user engagements, 926 social platform posts about COVID-19, and ground truth labels. The dataset is available at: URL", "edit_actions": [{"type": "R", "before": "3,235 news, 294,692", "after": "4,251 news, 296,000", "start_char_pos": 742, "end_char_pos": 761, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "851", "after": "926", "start_char_pos": 788, "end_char_pos": 791, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 151, 279, 437, 726, 854], "domain": "arxiv"}
{"doc_id": "2006.00995", "revision_depth": 1, "before_revision": "A growing body of work makes use of probing in order to investigate the working of neural models, often considered black boxes. Recently, an ongoing debate emerged surrounding the limitations of the probing paradigm. In this work, we point out the inability to infer behavioral conclusions from probing results, and offer an alternative method which is focused on how the information is being used, rather than on what information is encoded. Our method, Amnesic Probing, follows the intuition that the utility of a property for a given task can be assessed by measuring the influence of a causal intervention which removes it from the representation. Equipped with this new analysis tool, we can now ask questions that were not possible before, e.g. is part-of-speech information important for word prediction? We perform a series of analyses on BERT to answer these types of questions. Our findings demonstrate that conventional probing performance is not correlated to task importance, and we call for increased scrutiny of claims that draw behavioral or causal conclusions from probing results.", "after_revision": "A growing body of work makes use of probing in order to investigate the working of neural models, often considered black boxes. Recently, an ongoing debate emerged surrounding the limitations of the probing paradigm. In this work, we point out the inability to infer behavioral conclusions from probing results, and offer an alternative method which focuses on how the information is being used, rather than on what information is encoded. Our method, Amnesic Probing, follows the intuition that the utility of a property for a given task can be assessed by measuring the influence of a causal intervention which removes it from the representation. Equipped with this new analysis tool, we can ask questions that were not possible before, e.g. is part-of-speech information important for word prediction? We perform a series of analyses on BERT to answer these types of questions. Our findings demonstrate that conventional probing performance is not correlated to task importance, and we call for increased scrutiny of claims that draw behavioral or causal conclusions from probing results.", "edit_actions": [{"type": "R", "before": "is focused", "after": "focuses", "start_char_pos": 350, "end_char_pos": 360, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "D", "before": "now", "after": null, "start_char_pos": 697, "end_char_pos": 700, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 127, 216, 442, 651, 811, 887], "domain": "arxiv"}
{"doc_id": "2006.00995", "revision_depth": 2, "before_revision": "A growing body of work makes use of probing in order to investigate the working of neural models, often considered black boxes. Recently, an ongoing debate emerged surrounding the limitations of the probing paradigm. In this work, we point out the inability to infer behavioral conclusions from probing results , and offer an alternative method which focuses on how the information is being used, rather than on what information is encoded. Our method, Amnesic Probing, follows the intuition that the utility of a property for a given task can be assessed by measuring the influence of a causal intervention which removes it from the representation. Equipped with this new analysis tool, we can ask questions that were not possible before, e.g. is part-of-speech information important for word prediction? We perform a series of analyses on BERT to answer these types of questions. Our findings demonstrate that conventional probing performance is not correlated to task importance, and we call for increased scrutiny of claims that draw behavioral or causal conclusions from probing results.", "after_revision": "A growing body of work makes use of probing to investigate the working of neural models, often considered black boxes. Recently, an ongoing debate emerged surrounding the limitations of the probing paradigm. In this work, we point out the inability to infer behavioral conclusions from probing results and offer an alternative method that focuses on how the information is being used, rather than on what information is encoded. Our method, Amnesic Probing, follows the intuition that the utility of a property for a given task can be assessed by measuring the influence of a causal intervention that removes it from the representation. Equipped with this new analysis tool, we can ask questions that were not possible before, e.g. is part-of-speech information important for word prediction? We perform a series of analyses on BERT to answer these types of questions. Our findings demonstrate that conventional probing performance is not correlated to task importance, and we call for increased scrutiny of claims that draw behavioral or causal conclusions from probing results.", "edit_actions": [{"type": "D", "before": "in order", "after": null, "start_char_pos": 44, "end_char_pos": 52, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 311, "end_char_pos": 312, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "which", "after": "that", "start_char_pos": 345, "end_char_pos": 350, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "which", "after": "that", "start_char_pos": 608, "end_char_pos": 613, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 127, 216, 440, 649, 805, 881], "domain": "arxiv"}
{"doc_id": "2006.01095", "revision_depth": 1, "before_revision": "Artificial neural networks ( ANNS have shown much empirical success in solving perceptual tasks across various cognitive modalities. While they are only loosely inspired by the biological brain, recent studies report considerable similarities between representation extracted from task-optimized ANNS and neural populations in the brain. ANNS have subsequently become a popular model class to infer computational principles underlying complex cognitive functions, and in turn they have also emerged as a natural testbed for applying methods originally developed to probe information in neural populations. In this work, we utilize mean-field theoretic manifold analysis, a recent technique from computational neuroscience, to analyze the high dimensional geometry of language representations from large-scale contextual embedding models. We explore representations from different model families (BERT, RoBERTa, GPT-2, etc. ) and find evidence for emergence of linguistic manifold across layer depth (e.g., manifolds for part-of-speech and combinatory categorical grammar tags). We further observe that different encoding schemes used to obtain the representations lead to differences in whether these linguistic manifolds emerge in earlier or later layers of the network. In addition, we find that the emergence of linear separability in these manifolds is driven by a combined reduction of manifolds radius, dimensionality and inter-manifold correlations.", "after_revision": "Artificial neural networks ( ANNs) have shown much empirical success in solving perceptual tasks across various cognitive modalities. While they are only loosely inspired by the biological brain, recent studies report considerable similarities between representation extracted from task-optimized ANNs and neural populations in the brain. ANNs have subsequently become a popular model class to infer computational principles underlying complex cognitive functions, and in turn they have also emerged as a natural testbed for applying methods originally developed to probe information in neural populations. In this work, we utilize mean-field theoretic manifold analysis, a recent technique from computational neuroscience, to analyze the high dimensional geometry of language representations from large-scale contextual embedding models. We explore representations from different model families (BERT, RoBERTa, GPT-2, etc. ) and find evidence for emergence of linguistic manifold across layer depth (e.g., manifolds for part-of-speech and combinatory categorical grammar tags). We further observe that different encoding schemes used to obtain the representations lead to differences in whether these linguistic manifolds emerge in earlier or later layers of the network. In addition, we find that the emergence of linear separability in these manifolds is driven by a combined reduction of manifolds radius, dimensionality and inter-manifold correlations.", "edit_actions": [{"type": "R", "before": "ANNS", "after": "ANNs)", "start_char_pos": 29, "end_char_pos": 33, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "meaning-changed"]}, {"type": "R", "before": "ANNS", "after": "ANNs", "start_char_pos": 296, "end_char_pos": 300, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "ANNS", "after": "ANNs", "start_char_pos": 338, "end_char_pos": 342, "major_intent": "fluency", "raw_intents": ["meaning-changed", "fluency", "fluency"]}], "sents_char_pos": [0, 132, 337, 605, 837, 1077, 1271], "domain": "arxiv"}
{"doc_id": "2006.01095", "revision_depth": 2, "before_revision": "Artificial neural networks ( ANNs ) have shown much empirical success in solving perceptual tasks across various cognitive modalities. While they are only loosely inspired by the biological brain, recent studies report considerable similarities between representation extracted from task-optimized ANNs and neural populations in the brain. ANNs have subsequently become a popular model class to infer computational principles underlying complex cognitive functions, and in turn  they have also emerged as a natural testbed for applying methods originally developed to probe information in neural populations. In this work, we utilize mean-field theoretic manifold analysis, a recent technique from computational neuroscience  , to analyze the high dimensional geometry of language representations from large-scale contextual embedding models. We explore representations from different model families (BERT, RoBERTa, GPT-2 , etc.) and find evidence for emergence of linguistic manifold across layer depth (e.g., manifolds for part-of-speech and combinatory categorical grammar tags). We further observe that different encoding schemes used to obtain the representations lead to differences in whether these linguistic manifolds emerge in earlier or later layers of the network . In addition, we find that the emergence of linear separability in these manifolds is driven by a combined reduction of manifolds  radius, dimensionality and inter-manifold correlations.", "after_revision": "Deep neural networks ( DNNs ) have shown much empirical success in solving perceptual tasks across various cognitive modalities. While they are only loosely inspired by the biological brain, recent studies report considerable similarities between representations extracted from task-optimized DNNs and neural populations in the brain. DNNs have subsequently become a popular model class to infer computational principles underlying complex cognitive functions, and in turn , they have also emerged as a natural testbed for applying methods originally developed to probe information in neural populations. In this work, we utilize mean-field theoretic manifold analysis, a recent technique from computational neuroscience that connects geometry of feature representations with linear separability of classes , to analyze language representations from large-scale contextual embedding models. We explore representations from different model families (BERT, RoBERTa, GPT , etc.) and find evidence for emergence of linguistic manifolds across layer depth (e.g., manifolds for part-of-speech tags), especially in ambiguous data (i.e, words with multiple part-of-speech tags, or part-of-speech classes including many words) . In addition, we find that the emergence of linear separability in these manifolds is driven by a combined reduction of manifolds ' radius, dimensionality and inter-manifold correlations.", "edit_actions": [{"type": "R", "before": "Artificial", "after": "Deep", "start_char_pos": 0, "end_char_pos": 10, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "others", "meaning-changed"]}, {"type": "R", "before": "ANNs", "after": "DNNs", "start_char_pos": 29, "end_char_pos": 33, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}, {"type": "R", "before": "representation", "after": "representations", "start_char_pos": 253, "end_char_pos": 267, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "ANNs", "after": "DNNs", "start_char_pos": 298, "end_char_pos": 302, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "ANNs", "after": "DNNs", "start_char_pos": 340, "end_char_pos": 344, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 478, "end_char_pos": 478, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "that connects geometry of feature representations with linear separability of classes", "start_char_pos": 725, "end_char_pos": 725, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "the high dimensional geometry of", "after": null, "start_char_pos": 739, "end_char_pos": 771, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "GPT-2", "after": "GPT", "start_char_pos": 916, "end_char_pos": 921, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "manifold", "after": "manifolds", "start_char_pos": 976, "end_char_pos": 984, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and combinatory categorical grammar tags). We further observe that different encoding schemes used to obtain the representations lead to differences in whether these linguistic manifolds emerge in earlier or later layers of the network", "after": "tags), especially in ambiguous data (i.e, words with multiple part-of-speech tags, or part-of-speech classes including many words)", "start_char_pos": 1040, "end_char_pos": 1275, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "A", "before": null, "after": "'", "start_char_pos": 1407, "end_char_pos": 1407, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}], "sents_char_pos": [0, 134, 339, 608, 842, 1082, 1277], "domain": "arxiv"}
{"doc_id": "2006.02163", "revision_depth": 1, "before_revision": "Recent unsupervised machine translation (UMT) systems usually employ three main principles: initialization, language modeling and iterative back-translation, though they may apply these principles differently. This work introduces another component to this framework : Multi-Agent Cross-translated Diversification (MACD). The method trains multiple UMT agents and then translates monolingual data back and forth using non-duplicative agents to acquire synthetic parallel data for supervised MT. MACD is applicable to all previous UMT approaches. In our experiments, the technique boosts the performance for some commonly used UMT methods by 1.5-2.0 BLEU. In particular, in WMT'14 English-French, WMT'16 German-English and English-Romanian, MACD outperforms cross-lingual masked language model pretraining by 2.3, 2.2 and 1.6 BLEU, respectively. It also yields 1.5 -3.3 BLEU improvements in IWSLT English-French and English-German translation tasks. Through extensive experimental analyses, we show that MACD is effective because it embraces data diversity while other similar variants do not.", "after_revision": "Recent unsupervised machine translation (UMT) systems usually employ three main principles: initialization, language modeling and iterative back-translation, though they may apply them differently. Crucially, iterative back-translation and denoising auto-encoding for language modeling provide data diversity to train the UMT systems. However, the gains from these diversification processes has seemed to plateau. We introduce a novel component to the standard UMT framework called Cross-model Back-translated Distillation (CBD), that is aimed to induce another level of data diversification that existing principles lack. CBD is applicable to all previous UMT approaches. In our experiments, it boosts the performance of the standard UMT methods by 1.5-2.0 BLEU. In particular, in WMT'14 English-French, WMT'16 German-English and English-Romanian, CBD outperforms cross-lingual masked language model (XLM) by 2.3, 2.2 and 1.6 BLEU, respectively. It also yields 1.5 --3.3 BLEU improvements in IWSLT English-French and English-German tasks. Through extensive experimental analyses, we show that CBD is effective because it embraces data diversity while other similar variants do not.", "edit_actions": [{"type": "R", "before": "these principles differently. This work introduces another component to this framework : Multi-Agent Cross-translated Diversification (MACD). The method trains multiple UMT agents and then translates monolingual data back and forth using non-duplicative agents to acquire synthetic parallel data for supervised MT. MACD", "after": "them differently. Crucially, iterative back-translation and denoising auto-encoding for language modeling provide data diversity to train the UMT systems. However, the gains from these diversification processes has seemed to plateau. We introduce a novel component to the standard UMT framework called Cross-model Back-translated Distillation (CBD), that is aimed to induce another level of data diversification that existing principles lack. CBD", "start_char_pos": 180, "end_char_pos": 499, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the technique", "after": "it", "start_char_pos": 566, "end_char_pos": 579, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "for some commonly used", "after": "of the standard", "start_char_pos": 603, "end_char_pos": 625, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "MACD", "after": "CBD", "start_char_pos": 740, "end_char_pos": 744, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "pretraining", "after": "(XLM)", "start_char_pos": 793, "end_char_pos": 804, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "-3.3", "after": "--3.3", "start_char_pos": 864, "end_char_pos": 868, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "translation", "after": null, "start_char_pos": 930, "end_char_pos": 941, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "MACD", "after": "CBD", "start_char_pos": 1003, "end_char_pos": 1007, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 209, 321, 545, 654, 844, 948], "domain": "arxiv"}
{"doc_id": "2006.02163", "revision_depth": 2, "before_revision": "Recent unsupervised machine translation (UMT) systems usually employ three main principles: initialization, language modeling and iterative back-translation, though they may apply them differently. Crucially, iterative back-translation and denoising auto-encoding for language modeling provide data diversity to train the UMT systems. However, the gains from these diversification processes has seemed to plateau. We introduce a novel component to the standard UMT framework called Cross-model Back-translated Distillation (CBD), that is aimed to induce another level of data diversification that existing principles lack. CBD is applicable to all previous UMT approaches. In our experiments, it boosts the performance of the standard UMT methods by 1.5-2.0 BLEU. In particular, in WMT'14 English-French, WMT'16 German-English and English-Romanian , CBD outperforms cross-lingual masked language model (XLM) by 2.3, 2.2 and 1.6 BLEU , respectively. It also yields 1.5--3.3 BLEU improvements in IWSLT English-French and English-German tasks. Through extensive experimental analyses, we show that CBD is effective because it embraces data diversity while other similar variants do not.", "after_revision": "Recent unsupervised machine translation (UMT) systems usually employ three main principles: initialization, language modeling and iterative back-translation, though they may apply them differently. Crucially, iterative back-translation and denoising auto-encoding for language modeling provide data diversity to train the UMT systems. However, the gains from these diversification processes has seemed to plateau. We introduce a novel component to the standard UMT framework called Cross-model Back-translated Distillation (CBD), that is aimed to induce another level of data diversification that existing principles lack. CBD is applicable to all previous UMT approaches. In our experiments, CBD achieves the state of the art in the WMT'14 English-French, WMT'16 English-German and English-Romanian bilingual unsupervised translation tasks, with 38.2, 30.1, and 36.3 BLEU respectively. It also yields 1.5--3.3 BLEU improvements in IWSLT English-French and English-German tasks. Through extensive experimental analyses, we show that CBD is effective because it embraces data diversity while other similar variants do not.", "edit_actions": [{"type": "R", "before": "it boosts the performance of the standard UMT methods by 1.5-2.0 BLEU. In particular, in", "after": "CBD achieves the state of the art in the", "start_char_pos": 693, "end_char_pos": 781, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "German-English", "after": "English-German", "start_char_pos": 812, "end_char_pos": 826, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": ", CBD outperforms cross-lingual masked language model (XLM) by 2.3, 2.2 and 1.6 BLEU ,", "after": "bilingual unsupervised translation tasks, with 38.2, 30.1, and 36.3 BLEU", "start_char_pos": 848, "end_char_pos": 934, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 197, 334, 413, 622, 672, 763, 948, 1040], "domain": "arxiv"}
{"doc_id": "2006.02876", "revision_depth": 1, "before_revision": "Improving neural machine translation (NMT) models using the back-translations of the monolingual target data (synthetic parallel data) is currently the state-of-the-art approach for training improved translation systems. The quality of the backward system - which is trained on the available parallel data and used for the back-translation - has been shown in many studies to affect the performance of the final NMT model. In low resource conditions, the available parallel data is usually not enough to train a backward model that can produce the qualitative synthetic data needed to train a standard translation model . This work proposes a self-training strategy where the output of the backward model is used to improve the model itself through the forward translation technique. The technique was shown to improve baseline low resource IWSLT'14 English-German and IWSLT'15 English-Vietnamese backward translation models by 11.06 and 1.5 BLEUs respectively. The synthetic data generated by the improved English-German backward model was used to train a forward model which out-performed another forward model trained using standard back-translation by 2.7 BLEU .", "after_revision": "Improving neural machine translation (NMT) models using the back-translations of the monolingual target data (synthetic parallel data) is currently the state-of-the-art approach for training improved translation systems. The target-side side monolingual data has been used in the back-translation approach to improve the forward (target) translation model. Whereas the success of the approach heavily relies on the additional parallel data generating model -- the backward model -- the aim of the approach is only targeted at improving the forward model. The back-translation approach was designed primarily to benefit from an additional data whose source-side is synthetic. But research works have shown that translation models can also benefit from additional data whose target-side is synthetic . This work proposes the use of the target-side data throughout the back-translation approach to improve both the backward and forward models. We explored using only the target-side monolingual data to improve the backward model through forward translation and the forward model through back-translation. Experimental results on English-German and English-Vietnamese low resource neural machine translation showed that the proposed approach outperforms baselines that use the traditional back-translation method .", "edit_actions": [{"type": "R", "before": "quality of the backward system - which is trained on the available parallel data and used for the", "after": "target-side side monolingual data has been used in the", "start_char_pos": 225, "end_char_pos": 322, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "- has been shown in many studies to affect the performance of the final NMT model. In low resource conditions, the available parallel data is usually not enough to train a backward model that can produce the qualitative synthetic data needed to train a standard translation model", "after": "approach to improve the forward (target) translation model. Whereas the success of the approach heavily relies on the additional parallel data generating model -- the backward model -- the aim of the approach is only targeted at improving the forward model. The back-translation approach was designed primarily to benefit from an additional data whose source-side is synthetic. But research works have shown that translation models can also benefit from additional data whose target-side is synthetic", "start_char_pos": 340, "end_char_pos": 619, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "a self-training strategy where the output of the backward model is used", "after": "the use of the target-side data throughout the back-translation approach to improve both the backward and forward models. We explored using only the target-side monolingual data", "start_char_pos": 641, "end_char_pos": 712, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "model itself through the forward translation technique. The technique was shown to improve baseline low resource IWSLT'14", "after": "backward model through forward translation and the forward model through back-translation. Experimental results on", "start_char_pos": 728, "end_char_pos": 849, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "D", "before": "IWSLT'15", "after": null, "start_char_pos": 869, "end_char_pos": 877, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "backward translation models by 11.06 and 1.5 BLEUs respectively. The synthetic data generated by the improved English-German backward model was used to train a forward model which out-performed another forward model trained using standard", "after": "low resource neural machine translation showed that the proposed approach outperforms baselines that use the traditional", "start_char_pos": 897, "end_char_pos": 1135, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "by 2.7 BLEU", "after": "method", "start_char_pos": 1153, "end_char_pos": 1164, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 220, 422, 621, 783, 961]}
{"doc_id": "2006.02876", "revision_depth": 2, "before_revision": "Improving neural machine translation (NMT) models using the back-translations of the monolingual target data (synthetic parallel data) is currently the state-of-the-art approach for training improved translation systems. The target-side side monolingual data has been used in the back-translation approach to improve the forward (target) translation model. Whereas the success of the approach heavily relies on the additional parallel data generating model -- the backward model -- the aim of the approach is only targeted at improving the forward model. The back-translation approach was designed primarily to benefit from an additional data whose source-side is synthetic. But research works have shown that translation models can also benefit from additional data whose target-side is synthetic . This work proposes the use of the target-side data throughout the back-translation approach to improve both the backward and forward models. We explored using only the target-side monolingual data to improve the backward model through forward translation and the forward model through back-translation. Experimental results on English-German and  English-Vietnamese low resource neural machine translation showed that the proposed approach outperforms baselines that use the traditional back-translation method .", "after_revision": "Improving neural machine translation (NMT) models using the back-translations of the monolingual target data (synthetic parallel data) is currently the state-of-the-art approach for training improved translation systems. The quality of the backward system - which is trained on the available parallel data and used for the back-translation - has been shown in many studies to affect the performance of the final NMT model. In low resource conditions, the available parallel data is usually not enough to train a backward model that can produce the qualitative synthetic data needed to train a standard translation model . This work proposes a self-training strategy where the output of the backward model is used to improve the model itself through the forward translation technique. The technique was shown to improve baseline low resource IWSLT'14 English-German and IWSLT'15 English-Vietnamese backward translation models by 11.06 and 1.5 BLEUs respectively. The synthetic data generated by the improved English-German backward model was used to train a forward model which out-performed another forward model trained using standard back-translation by 2.7 BLEU .", "edit_actions": [{"type": "R", "before": "target-side side monolingual data has been used in the", "after": "quality of the backward system - which is trained on the available parallel data and used for the", "start_char_pos": 225, "end_char_pos": 279, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "approach to improve the forward (target) translation model. Whereas the success of the approach heavily relies on the additional parallel data generating model -- the backward model -- the aim of the approach is only targeted at improving the forward model. The back-translation approach was designed primarily to benefit from an additional data whose source-side is synthetic. But research works have shown that translation models can also benefit from additional data whose target-side is synthetic", "after": "- has been shown in many studies to affect the performance of the final NMT model. In low resource conditions, the available parallel data is usually not enough to train a backward model that can produce the qualitative synthetic data needed to train a standard translation model", "start_char_pos": 297, "end_char_pos": 797, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "the use of the target-side data throughout the back-translation approach to improve both the backward and forward models. We explored using only the target-side monolingual data", "after": "a self-training strategy where the output of the backward model is used", "start_char_pos": 819, "end_char_pos": 996, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "backward model through forward translation and the forward model through back-translation. Experimental results on", "after": "model itself through the forward translation technique. The technique was shown to improve baseline low resource IWSLT'14", "start_char_pos": 1012, "end_char_pos": 1126, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "IWSLT'15", "start_char_pos": 1146, "end_char_pos": 1146, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "low resource neural machine translation showed that the proposed approach outperforms baselines that use the traditional", "after": "backward translation models by 11.06 and 1.5 BLEUs respectively. The synthetic data generated by the improved English-German backward model was used to train a forward model which out-performed another forward model trained using standard", "start_char_pos": 1166, "end_char_pos": 1286, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "method", "after": "by 2.7 BLEU", "start_char_pos": 1304, "end_char_pos": 1310, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 220, 356, 554, 674, 799, 940, 1102]}
{"doc_id": "2006.03644", "revision_depth": 1, "before_revision": "Stance detection on social media is an emerging opinion mining paradigm for various social and political applications wheresentiment analysis might be seen sub-optimal. This paper surveys the work on stance detection and situates its usage withincurrent opinion mining techniques in social media. An exhaustive review of stance detection techniques on social media ispresented ,including the task definition, the different types of targets in stance detection, the features set used, and the variousmachine learning approaches applied. The survey reports the state-of-the-art results on the existing benchmark datasets onstance detection, and discusses the most effective approaches. In addition, this study explores the emerging trends and the different applications of stance detection on social media. The study concludes by providing discussion of the gabs in the current existing research and highlighting the possible futuredirections for stance detection on social media ", "after_revision": "Stance detection on social media is an emerging opinion mining paradigm for various social and political applications where sentiment analysis might be sub-optimal. This paper surveys the work on stance detection and situates its usage within current opinion mining techniques in social media. An exhaustive review of stance detection techniques on social media is presented ,including the task definition, the different types of targets in stance detection, the features set used, and the various machine learning approaches applied. The survey reports the state-of-the-art results on the existing benchmark datasets on stance detection, and discusses the most effective approaches. In addition, this study explores the emerging trends and the different applications of stance detection on social media. The study concludes by providing discussion of the gaps in the current existing research and highlighting the possible future directions for stance detection on social media .", "edit_actions": [{"type": "R", "before": "wheresentiment", "after": "where sentiment", "start_char_pos": 118, "end_char_pos": 132, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "seen", "after": null, "start_char_pos": 151, "end_char_pos": 155, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "withincurrent", "after": "within current", "start_char_pos": 240, "end_char_pos": 253, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "ispresented", "after": "is presented", "start_char_pos": 365, "end_char_pos": 376, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "variousmachine", "after": "various machine", "start_char_pos": 492, "end_char_pos": 506, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "onstance", "after": "on stance", "start_char_pos": 619, "end_char_pos": 627, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "gabs", "after": "gaps", "start_char_pos": 856, "end_char_pos": 860, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "futuredirections", "after": "future directions", "start_char_pos": 924, "end_char_pos": 940, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 978, "end_char_pos": 978, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 168, 296, 535, 683, 804], "domain": "arxiv"}
{"doc_id": "2006.03654", "revision_depth": 1, "before_revision": "Recent progress in pre-trained neural language models has significantly improved the performance of many natural language processing (NLP) tasks. In this paper we propose a new model architecture DeBERTa(Decoding-enhanced BERT with disentangled attention) that improves the BERT and RoBERTa models using two novel techniques. The first is the disentangled attention mechanism, where each word is represented using two vectors that encode its content and position, respectively, and the attention weights among words are computed using disentangled matrices on their contents and relative positions. Second, an enhanced mask decoder is used to replace the output softmax layer to predict the masked tokens for model pretraining . We show that these two techniques significantly improve the efficiency of model pre-training and performance of downstream tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half of the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9\\% (90.2\\% vs. 91.1\\%), on SQuAD v2.0 by +2.3\\% (88.4\\% vs. 90.7\\%) and RACE by +3.6\\% (83.2\\% vs. 86.8\\%). The DeBERTa code and pre-trained models will be made publicly available at URL", "after_revision": "Recent progress in pre-trained neural language models has significantly improved the performance of many natural language processing (NLP) tasks. In this paper wepropose a new model architecture DeBERTa(Decoding-enhanced BERT with dis-entangled attention) that improves the BERT and RoBERTa models using two novel techniques. The first is the disentangled attention mechanism, where each word is represented using two vectors that encode its content and position, respectively, and the attention weights among words are computed using disentangled matrices on their contents and relative positions. Second, an enhanced mask decoder is used to incorporate absolute positions in the decoding layer to predict the masked tokens in model pre-training . We show that these two techniques significantly improve the efficiency of model pre-training and the performance of both natural languageunderstand (NLU) and natural langauge generation (NLG) tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half of the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9\\% (90.2\\% vs. 91.1\\%), on SQuAD v2.0 by +2.3\\% (88.4\\% vs. 90.7\\%) and RACE by +3.6\\% (83.2\\% vs. 86.8\\%). Notably, we scale up DeBERTa to 1.5 billion parameters and it substantially outperforms Google's T5 with 11 billionparameters on the SuperGLUE benchmark (Wang et al., 2019a) and, for the first time, surpasses the human performance (89.9 vs. 89.8).", "edit_actions": [{"type": "R", "before": "we propose", "after": "wepropose", "start_char_pos": 160, "end_char_pos": 170, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": "disentangled", "after": "dis-entangled", "start_char_pos": 232, "end_char_pos": 244, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "replace the output softmax", "after": "incorporate absolute positions in the decoding", "start_char_pos": 643, "end_char_pos": 669, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "for model pretraining", "after": "in model pre-training", "start_char_pos": 705, "end_char_pos": 726, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "performance of downstream", "after": "the performance of both natural languageunderstand (NLU) and natural langauge generation (NLG)", "start_char_pos": 826, "end_char_pos": 851, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "The DeBERTa code and pre-trained models will be made publicly available at URL", "after": "Notably, we scale up DeBERTa to 1.5 billion parameters and it substantially outperforms Google's T5 with 11 billionparameters on the SuperGLUE benchmark (Wang et al., 2019a) and, for the first time, surpasses the human performance (89.9 vs. 89.8).", "start_char_pos": 1144, "end_char_pos": 1222, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 145, 325, 598, 728, 858, 1143], "domain": "arxiv"}
{"doc_id": "2006.04315", "revision_depth": 1, "before_revision": "Visual Question Answering (VQA ) models tend to rely on the language bias and thus fail to learn the reasoning from visual knowledge , which is however the original intention of VQA . In this paper, we propose a novel cause-effect look at the language bias , where the bias is formulated as the direct  effect of question on answer from the view of causal inference. The effectcan be captured by counterfactual VQA, where the image had not existed in an imagined scenario. Our proposed cause-effect look 1) is general to any baseline VQA architecture , 2) achieves significant improvement on the language-bias sensitive VQA-CP dataset , and 3) fills the theoretical gap in recent language prior based works .", "after_revision": "Recent VQA models may tend to rely on language bias as a shortcut and thus fail to sufficiently learn the multi-modal knowledge from both vision and language . In this paper, we investigate how to capture and mitigate language bias in VQA. Motivated by causal effects, we proposed a novel counterfactual inference framework, which enables us to capture the language bias as the direct causal effect of questions on answers and reduce the language bias by subtracting the direct language effect from the total causal effect . Experiments demonstrate that our proposed counterfactual inference framework 1) is general to various VQA backbones and fusion strategies , 2) achieves competitive performance on the language-bias sensitive VQA-CP dataset while performs robustly on the balanced VQA v2 dataset .", "edit_actions": [{"type": "R", "before": "Visual Question Answering (VQA ) models", "after": "Recent VQA models may", "start_char_pos": 0, "end_char_pos": 39, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "the language bias", "after": "language bias as a shortcut", "start_char_pos": 56, "end_char_pos": 73, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "learn the reasoning from visual knowledge , which is however the original intention of VQA", "after": "sufficiently learn the multi-modal knowledge from both vision and language", "start_char_pos": 91, "end_char_pos": 181, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "propose a novel cause-effect look at", "after": "investigate how to capture and mitigate language bias in VQA. Motivated by causal effects, we proposed a novel counterfactual inference framework, which enables us to capture", "start_char_pos": 202, "end_char_pos": 238, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ", where the bias is formulated", "after": null, "start_char_pos": 257, "end_char_pos": 287, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "causal effect of questions on answers and reduce the language bias by subtracting the direct language effect from the total causal", "start_char_pos": 302, "end_char_pos": 302, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "of question on answer from the view of causal inference. The effectcan be captured by counterfactual VQA, where the image had not existed in an imagined scenario. Our proposed cause-effect look", "after": ". Experiments demonstrate that our proposed counterfactual inference framework", "start_char_pos": 310, "end_char_pos": 503, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "meaning-changed"]}, {"type": "R", "before": "any baseline VQA architecture", "after": "various VQA backbones and fusion strategies", "start_char_pos": 521, "end_char_pos": 550, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "significant improvement", "after": "competitive performance", "start_char_pos": 565, "end_char_pos": 588, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": ", and 3) fills the theoretical gap in recent language prior based works", "after": "while performs robustly on the balanced VQA v2 dataset", "start_char_pos": 635, "end_char_pos": 706, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 183, 366, 472], "domain": "arxiv"}
{"doc_id": "2006.04315", "revision_depth": 2, "before_revision": "Recent VQA models may tend to rely on language bias as a shortcut and thus fail to sufficiently learn the multi-modal knowledge from both vision and language.  In this paper, we investigate how to capture and mitigate language bias in VQA. Motivated by causal effects, we proposed a novel counterfactual inference framework, which enables us to capture the language bias as the direct causal effect of questions on answers and reduce the language bias by subtracting the direct language effect from the total causal effect. Experiments demonstrate that our proposed counterfactual inference framework 1) is general to various VQA backbones and fusion strategies, 2) achieves competitive performance on the language-bias sensitive VQA-CP dataset while performs robustly on the balanced VQA v2 dataset .", "after_revision": " VQA models may tend to rely on language bias as a shortcut and thus fail to sufficiently learn the multi-modal knowledge from both vision and language. Recent debiasing methods proposed to exclude the language prior during inference. However, they fail to disentangle the \"good\" language context and \"bad\" language bias from the whole. In this paper, we investigate how to mitigate language bias in VQA. Motivated by causal effects, we proposed a novel counterfactual inference framework, which enables us to capture the language bias as the direct causal effect of questions on answers and reduce the language bias by subtracting the direct language effect from the total causal effect. Experiments demonstrate that our proposed counterfactual inference framework 1) is general to various VQA backbones and fusion strategies, 2) achieves competitive performance on the language-bias sensitive VQA-CP dataset while performs robustly on the balanced VQA v2 dataset without any augmented data. The code is available at URL", "edit_actions": [{"type": "D", "before": "Recent", "after": null, "start_char_pos": 0, "end_char_pos": 6, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Recent debiasing methods proposed to exclude the language prior during inference. However, they fail to disentangle the \"good\" language context and \"bad\" language bias from the whole.", "start_char_pos": 159, "end_char_pos": 159, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "capture and", "after": null, "start_char_pos": 197, "end_char_pos": 208, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": ".", "after": "without any augmented data. The code is available at URL", "start_char_pos": 800, "end_char_pos": 801, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 158, 239, 523], "domain": "arxiv"}
{"doc_id": "2006.06814", "revision_depth": 1, "before_revision": "Designing task-oriented dialogue systems is a challenging research topic, since it needs not only to generate utterances fulfilling user requests but also to guarantee the comprehensibility. Many previous works trained end-to-end (E2E) models with supervised learning (SL), however, the bias in annotated system utterances remains as a bottleneck. Reinforcement learning (RL) deals with the problem through using non-differentiable evaluation metrics (e.g., the success rate) as rewards. Nonetheless, existing works with RL showed that the comprehensibility of generated system utterances could be corrupted when improving the performance on fulfilling user requests. In our work, we (1) propose modelling the hierarchical structure between dialogue policy and natural language generator (NLG) with the option framework, called HDNO  ; (2) train HDNO with hierarchical reinforcement learning (HRL), as well as suggest alternating updates between dialogue policy and NLG during HRL inspired by fictitious play, to preserve the comprehensibility of generated system utterances while improving fulfilling user requests ; and (3) propose using a discriminator modelled with language models as an additional reward to further improve the comprehensibility. We test HDNO on MultiWoz 2.0 and MultiWoz 2.1, the datasets on multi-domain dialogues, in comparison with word-level E2E model trained with RL, LaRL and HDSA, showing a significant improvement on the total performance evaluated with automatic metrics .", "after_revision": "Designing task-oriented dialogue systems is a challenging research topic, since it needs not only to generate utterances fulfilling user requests but also to guarantee the comprehensibility. Many previous works trained end-to-end (E2E) models with supervised learning (SL), however, the bias in annotated system utterances remains as a bottleneck. Reinforcement learning (RL) deals with the problem through using non-differentiable evaluation metrics (e.g., the success rate) as rewards. Nonetheless, existing works with RL showed that the comprehensibility of generated system utterances could be corrupted when improving the performance on fulfilling user requests. In o gur work, we (1) propose modelling the hierarchical structure between dialogue policy and natural language generator (NLG) with the option framework, called HDNO , where the latent dialogue act is applied to avoid designing specific dialogue act representations ; (2) train HDNO via hierarchical reinforcement learning (HRL), as well as suggest the asynchronous updates between dialogue policy and NLG during training to theoretically guarantee their convergence to a local maximizer ; and (3) propose using a discriminator modelled with language models as an additional reward to further improve the comprehensibility. We test HDNO on MultiWoz 2.0 and MultiWoz 2.1, the datasets on multi-domain dialogues, in comparison with word-level E2E model trained with RL, LaRL and HDSA, showing improvements on the performance evaluated by automatic evaluation metrics and human evaluation. Finally, we demonstrate the semantic meanings of latent dialogue acts to show the ability of explanation .", "edit_actions": [{"type": "R", "before": "our", "after": "o gur", "start_char_pos": 671, "end_char_pos": 674, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "A", "before": null, "after": ", where the latent dialogue act is applied to avoid designing specific dialogue act representations", "start_char_pos": 833, "end_char_pos": 833, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "with", "after": "via", "start_char_pos": 851, "end_char_pos": 855, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "alternating", "after": "the asynchronous", "start_char_pos": 918, "end_char_pos": 929, "major_intent": "style", "raw_intents": ["style", "meaning-changed", "style"]}, {"type": "R", "before": "HRL inspired by fictitious play, to preserve the comprehensibility of generated system utterances while improving fulfilling user requests", "after": "training to theoretically guarantee their convergence to a local maximizer", "start_char_pos": 977, "end_char_pos": 1115, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "a significant improvement on the total performance evaluated with automatic metrics", "after": "improvements on the performance evaluated by automatic evaluation metrics and human evaluation. Finally, we demonstrate the semantic meanings of latent dialogue acts to show the ability of explanation", "start_char_pos": 1419, "end_char_pos": 1502, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 190, 347, 487, 667, 835, 1117, 1251], "domain": "arxiv"}
{"doc_id": "2006.06814", "revision_depth": 2, "before_revision": "Designing task-oriented dialogue systems is a challenging research topic, since it needs not only to generate utterances fulfilling user requests but also to guarantee the comprehensibility. Many previous works trained end-to-end (E2E) models with supervised learning (SL), however, the bias in annotated system utterances remains as a bottleneck. Reinforcement learning (RL) deals with the problem through using non-differentiable evaluation metrics (e.g., the success rate) as rewards. Nonetheless, existing works with RL showed that the comprehensibility of generated system utterances could be corrupted when improving the performance on fulfilling user requests. In o gur work, we (1) propose modelling the hierarchical structure between dialogue policy and natural language generator (NLG) with the option framework, called HDNO, where the latent dialogue act is applied to avoid designing specific dialogue act representations; (2) train HDNO via hierarchical reinforcement learning (HRL), as well as suggest the asynchronous updates between dialogue policy and NLG during training to theoretically guarantee their convergence to a local maximizer; and (3) propose using a discriminator modelled with language models as an additional reward to further improve the comprehensibility. We test HDNO on MultiWoz 2.0 and MultiWoz 2.1, the datasets on multi-domain dialogues, in comparison with word-level E2E model trained with RL, LaRL and HDSA, showing improvements on the performance evaluated by automatic evaluation metrics and human evaluation. Finally, we demonstrate the semantic meanings of latent dialogue acts to show the ability of explanation .", "after_revision": "Designing task-oriented dialogue systems is a challenging research topic, since it needs not only to generate utterances fulfilling user requests but also to guarantee the comprehensibility. Many previous works trained end-to-end (E2E) models with supervised learning (SL), however, the bias in annotated system utterances remains as a bottleneck. Reinforcement learning (RL) deals with the problem through using non-differentiable evaluation metrics (e.g., the success rate) as rewards. Nonetheless, existing works with RL showed that the comprehensibility of generated system utterances could be corrupted when improving the performance on fulfilling user requests. In our work, we (1) propose modelling the hierarchical structure between dialogue policy and natural language generator (NLG) with the option framework, called HDNO, where the latent dialogue act is applied to avoid designing specific dialogue act representations; (2) train HDNO via hierarchical reinforcement learning (HRL), as well as suggest the asynchronous updates between dialogue policy and NLG during training to theoretically guarantee their convergence to a local maximizer; and (3) propose using a discriminator modelled with language models as an additional reward to further improve the comprehensibility. We test HDNO on MultiWoz 2.0 and MultiWoz 2.1, the datasets on multi-domain dialogues, in comparison with word-level E2E model trained with RL, LaRL and HDSA, showing improvements on the performance evaluated by automatic evaluation metrics and human evaluation. Finally, we demonstrate the semantic meanings of latent dialogue acts to show the explanability for HDNO .", "edit_actions": [{"type": "R", "before": "o gur", "after": "our", "start_char_pos": 671, "end_char_pos": 676, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "ability of explanation", "after": "explanability for HDNO", "start_char_pos": 1635, "end_char_pos": 1657, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}], "sents_char_pos": [0, 190, 347, 487, 667, 934, 1155, 1289, 1552], "domain": "arxiv"}
{"doc_id": "2006.10598", "revision_depth": 1, "before_revision": "We present Shapeshifter Networks (SSNs), a flexible neural network framework that improves performance and reduces memory requirements on a diverse set of scenarios over standard neural networks. Our approach is based on the observation that many neural networks are severely overparameterized, resulting in significant waste in computational resources as well as being susceptible to overfitting. SSNs address this by learning where and how to share parameters between layers in a neural network while avoiding degenerate solutions that result in underfitting. Specifically, we automatically construct parameter groups that identify where parameter sharing is most beneficial. Then, we map each group's weights to construct layerswith learned combinations of candidates from a shared parameter pool. SSNs can share parameters across layers even when they have different sizes , perform different operations , and/or operate on features from different modalities . We evaluate our approach on a diverse set of tasks , including image classification, bidirectional image-sentence retrieval, and phrase grounding, creating high performing models even when using as little as 1\\% of the parameters . We also apply SSNs to knowledge distillation, where we obtain state-of-the-art results when combined with traditional distillation methods .", "after_revision": "Fitting a model into GPU memory during training is an increasing concern as models continue to grow. To address this issue, we present Shapeshifter Networks (SSNs), a flexible neural network framework that decouples layers from model weights, enabling us to implement any neural network with an arbitrary number of parameters. In SSNs each layer obtains weights from a parameter store that decides where and how to allocate parameters to layers. This can result in sharing parameters across layers even when they have different sizes or perform different operations . SSNs do not require any modifications to a model's loss function or architecture, making them easy to use. Our approach can create parameter efficient networks by using a relatively small number of weights, or can improve a model's performance by adding additional model capacity during training without affecting the computational resources required at test time . We evaluate SSNs using seven network architectures across diverse tasks that include image classification, bidirectional image-sentence retrieval, and phrase grounding, creating high performing models even when using as little as 1\\% of the parameters .", "edit_actions": [{"type": "R", "before": "We", "after": "Fitting a model into GPU memory during training is an increasing concern as models continue to grow. To address this issue, we", "start_char_pos": 0, "end_char_pos": 2, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "improves performance and reduces memory requirements on a diverse set of scenarios over standard neural networks. Our approach is based on the observation that many neural networks are severely overparameterized, resulting in significant waste in computational resources as well as being susceptible to overfitting. SSNs address this by learning", "after": "decouples layers from model weights, enabling us to implement any neural network with an arbitrary number of parameters. In SSNs each layer obtains weights from a parameter store that decides", "start_char_pos": 82, "end_char_pos": 427, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "share parameters between layers in a neural network while avoiding degenerate solutions that result in underfitting. Specifically, we automatically construct parameter groups that identify where parameter sharing is most beneficial. Then, we map each group's weights to construct layerswith learned combinations of candidates from a shared parameter pool. SSNs can share", "after": "allocate parameters to layers. This can result in sharing", "start_char_pos": 445, "end_char_pos": 815, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ",", "after": "or", "start_char_pos": 877, "end_char_pos": 878, "major_intent": "fluency", "raw_intents": ["fluency", "meaning-changed", "fluency"]}, {"type": "R", "before": ", and/or operate on features from different modalities", "after": ". SSNs do not require any modifications to a model's loss function or architecture, making them easy to use. Our approach can create parameter efficient networks by using a relatively small number of weights, or can improve a model's performance by adding additional model capacity during training without affecting the computational resources required at test time", "start_char_pos": 908, "end_char_pos": 962, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "our approach on a diverse set of tasks , including", "after": "SSNs using seven network architectures across diverse tasks that include", "start_char_pos": 977, "end_char_pos": 1027, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ". We also apply SSNs to knowledge distillation, where we obtain state-of-the-art results when combined with traditional distillation methods", "after": null, "start_char_pos": 1195, "end_char_pos": 1335, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 195, 397, 561, 677, 800, 964, 1196], "domain": "arxiv"}
{"doc_id": "2006.10598", "revision_depth": 2, "before_revision": "Fitting a model into GPU memory during training is an increasing concern as models continue to grow. To address this issue, we present Shapeshifter Networks (SSNs), a flexible neural network framework that decouples layers from model weights, enabling us to implement any neural network with an arbitrary number of parameters . In SSNseach layer obtains weights from a parameter store that decides where and how to allocate parameters to layers . This can result in sharing parameters across layers even when they have different sizes or perform different operations. SSNs do not require any modifications to a model's loss function or architecture  , making them easy to use. Our approach can create parameter efficient networks by using a relatively small number of weights, or can improve a model's performance by adding additional model capacity during training without affecting the computational resources required at test time. We evaluate SSNs  using seven network architectures across diverse tasks that include image classification, bidirectional image-sentence retrieval, and phrase grounding, creating high performing models even when using as little as 1\\% of the parameters.", "after_revision": "Fitting a model into GPU memory during training is an increasing concern as models continue to grow. Parameter sharing can reduce memory requirements, but existing methods only share parameters between identical layers, limiting their impact. This paper removes these restrictions with a novel task called Neural Parameter Allocation Search (NPAS), where the goal is to generate weights for a network using a given parameter budget. NPAS requires new techniques to morph available parameters to fit any architecture. To address this new task we introduce Shapeshifter Networks (SSNs), which automatically learns where and how to share parameters between all layers in a network, even between layers of varying sizes and operations. SSNs do not require any loss function or architecture modifications , making them easy to use. We evaluate SSNs in key NPAS settings using seven network architectures across diverse tasks including image classification, bidirectional image-sentence retrieval, and phrase grounding, creating high performing models even when using as little as 1\\% of the parameters.", "edit_actions": [{"type": "R", "before": "To address this issue, we present Shapeshifter Networks (SSNs), a flexible neural network framework that decouples layers from model weights, enabling us to implement any neural network with an arbitrary number of parameters . In SSNseach layer obtains weights from a parameter store that decides", "after": "Parameter sharing can reduce memory requirements, but existing methods only share parameters between identical layers, limiting their impact. This paper removes these restrictions with a novel task called Neural Parameter Allocation Search (NPAS), where the goal is to generate weights for a network using a given parameter budget. NPAS requires new techniques to morph available parameters to fit any architecture. To address this new task we introduce Shapeshifter Networks (SSNs), which automatically learns", "start_char_pos": 101, "end_char_pos": 397, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "allocate parameters to layers . This can result in sharing parameters across layers even when they have different sizes or perform different", "after": "share parameters between all layers in a network, even between layers of varying sizes and", "start_char_pos": 415, "end_char_pos": 555, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "modifications to a model's", "after": null, "start_char_pos": 592, "end_char_pos": 618, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "modifications", "start_char_pos": 649, "end_char_pos": 649, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "Our approach can create parameter efficient networks by using a relatively small number of weights, or can improve a model's performance by adding additional model capacity during training without affecting the computational resources required at test time.", "after": null, "start_char_pos": 677, "end_char_pos": 934, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "in key NPAS settings", "start_char_pos": 952, "end_char_pos": 952, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "that include", "after": "including", "start_char_pos": 1008, "end_char_pos": 1020, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}], "sents_char_pos": [0, 100, 327, 446, 567, 676, 934], "domain": "arxiv"}
{"doc_id": "2006.11477", "revision_depth": 1, "before_revision": "We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. We set a new state of the art on both the 100 hour subset of Librispeech as well as on TIMIT phoneme recognition . When lowering the amount of labeled data to one hour, our model outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 5.7 / 10.1 WER on the noisy/clean test sets of Librispeech. This demonstrates the feasibility of speech recognition with limited amounts of labeled data . Fine-tuning on all of Librispeech achieves 1.9/3.5 WER using a simple baseline model architecture. We will release code and models .", "after_revision": "We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/noisy test sets . When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 5.2 / 8.6 WER on the noisy/clean test sets of Librispeech. This demonstrates the feasibility of speech recognition with limited amounts of labeled data .", "edit_actions": [{"type": "R", "before": "We set a new state of the art on both the 100 hour subset of Librispeech as well as on TIMIT phoneme recognition", "after": "Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/noisy test sets", "start_char_pos": 388, "end_char_pos": 500, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "our model", "after": "wav2vec 2.0", "start_char_pos": 557, "end_char_pos": 566, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "5.7", "after": "5.2", "start_char_pos": 775, "end_char_pos": 778, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "10.1", "after": "8.6", "start_char_pos": 781, "end_char_pos": 785, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ". Fine-tuning on all of Librispeech achieves 1.9/3.5 WER using a simple baseline model architecture. We will release code and models", "after": null, "start_char_pos": 928, "end_char_pos": 1060, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 387, 502, 672, 834, 929, 1028], "domain": "arxiv"}
{"doc_id": "2006.11477", "revision_depth": 2, "before_revision": "We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/ noisy test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 5.2 / 8.6 WERon the noisy/clean test sets of Librispeech . This demonstrates the feasibility of speech recognition with limited amounts of labeled data.", "after_revision": "We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/ other test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 4.8 / 8.2 WER . This demonstrates the feasibility of speech recognition with limited amounts of labeled data.", "edit_actions": [{"type": "R", "before": "noisy", "after": "other", "start_char_pos": 472, "end_char_pos": 477, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "5.2", "after": "4.8", "start_char_pos": 763, "end_char_pos": 766, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "8.6 WERon the noisy/clean test sets of Librispeech", "after": "8.2 WER", "start_char_pos": 769, "end_char_pos": 819, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 387, 488, 660, 821], "domain": "arxiv"}
{"doc_id": "2006.15595", "revision_depth": 1, "before_revision": "How to explicitly encode positional information into neural networks is an important problem in natural language processing. In the Transformer model , the positional information is simply encoded as embedding vectors, which are used in the input layer, or encoded as a bias term in the self-attention module. In this work, we investigate the problems in the previous formulations and propose a new positional encoding method for BERT called Transformer with Untied Positional Encoding (TUPE). Different from all other works, TUPE only uses the word embedding as input. In the self-attention module, the word  correlation and positional correlation are computed separately with different parameterizations and then added together. This design removes the noisy word-position correlation and gives more expressiveness to characterize the relationship between words/positions by using different projection matrices. Furthermore, TUPE unties the [CLS] symbol from other positions to provide it with a more specific role to capture the global representation of the sentence. Extensive experiments and ablation studies on GLUE benchmark demonstrate the effectiveness and efficiency of the proposed method: TUPE outperforms several baselines on almost all tasks by a large margin. In particular, it can achieve a higher score than baselines while only using 30\\% pre-training computational costs. We release our code at URL", "after_revision": "How to explicitly encode positional information into neural networks is important in learning the representation of natural languages, such as BERT. Based on the Transformer architecture , the positional information is simply encoded as embedding vectors, which are used in the input layer, or encoded as a bias term in the self-attention module. In this work, we investigate the problems in the previous formulations and propose a new positional encoding method for BERT called Transformer with Untied Positional Encoding (TUPE). Different from all other works, TUPE only uses the word embedding as input. In the self-attention module, the word contextual correlation and positional correlation are computed separately with different parameterizations and then added together. This design removes the addition over heterogeneous embeddings in the input, which may potentially bring randomness, and gives more expressiveness to characterize the relationship between words/positions by using different projection matrices. Furthermore, TUPE unties the [CLS] symbol from other positions to provide it with a more specific role to capture the global representation of the sentence. Extensive experiments and ablation studies on GLUE benchmark demonstrate the effectiveness and efficiency of the proposed method: TUPE outperforms several baselines on almost all tasks by a large margin. In particular, it can achieve a higher score than baselines while only using 30\\% pre-training computational costs. We release our code at URL", "edit_actions": [{"type": "R", "before": "an important problem in natural language processing. In the Transformer model", "after": "important in learning the representation of natural languages, such as BERT. Based on the Transformer architecture", "start_char_pos": 72, "end_char_pos": 149, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "contextual", "start_char_pos": 609, "end_char_pos": 609, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "noisy word-position correlation", "after": "addition over heterogeneous embeddings in the input, which may potentially bring randomness,", "start_char_pos": 755, "end_char_pos": 786, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 124, 309, 493, 569, 730, 913, 1070, 1274, 1390], "domain": "arxiv"}
{"doc_id": "2007.00576", "revision_depth": 1, "before_revision": "To combat COVID-19,  clinicians and scientists all need to digest the vast amount of relevant biomedical knowledge in literature to understand the disease mechanism and the related biological functions. We have developed a novel and comprehensive knowledge discovery framework, COVID-KG, which leverages novel semantic representation and external ontologies to represent text and images in the input literature data, and then performs various extraction components to  extract fine-grained multimedia knowledge elements (entities, relations and events)  . We then exploit the constructed multimedia KGs for question answering and report generation, using drug repurposing as a case study. Our framework also provides detailed contextual sentences, subfigures and knowledge subgraphs as evidence. All of the data, KGs, resources, and shared services are publicly available.", "after_revision": "To combat COVID-19, both clinicians and scientists need to digest the vast amount of relevant biomedical knowledge in literature to understand the disease mechanism and the related biological functions. We have developed a novel and comprehensive knowledge discovery framework, COVID-KG to extract fine-grained multimedia knowledge elements (entities, relations and events) from scientific literature . We then exploit the constructed multimedia knowledge graphs (KGs) for question answering and report generation, using drug repurposing as a case study. Our framework also provides detailed contextual sentences, subfigures and knowledge subgraphs as evidence. All of the data, KGs, reports, resources and shared services are publicly available.", "edit_actions": [{"type": "A", "before": null, "after": "both", "start_char_pos": 20, "end_char_pos": 20, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "all", "after": null, "start_char_pos": 47, "end_char_pos": 50, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "COVID-KG, which leverages novel semantic representation and external ontologies to represent text and images in the input literature data, and then performs various extraction components to", "after": "COVID-KG", "start_char_pos": 278, "end_char_pos": 467, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "to", "start_char_pos": 468, "end_char_pos": 468, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "A", "before": null, "after": "from scientific literature", "start_char_pos": 553, "end_char_pos": 553, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "KGs", "after": "knowledge graphs (KGs)", "start_char_pos": 599, "end_char_pos": 602, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "resources,", "after": "reports, resources", "start_char_pos": 818, "end_char_pos": 828, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 202, 555, 688, 795], "domain": "arxiv"}
{"doc_id": "2007.00576", "revision_depth": 2, "before_revision": "To combat COVID-19, both clinicians and scientists need to digest the vast amount of relevant biomedical knowledge in  literature to understand the disease mechanism and the related biological functions. We have developed a novel and comprehensive knowledge discovery framework, textbf COVID-KG to extract fine-grained multimedia knowledge elements (entities  , relations and events) from scientific literature. We then exploit the constructed multimedia knowledge graphs (KGs) for question answering and report generation, using drug repurposing as a case study. Our framework also provides detailed contextual sentences, subfigures  and knowledge subgraphs as evidence . All of the data, KGs, reports, resources and shared services are publicly available .", "after_revision": "To combat COVID-19, both clinicians and scientists need to digest vast amounts of relevant biomedical knowledge in scientific literature to understand the disease mechanism and related biological functions. We have developed a novel and comprehensive knowledge discovery framework, COVID-KG to extract fine-grained multimedia knowledge elements (entities and their visual chemical structures, relations , and events) from scientific literature. We then exploit the constructed multimedia knowledge graphs (KGs) for question answering and report generation, using drug repurposing as a case study. Our framework also provides detailed contextual sentences, subfigures , and knowledge subgraphs as evidence .", "edit_actions": [{"type": "R", "before": "the vast amount", "after": "vast amounts", "start_char_pos": 66, "end_char_pos": 81, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "scientific", "start_char_pos": 118, "end_char_pos": 118, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 170, "end_char_pos": 173, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "D", "before": "textbf", "after": null, "start_char_pos": 279, "end_char_pos": 285, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "and their visual chemical structures, relations", "start_char_pos": 359, "end_char_pos": 359, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "D", "before": "relations", "after": null, "start_char_pos": 362, "end_char_pos": 371, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 634, "end_char_pos": 634, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "D", "before": ". All of the data, KGs, reports, resources and shared services are publicly available", "after": null, "start_char_pos": 671, "end_char_pos": 756, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 203, 411, 563, 672], "domain": "arxiv"}
{"doc_id": "2007.04508", "revision_depth": 1, "before_revision": "Using the presence or frequency of keywords is a classic approach in the formal analysis of text, but has the drawback of glossing over the relationality of word meanings. Word embedding models overcome this problem by constructing a standardized meaning space where words are assigned a location based on relations of similarity to , and difference from, other words based on how they are used in natural language samples. We show how word embeddings  can be put to the task of interpretation via two kinds of navigation. First, one can hold terms constant and measure how the embedding space moves around them--much like astronomers measured the changing of celestial bodies with the seasons. Second, one can also hold the embedding space constant and see how documents or authors move relative to it--just as ships use the stars on a given night to determine their location. Using the empirical case of immigration discourse in the United States, we demonstrate the merits of these two broad strategies to advance formal approaches to cultural analysis .", "after_revision": "Using the frequency of keywords is a classic approach in the formal analysis of text, but has the drawback of glossing over the relationality of word meanings. Word embedding models overcome this problem by constructing a standardized and continuous \"meaning-space\" where words are assigned a location based on relations of similarity to other words based on how they are used in natural language samples. We show how word embeddings are commensurate with prevailing theories of meaning in sociology and can be put to the task of interpretation via two kinds of navigation. First, one can hold terms constant and measure how the embedding space moves around them -- much like astronomers measured the changing of celestial bodies with the seasons. Second, one can also hold the embedding space constant and see how documents or authors move relative to it -- just as ships use the stars on a given night to determine their location. Using the empirical case of immigration discourse in the United States, we demonstrate the merits of these two broad strategies for advancing important topics in cultural theory, including social marking, media fields, echo chambers, and cultural diffusion and change more broadly .", "edit_actions": [{"type": "D", "before": "presence or", "after": null, "start_char_pos": 10, "end_char_pos": 21, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "meaning space", "after": "and continuous \"meaning-space\"", "start_char_pos": 247, "end_char_pos": 260, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": ", and difference from,", "after": null, "start_char_pos": 333, "end_char_pos": 355, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "are commensurate with prevailing theories of meaning in sociology and", "start_char_pos": 452, "end_char_pos": 452, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "them--much", "after": "them -- much", "start_char_pos": 607, "end_char_pos": 617, "major_intent": "others", "raw_intents": ["clarity", "others", "others"]}, {"type": "R", "before": "it--just", "after": "it -- just", "start_char_pos": 800, "end_char_pos": 808, "major_intent": "others", "raw_intents": ["fluency", "others", "others"]}, {"type": "R", "before": "to advance formal approaches to cultural analysis", "after": "for advancing important topics in cultural theory, including social marking, media fields, echo chambers, and cultural diffusion and change more broadly", "start_char_pos": 1006, "end_char_pos": 1055, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 171, 423, 522, 694, 877]}
{"doc_id": "2007.06225", "revision_depth": 1, "before_revision": "Motivation: NLP continues improving substantially through auto-regressive and auto-encoding Language Models . These LMsrequire expensive computing resources for self-supervised or un-supervised learning from huge unlabelled text corpora. The information learned is transferred through so-called embeddings to downstream prediction tasks. Bioinformatics provide vast gold-mines of structured and sequentially ordered text data leading to extraordinarily successful protein sequence LMs that promise new frontiers for generative and predictive tasks at low inference cost . Here, we addressed two questions: (1) To which extent can HPC up-scale protein LMs to larger databases and larger models? (2) To which extent can LMs extract features from single proteins to get closer to the performance of methods using evolutionary information? Methodology: Here, we trained two auto-regressive language models (Transformer-XL and XLNet) and two auto-encoder models ( BERT and Albert) using 80 billion amino acids from 200 million protein sequences (UniRef100) and 393 billion amino acids  from 2.1 billion protein sequences ( BFD ). The LMs were trained on the Summit supercomputer , using 5616 GPUs  and one TPU Pod , using V3-512 cores. Results: The results of training these LMs on proteins was assessed by predicting secondary structure in three- and eight-states ( Q3= 75-83, Q8= 63-72), localization for 10 cellular compartments (Q10=74) and whether a protein is membrane-bound or water-soluble (Q2=89). Dimensionality reduction revealed that the LM-embeddings from unlabelled data (only protein sequences) captured important biophysical properties of the protein alphabet, namely the amino acids, and their well orchestrated interplay in governing the shapeof proteins. In the analogy of NLP, this implied having learned some of the grammar of the language of life realized in protein sequences. ", "after_revision": "Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs . Here, we trained two auto-regressive language models (Transformer-XL , XLNet) and two auto-encoder models ( Bert, Albert) on data from UniRef and BFD containing up to 393 billion amino acids (words) from 2.1 billion protein sequences ( 22- and 112-times the entire English Wikipedia ). The LMs were trained on the Summit supercomputer at Oak Ridge National Laboratory (ORNL), using 936 nodes (total 5616 GPUs ) and one TPU Pod ( V3-512 or V3-1024). We validated the advantage of up-scaling LMs to larger models supported by bigger data by predicting secondary structure (3-states: Q3= 76-84, 8-states: Q8= 65-73), sub-cellular localization for 10 cellular compartments (Q10=74) and whether a protein is membrane-bound or water-soluble (Q2=89). Dimensionality reduction revealed that the LM-embeddings from unlabeled data (only protein sequences) captured important biophysical properties governing protein shape. This implied learning some of the grammar of the language of life realized in protein sequences. The successful up-scaling of protein LMs through HPC to larger data sets slightly reduced the gap between models trained on evolutionary information and LMs. The official GitHub repository: URL", "edit_actions": [{"type": "R", "before": "Motivation: NLP continues improving substantially through auto-regressive and auto-encoding Language Models . These LMsrequire expensive computing resources for self-supervised or un-supervised learning from huge unlabelled text corpora. The information learned is transferred through so-called embeddings to downstream prediction tasks. Bioinformatics provide vast", "after": "Computational biology and bioinformatics provide vast data", "start_char_pos": 0, "end_char_pos": 365, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "of structured and sequentially ordered text data leading to extraordinarily successful protein sequence LMs that promise new frontiers for generative and predictive tasks", "after": "from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers", "start_char_pos": 377, "end_char_pos": 547, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "cost", "after": "costs", "start_char_pos": 565, "end_char_pos": 569, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "D", "before": "addressed two questions: (1) To which extent can HPC up-scale protein LMs to larger databases and larger models? (2) To which extent can LMs extract features from single proteins to get closer to the performance of methods using evolutionary information? Methodology: Here, we", "after": null, "start_char_pos": 581, "end_char_pos": 857, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "and", "after": ",", "start_char_pos": 918, "end_char_pos": 921, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "coherence"]}, {"type": "R", "before": "BERT and Albert) using 80 billion amino acids from 200 million protein sequences (UniRef100) and", "after": "Bert, Albert) on data from UniRef and BFD containing up to", "start_char_pos": 959, "end_char_pos": 1055, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "(words)", "start_char_pos": 1080, "end_char_pos": 1080, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "style"]}, {"type": "R", "before": "BFD", "after": "22- and 112-times the entire English Wikipedia", "start_char_pos": 1118, "end_char_pos": 1121, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", using", "after": "at Oak Ridge National Laboratory (ORNL), using 936 nodes (total", "start_char_pos": 1174, "end_char_pos": 1181, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ")", "start_char_pos": 1192, "end_char_pos": 1192, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", using", "after": "(", "start_char_pos": 1209, "end_char_pos": 1216, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "cores. Results: The results of training these LMs on proteins was assessed by", "after": "or V3-1024). We validated the advantage of up-scaling LMs to larger models supported by bigger data by", "start_char_pos": 1224, "end_char_pos": 1301, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "in three- and eight-states (", "after": "(3-states:", "start_char_pos": 1333, "end_char_pos": 1361, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "75-83,", "after": "76-84, 8-states:", "start_char_pos": 1366, "end_char_pos": 1372, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "63-72),", "after": "65-73), sub-cellular", "start_char_pos": 1377, "end_char_pos": 1384, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "unlabelled", "after": "unlabeled", "start_char_pos": 1564, "end_char_pos": 1574, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "of the protein alphabet, namely the amino acids, and their well orchestrated interplay in governing the shapeof proteins. In the analogy of NLP, this implied having learned", "after": "governing protein shape. This implied learning", "start_char_pos": 1647, "end_char_pos": 1819, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "The successful up-scaling of protein LMs through HPC to larger data sets slightly reduced the gap between models trained on evolutionary information and LMs. The official GitHub repository: URL", "start_char_pos": 1895, "end_char_pos": 1895, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 109, 237, 337, 571, 693, 835, 1124, 1230, 1501, 1768], "domain": "arxiv"}
{"doc_id": "2008.01766", "revision_depth": 1, "before_revision": "Machines show an increasingly broad set of linguistic competencies, thanks to recent progress in Natural Language Processing (NLP). Many algorithms stem from past computational work in psychology, raising the question of whether they understand words as people do. In this paper, we compare how humans and machines represent the meaning of words. We argue that contemporary NLP systems are promising models of human word similarity, but they fall short in many other respects. Current models are too strongly linked to the text-based patterns in large corpora, and too weakly linked to the desires, goals, and beliefs that people use words in order to express . Word meanings must also be grounded in vision and action, and capable of flexible combinations , in ways that current systems are not. We pose concrete challenges for developing machines with a more human-like, conceptual basis for word meaning. We also discuss implications for cognitive science and NLP.", "after_revision": "Machines show an increasingly broad set of linguistic competencies, thanks to recent progress in Natural Language Processing (NLP). Many algorithms stem from past computational work in psychology, raising the question of whether they understand words as people do. In this paper, we compare how humans and machines represent the meaning of words. We argue that contemporary NLP systems are promising models of human word similarity, but they fall short in many other respects. Current models are too strongly linked to the text-based patterns in large corpora, and too weakly linked to the desires, goals, and beliefs that people express through words . Word meanings must also be grounded in vision and action, and capable of flexible combinations in ways that current systems are not. We pose concrete challenges for developing machines with a more human-like, conceptual basis for word meaning. We also discuss implications for cognitive science and NLP.", "edit_actions": [{"type": "R", "before": "use words in order to express", "after": "express through words", "start_char_pos": 630, "end_char_pos": 659, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 757, "end_char_pos": 758, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}], "sents_char_pos": [0, 131, 264, 346, 476, 661, 796, 907], "domain": "arxiv"}
{"doc_id": "2008.01766", "revision_depth": 2, "before_revision": "Machines show an increasingly broad set of linguistic competencies, thanks to recent progress in Natural Language Processing (NLP). Many algorithms stem from past computational work in psychology, raising the question of whether they understand words as people do . In this paper , we compare how humans and machines represent the meaning of words. We argue that contemporary NLP systems are promising models of human word similarity, but they fall short in many other respects. Current models are too strongly linked to the text-based patterns in large corpora, and too weakly linked to the desires, goals, and beliefs that people express through words. Word meanings must also be grounded in vision and action , and capable of flexible combinations in ways that current systems are not. We pose concrete challenges for developing machines with a more human-like, conceptual basis for word meaning . We also discuss implications for cognitive science and NLP .", "after_revision": "Machines have achieved a broad and growing set of linguistic competencies, thanks to recent progress in Natural Language Processing (NLP). Psychologists have shown increasing interest in such models, comparing their output to psychological judgments such as similarity, association, priming, and comprehension, raising the question of whether the models could serve as psychological theories . In this article , we compare how humans and machines represent the meaning of words. We argue that contemporary NLP systems are fairly successful models of human word similarity, but they fall short in many other respects. Current models are too strongly linked to the text-based patterns in large corpora, and too weakly linked to the desires, goals, and beliefs that people express through words. Word meanings must also be grounded in perception and action and be capable of flexible combinations in ways that current systems are not. We discuss more promising approaches to grounding NLP systems and argue that they will be more successful with a more human-like, conceptual basis for word meaning .", "edit_actions": [{"type": "R", "before": "show an increasingly broad", "after": "have achieved a broad and growing", "start_char_pos": 9, "end_char_pos": 35, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "Many algorithms stem from past computational work in psychology,", "after": "Psychologists have shown increasing interest in such models, comparing their output to psychological judgments such as similarity, association, priming, and comprehension,", "start_char_pos": 132, "end_char_pos": 196, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "they understand words as people do", "after": "the models could serve as psychological theories", "start_char_pos": 229, "end_char_pos": 263, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "paper", "after": "article", "start_char_pos": 274, "end_char_pos": 279, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "promising", "after": "fairly successful", "start_char_pos": 392, "end_char_pos": 401, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "vision and action , and", "after": "perception and action and be", "start_char_pos": 694, "end_char_pos": 717, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "pose concrete challenges for developing machines", "after": "discuss more promising approaches to grounding NLP systems and argue that they will be more successful", "start_char_pos": 792, "end_char_pos": 840, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "D", "before": ". We also discuss implications for cognitive science and NLP", "after": null, "start_char_pos": 899, "end_char_pos": 959, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 131, 265, 348, 478, 654, 788, 900], "domain": "arxiv"}
{"doc_id": "2008.07905", "revision_depth": 1, "before_revision": "Non-autoregressive neural machine translation achieves remarkable inference acceleration compared to autoregressive models. However, current non-autoregressive models still fall behind their autoregressive counterparts in prediction accuracy. We attribute the accuracy gaps to two disadvantages of non-autoregressive models : a) learning simultaneous generation under the overly strong conditional independence assumption; b) lacking explicit target language modeling. In this paper , we propose Glancing Transformer (GLAT) to address the above disadvantages, which reduces the difficulty of learning simultaneous generation and introduces explicit target language modeling in the non-autoregressive setting at the same time . Experiments on several benchmarks demonstrate that our approach significantly improves the accuracy of non-autoregressive models without sacrificing any inference efficiency . In particular, GLAT achieves 30.91 BLEU on WMT 2014 German-English, which narrows the gap between autoregressive models and non-autoregressive models to less than 0.5 BLEU score .", "after_revision": "Although non-autoregressive models with one-iteration generation achieve remarkable inference speed-up, they still fall behind their autoregressive counterparts in prediction accuracy. The non-autoregressive models with the best accuracy currently rely on multiple decoding iterations, which largely sacrifice the inference speed of non-autoregressive models . Inspired by the way of learning word dependencies in autoregressive and iterative-decoding models , we propose Glancing Transformer (GLAT) with a glancing language model (GLM), which learns to capture the word dependency gradually . Experiments on three benchmarks demonstrate that our approach can significantly improve the accuracy of non-autoregressive models without multiple decoding iterations . In particular, GLAT achieves state-of-the-art results among non-iterative models and even outperforms top iterative counterparts in some specific benchmarks .", "edit_actions": [{"type": "R", "before": "Non-autoregressive neural machine translation achieves remarkable inference acceleration compared to autoregressive models. However, current", "after": "Although", "start_char_pos": 0, "end_char_pos": 140, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "models", "after": "models with one-iteration generation achieve remarkable inference speed-up, they", "start_char_pos": 160, "end_char_pos": 166, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "We attribute the accuracy gaps to two disadvantages", "after": "The non-autoregressive models with the best accuracy currently rely on multiple decoding iterations, which largely sacrifice the inference speed", "start_char_pos": 243, "end_char_pos": 294, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": ": a) learning simultaneous generation under the overly strong conditional independence assumption; b) lacking explicit target language modeling. In this paper", "after": ". Inspired by the way of learning word dependencies in autoregressive and iterative-decoding models", "start_char_pos": 324, "end_char_pos": 482, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "to address the above disadvantages, which reduces the difficulty of learning simultaneous generation and introduces explicit target language modeling in the non-autoregressive setting at the same time", "after": "with a glancing language model (GLM), which learns to capture the word dependency gradually", "start_char_pos": 524, "end_char_pos": 724, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "several", "after": "three", "start_char_pos": 742, "end_char_pos": 749, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "significantly improves", "after": "can significantly improve", "start_char_pos": 791, "end_char_pos": 813, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "sacrificing any inference efficiency", "after": "multiple decoding iterations", "start_char_pos": 864, "end_char_pos": 900, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "30.91 BLEU on WMT 2014 German-English, which narrows the gap between autoregressive models and non-autoregressive models to less than 0.5 BLEU score", "after": "state-of-the-art results among non-iterative models and even outperforms top iterative counterparts in some specific benchmarks", "start_char_pos": 932, "end_char_pos": 1080, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 123, 242, 422, 468, 902], "domain": "arxiv"}
{"doc_id": "2008.07905", "revision_depth": 2, "before_revision": "Although non-autoregressive models with one-iteration generation achieve remarkable inference speed-up, they still fall behind their autoregressive counterparts in prediction accuracy. The non-autoregressive models with the best accuracy currently rely on multiple decoding iterations, which largely sacrifice the inference speed of non-autoregressive models. Inspired by the way of learning word dependencies in autoregressive and iterative-decoding models, we propose Glancing Transformer (GLAT) with a glancing language model (GLM), which learns to capture the word dependency gradually . Experiments on three benchmarks demonstrate that our approach can significantly improve the accuracy of non-autoregressive models without multiple decoding iterations. In particular, GLAT achieves state-of-the-art results among non-iterative models and even outperforms top iterative counterparts in some specific benchmarks .", "after_revision": "Recent work on non-autoregressive neural machine translation (NAT) aims at improving the efficiency by parallel decoding without sacrificing the quality. However, existing NAT methods are either inferior to Transformer or require multiple decoding passes, leading to reduced speedup. We propose the Glancing Language Model (GLM), a method to learn word interdependency for single-pass parallel generation models. With GLM, we develop Glancing Transformer (GLAT) for machine translation. With only single-pass parallel decoding, GLAT is able to generate high-quality translation with 8-15 times speedup . Experiments on multiple WMT language directions show that GLAT outperforms all previous single pass non-autoregressive methods, and is nearly comparable to Transformer, reducing the gap to 0.25-0.9 BLEU points .", "edit_actions": [{"type": "R", "before": "Although", "after": "Recent work on", "start_char_pos": 0, "end_char_pos": 8, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "models with one-iteration generation achieve remarkable inference speed-up, they still fall behind their autoregressive counterparts in prediction accuracy. The non-autoregressive models with the best accuracy currently rely on multiple decoding iterations, which largely sacrifice the inference speed of non-autoregressive models. Inspired by the way of learning word dependencies in autoregressive and iterative-decoding models, we propose", "after": "neural machine translation (NAT) aims at improving the efficiency by parallel decoding without sacrificing the quality. However, existing NAT methods are either inferior to Transformer or require multiple decoding passes, leading to reduced speedup. We propose the Glancing Language Model (GLM), a method to learn word interdependency for single-pass parallel generation models. With GLM, we develop", "start_char_pos": 28, "end_char_pos": 469, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "with a glancing language model (GLM), which learns to capture the word dependency gradually", "after": "for machine translation. With only single-pass parallel decoding, GLAT is able to generate high-quality translation with 8-15 times speedup", "start_char_pos": 498, "end_char_pos": 589, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "three benchmarks demonstrate that our approach can significantly improve the accuracy of", "after": "multiple WMT language directions show that GLAT outperforms all previous single pass", "start_char_pos": 607, "end_char_pos": 695, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "models without multiple decoding iterations. In particular, GLAT achieves state-of-the-art results among non-iterative models and even outperforms top iterative counterparts in some specific benchmarks", "after": "methods, and is nearly comparable to Transformer, reducing the gap to 0.25-0.9 BLEU points", "start_char_pos": 715, "end_char_pos": 916, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 184, 359, 759], "domain": "arxiv"}
{"doc_id": "2008.11015", "revision_depth": 1, "before_revision": "It is common for people to create different types of charts to explore a multi-dimensional dataset (table). However, to build an intelligent assistant that recommends commonly composed charts, the fundamental problems of \"multi-dialect\" unification , imbalanced data and open vocabulary exist . In this paper, we propose Table2Charts framework which learns common patterns from a large corpus of (table, charts) pairs. Based on deep Q-learning with copying mechanism and heuristic searching, Table2Charts does table-to-sequence generation, where each sequence follows a chart template. On a large spreadsheet corpus with 196k tables and 306k charts, we show that Table2Charts could learn a shared representation of table fields so that tasks on different chart types could mutually enhance each other. Table2Charts has >0.61 recall at top-3 and >0.49 recall at top-1 for both single-type and multi-type chart recommendation tasks .", "after_revision": "It is common for people to create different types of charts to explore a multi-dimensional dataset (table). However, to build a real-world intelligent assistant that recommends commonly composed charts, it should take the challenges of efficiency , imbalanced data hungry and table context into consideration . In this paper, we propose Table2Charts framework which learns common patterns from a large corpus of (table, charts) pairs. Based on deep Q-learning with copying mechanism and heuristic searching, Table2Charts does table-to-sequence generation, where each sequence follows a chart template. On a large spreadsheet corpus with 167k tables and 271k charts, we show that Table2Charts could learn a shared representation of table fields so that tasks on different chart types could mutually enhance each other. Table2Charts outperforms other chart recommendation systems in both multi-type task (with almost doubled recall numbers R@3=0.62 and R@1=0.44) and human evaluations .", "edit_actions": [{"type": "R", "before": "an", "after": "a real-world", "start_char_pos": 126, "end_char_pos": 128, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "the fundamental problems of \"multi-dialect\" unification", "after": "it should take the challenges of efficiency", "start_char_pos": 193, "end_char_pos": 248, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "and open vocabulary exist", "after": "hungry and table context into consideration", "start_char_pos": 267, "end_char_pos": 292, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "196k tables and 306k", "after": "167k tables and 271k", "start_char_pos": 621, "end_char_pos": 641, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "has >0.61 recall at top-3 and >0.49 recall at top-1 for both single-type and", "after": "outperforms other chart recommendation systems in both", "start_char_pos": 815, "end_char_pos": 891, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "chart recommendation tasks", "after": "task (with almost doubled recall numbers R@3=0.62 and R@1=0.44) and human evaluations", "start_char_pos": 903, "end_char_pos": 929, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 107, 294, 418, 585, 801], "domain": "arxiv"}
{"doc_id": "2008.11015", "revision_depth": 2, "before_revision": "It is common for people to create different types of charts to explore a multi-dimensional dataset (table). However, to build a real-world intelligent assistant that recommends commonly composed charts , it should take the challenges of efficiency, imbalanced data hungry and table context into consideration. In this paper, we propose Table2Charts framework which learns common patterns from a large corpus of (table, charts) pairs. Based on deep Q-learning with copying mechanism and heuristic searching, Table2Charts does table-to-sequence generation, where each sequence follows a chart template. On a large spreadsheet corpus with 167k tables and 271k charts, we show that Table2Charts could learn a shared representation of table fields so that  tasks on different chart types could mutually enhance each other. Table2Charts outperforms other chart recommendation systems in both multi-type task (with almost doubled recall numbers R@3= 0.62 and R@1= 0.44 ) and human evaluations.", "after_revision": "It is common for people to create different types of charts to explore a multi-dimensional dataset (table). However, to recommend commonly composed charts in real world, one should take the challenges of efficiency, imbalanced data and table context into consideration. In this paper, we propose Table2Charts framework which learns common patterns from a large corpus of (table, charts) pairs. Based on deep Q-learning with copying mechanism and heuristic searching, Table2Charts does table-to-sequence generation, where each sequence follows a chart template. On a large spreadsheet corpus with 165k tables and 266k charts, we show that Table2Charts could learn a shared representation of table fields so that recommendation tasks on different chart types could mutually enhance each other. Table2Charts outperforms other chart recommendation systems in both multi-type task (with doubled recall numbers R@3= 0.61 and R@1= 0.43 ) and human evaluations.", "edit_actions": [{"type": "R", "before": "build a real-world intelligent assistant that recommends", "after": "recommend", "start_char_pos": 120, "end_char_pos": 176, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", it", "after": "in real world, one", "start_char_pos": 202, "end_char_pos": 206, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "hungry", "after": null, "start_char_pos": 265, "end_char_pos": 271, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "167k tables and 271k", "after": "165k tables and 266k", "start_char_pos": 636, "end_char_pos": 656, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "recommendation", "start_char_pos": 751, "end_char_pos": 751, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "almost", "after": null, "start_char_pos": 908, "end_char_pos": 914, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "0.62", "after": "0.61", "start_char_pos": 943, "end_char_pos": 947, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "0.44", "after": "0.43", "start_char_pos": 957, "end_char_pos": 961, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 107, 309, 433, 600, 817], "domain": "arxiv"}
{"doc_id": "2008.11608", "revision_depth": 1, "before_revision": "Transformer-based language models have taken many fields in NLP by storm. BERT and its derivatives dominate most of the existing evaluation benchmarks, including those for Word Sense Disambiguation (WSD), thanks to their ability in capturing context-sensitive semantic nuances. However, there is still little knowledge about their capabilities and potential limitations for encoding and recovering word senses. In this article, we provide an in-depth quantitative and qualitative analysis of the celebrated BERT model with respect to lexical ambiguity. One of the main conclusions of our analysis is that BERT performs a decent job in capturing high-level sense distinctions  , even when a limited number of examples is available for each word sense. Our analysis also reveals that in some cases language models come close to solving coarse-grained noun disambiguation under ideal conditions in terms of availability of training data and computing resources. However, this scenario rarely occurs in real-world settings and, hence, many practical challenges remain even in the coarse-grained setting. We also perform an in-depth comparison of the two main language model based WSD strategies, i.e., fine-tuning and feature extraction, finding that the latter approach is more robust with respect to sense bias and it can better exploit limited available training data  .", "after_revision": "Transformer-based language models have taken many fields in NLP by storm. BERT and its derivatives dominate most of the existing evaluation benchmarks, including those for Word Sense Disambiguation (WSD), thanks to their ability in capturing context-sensitive semantic nuances. However, there is still little knowledge about their capabilities and potential limitations for encoding and recovering word senses. In this article, we provide an in-depth quantitative and qualitative analysis of the celebrated BERT model with respect to lexical ambiguity. One of the main conclusions of our analysis is that BERT captures high-level sense distinctions accurately , even when a limited number of examples is available for each word sense. Our analysis also reveals that in some cases language models come close to solving coarse-grained noun disambiguation under ideal conditions in terms of availability of training data and computing resources. However, this scenario rarely occurs in real-world settings and, hence, many practical challenges remain even in the coarse-grained setting. We also perform an in-depth comparison of the two main language model based WSD strategies, i.e., fine-tuning and feature extraction, finding that the latter approach is more robust with respect to sense bias and it can better exploit limited available training data . In fact, a simple feature extraction strategy based on the averaging of contextualized embeddings proves robust even using only three training sentences per word sense, with minimal improvements beyond this small number of examples .", "edit_actions": [{"type": "R", "before": "performs a decent job in capturing", "after": "captures", "start_char_pos": 610, "end_char_pos": 644, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "accurately", "start_char_pos": 675, "end_char_pos": 675, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ". In fact, a simple feature extraction strategy based on the averaging of contextualized embeddings proves robust even using only three training sentences per word sense, with minimal improvements beyond this small number of examples", "start_char_pos": 1367, "end_char_pos": 1367, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 73, 277, 410, 552, 750, 958, 1099], "domain": "arxiv"}
{"doc_id": "2008.11608", "revision_depth": 2, "before_revision": "Transformer-based language models have taken many fields in NLP by storm. BERT and its derivatives dominate most of the existing evaluation benchmarks, including those for Word Sense Disambiguation (WSD), thanks to their ability in capturing context-sensitive semantic nuances. However, there is still little knowledge about their capabilities and potential limitations for encoding and recovering word senses. In this article, we provide an in-depth quantitative and qualitative analysis of the celebrated BERT model with respect to lexical ambiguity. One of the main conclusions of our analysis is that BERT captures high-level sense distinctions accurately , even when a limited number of examples is available for each word sense. Our analysis also reveals that in some cases language models come close to solving coarse-grained noun disambiguation under ideal conditions in terms of availability of training data and computing resources. However, this scenario rarely occurs in real-world settings and, hence, many practical challenges remain even in the coarse-grained setting. We also perform an in-depth comparison of the two main language model based WSD strategies, i.e., fine-tuning and feature extraction, finding that the latter approach is more robust with respect to sense bias and it can better exploit limited available training data. In fact, a simple feature extraction strategy based on the averaging of contextualized embeddings proves robust even using only three training sentences per word sense, with minimal improvements beyond this small number of examples .", "after_revision": "Transformer-based language models have taken many fields in NLP by storm. BERT and its derivatives dominate most of the existing evaluation benchmarks, including those for Word Sense Disambiguation (WSD), thanks to their ability in capturing context-sensitive semantic nuances. However, there is still little knowledge about their capabilities and potential limitations in encoding and recovering word senses. In this article, we provide an in-depth quantitative and qualitative analysis of the celebrated BERT model with respect to lexical ambiguity. One of the main conclusions of our analysis is that BERT can accurately capture high-level sense distinctions , even when a limited number of examples is available for each word sense. Our analysis also reveals that in some cases language models come close to solving coarse-grained noun disambiguation under ideal conditions in terms of availability of training data and computing resources. However, this scenario rarely occurs in real-world settings and, hence, many practical challenges remain even in the coarse-grained setting. We also perform an in-depth comparison of the two main language model based WSD strategies, i.e., fine-tuning and feature extraction, finding that the latter approach is more robust with respect to sense bias and it can better exploit limited available training data. In fact, the simple feature extraction strategy of averaging contextualized embeddings proves robust even using only three training sentences per word sense, with minimal improvements obtained by increasing the size of this training data .", "edit_actions": [{"type": "R", "before": "for", "after": "in", "start_char_pos": 370, "end_char_pos": 373, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "captures", "after": "can accurately capture", "start_char_pos": 610, "end_char_pos": 618, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "accurately", "after": null, "start_char_pos": 649, "end_char_pos": 659, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "a", "after": "the", "start_char_pos": 1361, "end_char_pos": 1362, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "based on the averaging of", "after": "of averaging", "start_char_pos": 1398, "end_char_pos": 1423, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "beyond this small number of examples", "after": "obtained by increasing the size of this training data", "start_char_pos": 1547, "end_char_pos": 1583, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 73, 277, 410, 552, 734, 942, 1083, 1351], "domain": "arxiv"}
{"doc_id": "2009.03996", "revision_depth": 1, "before_revision": "Our goal is to construct mathematical operations that combine non-determinism measured from quantum randomness with computational determinism so that non-mechanistic behavior is preserved in the computation. Formally, some results about operations applied to computably enumerable (c.e.) and bi-immune sets are proven here, where the objective is for the operations to preserve bi-immunity. While developing rearrangement operations on the natural numbers, we discovered that the bi-immune rearrangements generate an uncountable subgroup of the  symmetric group on the natural numbers. The structure of this new subgroup  is unknown.", "after_revision": "Our goal is to construct mathematical operations that combine indeterminism measured from quantum randomness with computational determinism so that non-mechanistic behavior is preserved in the computation. Formally, some results about operations applied to computably enumerable (c.e.) and bi-immune sets are proven here, where the operations preserve bi-immunity. While developing rearrangement operations on the natural numbers, we discovered that the bi-immune rearrangements generate an uncountable subgroup of the infinite symmetric group on the natural numbers. We show that this new subgroup contains the bounded symmetric group on the natural numbers, and consequently is highly transitive. The complete structure of this new subgroup and its subgroups generated by one or more bi-immune rearrangements is unknown.", "edit_actions": [{"type": "R", "before": "non-determinism", "after": "indeterminism", "start_char_pos": 62, "end_char_pos": 77, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "objective is for the operations to", "after": "operations", "start_char_pos": 334, "end_char_pos": 368, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "infinite", "start_char_pos": 545, "end_char_pos": 545, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "The", "after": "We show that this new subgroup contains the bounded symmetric group on the natural numbers, and consequently is highly transitive. The complete", "start_char_pos": 586, "end_char_pos": 589, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "and its subgroups generated by one or more bi-immune rearrangements", "start_char_pos": 621, "end_char_pos": 621, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 207, 390, 585], "domain": "arxiv"}
{"doc_id": "2009.03996", "revision_depth": 2, "before_revision": "Our goal is to construct mathematical operations that combine indeterminism measured from quantum randomness with computational determinism so that non-mechanistic behavior is preserved in the computation. Formally, some results about operations applied to computably enumerable (c.e.) and bi-immune sets are proven here, where the operations preserve bi-immunity. While developing rearrangement operations on the natural numbers, we discovered that the bi-immune rearrangements generate an uncountable subgroup of the infinite symmetric group  } on the natural numbers . } We show that this new subgroup contains the bounded symmetric group on the natural numbers, and consequently is highly transitive.  } The complete structure of this new subgroup and its subgroups generated by one or more bi-immune rearrangements is unknown.", "after_revision": "Our goal is to construct mathematical operations that combine indeterminism measured from quantum randomness with computational determinism so that non-mechanistic behavior is preserved in the computation. Formally, some results about operations applied to computably enumerable (c.e.) and bi-immune sets are proven here, where the objective is for the operations to preserve bi-immunity. While developing rearrangement operations on the natural numbers, we discovered that the bi-immune rearrangements generate an uncountable subgroup of the infinite symmetric group (Sym(mathbb N})) on the natural numbers mathbb N}. This new uncountable subgroup is called the bi-immune symmetric group. We show that the bi-immune symmetric group contains the finitary symmetric group on the natural numbers, and consequently is highly transitive. Furthermore, the bi-immune symmetric group is dense in Sym(mathbb N}) with respect to the pointwise convergence topology. The complete structure of the bi-immune symmetric group and its subgroups generated by one or more bi-immune rearrangements is unknown.", "edit_actions": [{"type": "R", "before": "operations", "after": "objective is for the operations to", "start_char_pos": 332, "end_char_pos": 342, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "(Sym(mathbb", "start_char_pos": 544, "end_char_pos": 544, "major_intent": "others", "raw_intents": ["clarity", "others", "others"]}, {"type": "A", "before": null, "after": "N", "start_char_pos": 545, "end_char_pos": 545, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "A", "before": null, "after": "))", "start_char_pos": 546, "end_char_pos": 546, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": ".", "after": "mathbb", "start_char_pos": 570, "end_char_pos": 571, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "A", "before": null, "after": "N", "start_char_pos": 572, "end_char_pos": 572, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "A", "before": null, "after": ". This new uncountable subgroup is called the bi-immune symmetric group.", "start_char_pos": 573, "end_char_pos": 573, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "this new subgroup contains the bounded", "after": "the bi-immune symmetric group contains the finitary", "start_char_pos": 587, "end_char_pos": 625, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "Furthermore, the bi-immune symmetric group is dense in Sym(mathbb", "start_char_pos": 705, "end_char_pos": 705, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "N", "start_char_pos": 706, "end_char_pos": 706, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "A", "before": null, "after": ") with respect to the pointwise convergence topology.", "start_char_pos": 707, "end_char_pos": 707, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "this new subgroup", "after": "the bi-immune symmetric group", "start_char_pos": 734, "end_char_pos": 751, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 205, 364, 704]}
{"doc_id": "2009.05166", "revision_depth": 1, "before_revision": "Large-scale cross-lingual language models (LM), such as mBERT, Unicoder and XLM, have achieved great success in cross-lingual representation learning. However, when applied to zero-shot cross-lingual transfer tasks, most existing methods use only single-language input for LM finetuning, without leveraging the intrinsic cross-lingual alignment between different languages that is essential for multilingual tasks. In this paper, we propose FILTER, an enhanced fusion method that takes cross-lingual data as input for XLM finetuning. Specifically, FILTER first encodes text input in the source language and its translation in the target language independently in the shallow layers, then performs cross-lingual fusion to extract multilingual knowledge in the intermediate layers, and finally performs further language-specific encoding. During inference, the model makes predictions based on the text input in the target language and its translation in the source language. For simple tasks such as classification, translated text in the target language shares the same label as the source language. However, this shared label becomes less accurate or even unavailable for more complex tasks such as question answering, NER and POS tagging. For better model scalability, we further propose an additional KL-divergence self-teaching loss for model training, based on auto-generated soft pseudo-labels for translated text in the target language. Extensive experiments demonstrate that FILTER achieves new state of the art (77.0 on average) on the challenging multilingual multi-task benchmark, XTREME .", "after_revision": "Large-scale cross-lingual language models (LM), such as mBERT, Unicoder and XLM, have achieved great success in cross-lingual representation learning. However, when applied to zero-shot cross-lingual transfer tasks, most existing methods use only single-language input for LM finetuning, without leveraging the intrinsic cross-lingual alignment between different languages that is essential for multilingual tasks. In this paper, we propose FILTER, an enhanced fusion method that takes cross-lingual data as input for XLM finetuning. Specifically, FILTER first encodes text input in the source language and its translation in the target language independently in the shallow layers, then performs cross-lingual fusion to extract multilingual knowledge in the intermediate layers, and finally performs further language-specific encoding. During inference, the model makes predictions based on the text input in the target language and its translation in the source language. For simple tasks such as classification, translated text in the target language shares the same label as the source language. However, this shared label becomes less accurate or even unavailable for more complex tasks such as question answering, NER and POS tagging. For better model scalability, we further propose an additional KL-divergence self-teaching loss for model training, based on auto-generated soft pseudo-labels for translated text in the target language. Extensive experiments demonstrate that FILTER achieves new state of the art on two challenging multilingual multi-task benchmarks, XTREME and XGLUE .", "edit_actions": [{"type": "R", "before": "(77.0 on average) on the", "after": "on two", "start_char_pos": 1520, "end_char_pos": 1544, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "benchmark, XTREME", "after": "benchmarks, XTREME and XGLUE", "start_char_pos": 1581, "end_char_pos": 1598, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 150, 414, 533, 836, 973, 1099, 1240, 1443], "domain": "arxiv"}
{"doc_id": "2009.05166", "revision_depth": 2, "before_revision": "Large-scale cross-lingual language models (LM), such as mBERT, Unicoder and XLM, have achieved great success in cross-lingual representation learning. However, when applied to zero-shot cross-lingual transfer tasks, most existing methods use only single-language input for LM finetuning, without leveraging the intrinsic cross-lingual alignment between different languages that is essential for multilingual tasks. In this paper, we propose FILTER, an enhanced fusion method that takes cross-lingual data as input for XLM finetuning. Specifically, FILTER first encodes text input in the source language and its translation in the target language independently in the shallow layers, then performs cross-lingual fusion to extract multilingual knowledge in the intermediate layers, and finally performs further language-specific encoding. During inference, the model makes predictions based on the text input in the target language and its translation in the source language. For simple tasks such as classification, translated text in the target language shares the same label as the source language. However, this shared label becomes less accurate or even unavailable for more complex tasks such as question answering, NER and POS tagging. For better model scalability , we further propose an additional KL-divergence self-teaching loss for model training, based on auto-generated soft pseudo-labels for translated text in the target language. Extensive experiments demonstrate that FILTER achieves new state of the art on two challenging multilingual multi-task benchmarks, XTREME and XGLUE.", "after_revision": "Large-scale cross-lingual language models (LM), such as mBERT, Unicoder and XLM, have achieved great success in cross-lingual representation learning. However, when applied to zero-shot cross-lingual transfer tasks, most existing methods use only single-language input for LM finetuning, without leveraging the intrinsic cross-lingual alignment between different languages that proves essential for multilingual tasks. In this paper, we propose FILTER, an enhanced fusion method that takes cross-lingual data as input for XLM finetuning. Specifically, FILTER first encodes text input in the source language and its translation in the target language independently in the shallow layers, then performs cross-language fusion to extract multilingual knowledge in the intermediate layers, and finally performs further language-specific encoding. During inference, the model makes predictions based on the text input in the target language and its translation in the source language. For simple tasks such as classification, translated text in the target language shares the same label as the source language. However, this shared label becomes less accurate or even unavailable for more complex tasks such as question answering, NER and POS tagging. To tackle this issue , we further propose an additional KL-divergence self-teaching loss for model training, based on auto-generated soft pseudo-labels for translated text in the target language. Extensive experiments demonstrate that FILTER achieves new state of the art on two challenging multilingual multi-task benchmarks, XTREME and XGLUE.", "edit_actions": [{"type": "R", "before": "is", "after": "proves", "start_char_pos": 378, "end_char_pos": 380, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "cross-lingual", "after": "cross-language", "start_char_pos": 697, "end_char_pos": 710, "major_intent": "fluency", "raw_intents": ["fluency", "style", "clarity"]}, {"type": "R", "before": "For better model scalability", "after": "To tackle this issue", "start_char_pos": 1241, "end_char_pos": 1269, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 150, 414, 533, 836, 973, 1099, 1240, 1444], "domain": "arxiv"}
{"doc_id": "2009.05169", "revision_depth": 1, "before_revision": "We propose a novel method to sparsify attention in the Transformer model by learning to select the most-informative token representations , thus leveraging the model's information bottleneck with twofold strength. A careful analysis shows that the contextualization of encoded representations in our model is significantly more effective than in the original Transformer. We achieve a notable reduction in memory usage due to an improved differentiable top-k operator , making the model suitable to process long documents, as shown on an example of a summarization task .", "after_revision": "We propose a novel method to sparsify attention in the Transformer model by learning to select the most-informative token representations during the training process, thus focusing on task-specific parts of the input. A reduction of quadratic time and memory complexity to sublinear was achieved due to a robust differentiable top-k operator . For example, our experiments on a challenging summarization task of long documents show that our method is much faster and up to 16 times more memory efficient while significantly outperforming both dense and state-of-the-art sparse transformer models. The method can be effortlessly applied to many models used in NLP and CV, simultaneously with other improvements since representation pooling addresses a different aspect of the attention's complexity problem .", "edit_actions": [{"type": "R", "before": ", thus leveraging the model's information bottleneck with twofold strength. A careful analysis shows that the contextualization of encoded representations in our model is significantly more effective than in the original Transformer. We achieve a notable reduction in memory usage due to an improved", "after": "during the training process, thus focusing on task-specific parts of the input. A reduction of quadratic time and memory complexity to sublinear was achieved due to a robust", "start_char_pos": 138, "end_char_pos": 437, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ", making the model suitable to process long documents, as shown on an example of a summarization task", "after": ". For example, our experiments on a challenging summarization task of long documents show that our method is much faster and up to 16 times more memory efficient while significantly outperforming both dense and state-of-the-art sparse transformer models. The method can be effortlessly applied to many models used in NLP and CV, simultaneously with other improvements since representation pooling addresses a different aspect of the attention's complexity problem", "start_char_pos": 468, "end_char_pos": 569, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}], "sents_char_pos": [0, 213, 371], "domain": "arxiv"}
{"doc_id": "2009.05426", "revision_depth": 1, "before_revision": "The second edition of \"Semantic Relations Between Nominals\" ( by Vivi Nastase, Stan Szpakowicz, Preslav Nakov and Diarmuid 'O S'eaghdha ) will be published  by URLan & Claypool . A new Chapter 5 of the book  discusses relation classification/extraction in the deep-learning paradigm which arose after the first edition appeared. This is a preview of Chapter 5, made public by the kind permission of URLan & Claypool.", "after_revision": "The second edition of \"Semantic Relations Between Nominals\" by Vivi Nastase, Stan Szpakowicz, Preslav Nakov and Diarmuid 'O S'eaghdha will be published early in 2021 by URLan & Claypool URL A new Chapter 5 of the book , by Vivi Nastase and Stan Szpakowicz, discusses relation classification/extraction in the deep-learning paradigm which arose after the first edition appeared. This is Chapter 5, made public by the kind permission of URLan & Claypool.", "edit_actions": [{"type": "D", "before": "(", "after": null, "start_char_pos": 60, "end_char_pos": 61, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ")", "after": null, "start_char_pos": 136, "end_char_pos": 137, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "early in 2021", "start_char_pos": 156, "end_char_pos": 156, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ".", "after": "URL", "start_char_pos": 177, "end_char_pos": 178, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ", by Vivi Nastase and Stan Szpakowicz,", "start_char_pos": 207, "end_char_pos": 207, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "a preview of", "after": null, "start_char_pos": 337, "end_char_pos": 349, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 178, 328], "domain": "arxiv"}
{"doc_id": "2009.05664", "revision_depth": 1, "before_revision": "Existing commonsense reasoning datasets for AI and NLP tasks fail to address an important aspect of human life: cultural differences. In this work, we introduce an approach that extends prior work on crowdsourcing commonsense knowledge by incorporating differences in knowledge that are attributable to cultural or national groups. We demonstrate the technique by collecting commonsense knowledge that surrounds three fairly universal rituals---coming-of-age, marriage, and funerals---across three different national groups: the United States , India, and the Philippines. Our pilot study expands the different types of relationships identified by existing work in the field of commonsense reasoning for commonplace events, and uses these new types to gather information that distinguishes the knowledge of the different groups. It also moves us a step closer towards building a machine that doesn't assume a rigid framework of universal (and likely Western-biased) commonsense knowledge, but rather has the ability to reason in a contextually and culturally sensitive way. Our hope is that cultural knowledge of this sort will lead to more human-like performance in NLP tasks such as question answering (QA) and text understanding and generation.", "after_revision": "Existing commonsense reasoning datasets for AI and NLP tasks fail to address an important aspect of human life: cultural differences. In this work, we introduce an approach that extends prior work on crowdsourcing commonsense knowledge by incorporating differences in knowledge that are attributable to cultural or national groups. We demonstrate the technique by collecting commonsense knowledge that surrounds six fairly universal rituals---birth, coming-of-age, marriage, funerals, new year, and birthdays---across two national groups: the United States and India. Our study expands the different types of relationships identified by existing work in the field of commonsense reasoning for commonplace events, and uses these new types to gather information that distinguishes the knowledge of the different groups. It also moves us a step closer towards building a machine that doesn't assume a rigid framework of universal (and likely Western-biased) commonsense knowledge, but rather has the ability to reason in a contextually and culturally sensitive way. Our hope is that cultural knowledge of this sort will lead to more human-like performance in NLP tasks such as question answering (QA) and text understanding and generation.", "edit_actions": [{"type": "R", "before": "three fairly universal rituals---coming-of-age, marriage, and funerals---across three different", "after": "six fairly universal rituals---birth, coming-of-age, marriage, funerals, new year, and birthdays---across two", "start_char_pos": 412, "end_char_pos": 507, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", India, and the Philippines. Our pilot", "after": "and India. Our", "start_char_pos": 543, "end_char_pos": 582, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 133, 331, 572, 828, 1073]}
{"doc_id": "2009.08553", "revision_depth": 1, "before_revision": "Conventional sparse retrieval methods such as TF-IDF and BM25 are simple and efficient, but solely rely on lexical overlap and fail to conduct semantic matching. Recent dense retrieval methods learn latent representations to tackle the lexical mismatch problem, while being more computationally expensive and sometimes insufficient for exact matching as they embed the entire text sequence into a single vector with limited capacity. In this paper, we present Generation-Augmented Retrieval (GAR), a query expansion method that augments a query with relevant contexts through text generation. We demonstrate on open-domain question answering (QA) that the generated contexts significantly enrich the semantics of the queries and thus GAR with sparse representations (BM25) achieves comparable or better performance than the current state-of-the-art dense method DPR cite{karpukhin2020dense}. We show that generating various contexts of a query is beneficial as fusing their results consistently yields a better retrieval accuracy. Moreover, GAR achieves the state-of-the-art performance of extractive QA on the Natural Questions and TriviaQA datasets  when equipped with an extractive reader  .", "after_revision": "Conventional sparse retrieval methods such as TF-IDF and BM25 are simple and efficient, but solely rely on lexical overlap without semantic matching. Recent dense retrieval methods learn latent representations to tackle the lexical mismatch problem, while being more computationally expensive and insufficient for exact matching as they embed the text sequence into a single vector with limited capacity. In this paper, we present Generation-Augmented Retrieval (GAR), a query expansion method that augments a query with relevant contexts through text generation. We demonstrate on open-domain question answering that the generated contexts significantly enrich the semantics of the queries and thus GAR with sparse representations (BM25) achieves comparable or better performance than the state-of-the-art dense methods such as DPR cite{karpukhin2020dense}. We show that generating various contexts of a query is beneficial as fusing their results consistently yields better retrieval accuracy. Moreover, as sparse and dense representations are often complementary, GAR can be easily combined with DPR to achieve even better performance. Furthermore, GAR achieves the state-of-the-art performance on the Natural Questions and TriviaQA datasets under the extractive setting when equipped with an extractive reader , and consistently outperforms other retrieval methods when the same generative reader is used .", "edit_actions": [{"type": "R", "before": "and fail to conduct", "after": "without", "start_char_pos": 123, "end_char_pos": 142, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "sometimes", "after": null, "start_char_pos": 309, "end_char_pos": 318, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "entire", "after": null, "start_char_pos": 369, "end_char_pos": 375, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "(QA)", "after": null, "start_char_pos": 642, "end_char_pos": 646, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "current", "after": null, "start_char_pos": 824, "end_char_pos": 831, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "method", "after": "methods such as", "start_char_pos": 855, "end_char_pos": 861, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "a", "after": null, "start_char_pos": 1002, "end_char_pos": 1003, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "GAR", "after": "as sparse and dense representations are often complementary, GAR can be easily combined with DPR to achieve even better performance. Furthermore, GAR", "start_char_pos": 1041, "end_char_pos": 1044, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "of extractive QA", "after": null, "start_char_pos": 1087, "end_char_pos": 1103, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "under the extractive setting", "start_char_pos": 1151, "end_char_pos": 1151, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ", and consistently outperforms other retrieval methods when the same generative reader is used", "start_char_pos": 1192, "end_char_pos": 1192, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 161, 433, 592, 891, 1030], "domain": "arxiv"}
{"doc_id": "2009.09704", "revision_depth": 2, "before_revision": "An end-to-end speech-to-text translation (ST) takes audio in a source language and outputs the text in a target language. Inspired by neuroscience, humans have perception systems and cognitive systems to process different information , we propose LUT, Listen-Understand-Translate,  a unified framework with triple supervision  to decouple the end-to-end speech-to-text translation task. In addition to the target language sentence translation loss, LUT includes two auxiliary supervising signals to guide the acoustic encoder to extracts acoustic features from the input, and the semantic encoder to extract semantic features relevant to the source transcription text. We do experiments on English-French,  English-German and  English-Chinese speech translation benchmarks and the results demonstrate the reasonability of LUT . Our code is available at URL", "after_revision": "An end-to-end speech-to-text translation (ST) takes audio in a source language and outputs the text in a target language. Existing methods are limited by the amount of parallel corpus. Can we build a system to fully utilize signals in a parallel ST corpus? We are inspired by human understanding system which is composed of auditory perception and cognitive processing. In this paper , we propose Listen-Understand-Translate, (LUT), a unified framework with triple supervision signals to decouple the end-to-end speech-to-text translation task. LUT is able to guide the acoustic encoder to extract as much information from the auditory input. In addition, LUT utilizes a pre-trained BERT model to enforce the upper encoder to produce as much semantic information as possible, without extra data. We perform experiments on a diverse set of speech translation benchmarks, including Librispeech English-French, IWSLT English-German and TED English-Chinese . Our results demonstrate LUT achieves the state-of-the-art performance, outperforming previous methods. The code is available at URL", "edit_actions": [{"type": "R", "before": "Inspired by neuroscience, humans have perception systems and cognitive systems to process different information", "after": "Existing methods are limited by the amount of parallel corpus. Can we build a system to fully utilize signals in a parallel ST corpus? We are inspired by human understanding system which is composed of auditory perception and cognitive processing. In this paper", "start_char_pos": 122, "end_char_pos": 233, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "LUT,", "after": null, "start_char_pos": 247, "end_char_pos": 251, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "(LUT),", "start_char_pos": 281, "end_char_pos": 281, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "signals", "start_char_pos": 326, "end_char_pos": 326, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "In addition to the target language sentence translation loss, LUT includes two auxiliary supervising signals", "after": "LUT is able", "start_char_pos": 387, "end_char_pos": 495, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "extracts acoustic features from the input, and the semantic encoder to extract semantic features relevant to the source transcription text. We do experiments on", "after": "extract as much information from the auditory input. In addition, LUT utilizes a pre-trained BERT model to enforce the upper encoder to produce as much semantic information as possible, without extra data. We perform experiments on a diverse set of speech translation benchmarks, including Librispeech", "start_char_pos": 529, "end_char_pos": 689, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "IWSLT", "start_char_pos": 706, "end_char_pos": 706, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "TED", "start_char_pos": 726, "end_char_pos": 726, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "speech translation benchmarks and the results demonstrate the reasonability of LUT . Our", "after": ". Our results demonstrate LUT achieves the state-of-the-art performance, outperforming previous methods. The", "start_char_pos": 743, "end_char_pos": 831, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 121, 386, 668], "domain": "arxiv"}
{"doc_id": "2009.09737", "revision_depth": 2, "before_revision": "Speech-to-text translation (ST), which directly translates the source language speech to the target language text, has attracted intensive attention recently. However, the combination of speech recognition and machine translation in a single model poses a heavy burden on the direct cross-modal cross-lingual mapping. To reduce the learning difficulty, we propose COnSecutive Transcription and Translation (COSTT), an integral framework for speech-to-text translation.  Our method is verified on three mainstream datasets, including Augmented LibriSpeech English-French dataset, TED English-German dataset, and TED English-Chinese dataset. Experiments show that our proposed COSTT outperforms the previous state-of-the-art methods. Our code is available at URL", "after_revision": "Speech-to-text translation (ST), which directly translates the source language speech to the target language text, has attracted intensive attention recently. However, the combination of speech recognition and machine translation in a single model poses a heavy burden on the direct cross-modal cross-lingual mapping. To reduce the learning difficulty, we propose COnSecutive Transcription and Translation (COSTT), an integral approach for speech-to-text translation. The key idea is to generate source transcript and target translation text with a single decoder. It benefits the model training so that additional large parallel text corpus can be fully exploited to enhance the speech translation training. Our method is verified on three mainstream datasets, including Augmented LibriSpeech English-French dataset, TED English-German dataset, and TED English-Chinese dataset. Experiments show that our proposed COSTT outperforms the previous state-of-the-art methods. The code is available at URL", "edit_actions": [{"type": "R", "before": "framework", "after": "approach", "start_char_pos": 427, "end_char_pos": 436, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "The key idea is to generate source transcript and target translation text with a single decoder. It benefits the model training so that additional large parallel text corpus can be fully exploited to enhance the speech translation training.", "start_char_pos": 469, "end_char_pos": 469, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Our", "after": "The", "start_char_pos": 732, "end_char_pos": 735, "major_intent": "style", "raw_intents": ["style", "clarity", "fluency"]}], "sents_char_pos": [0, 158, 317, 468, 639, 731], "domain": "arxiv"}
{"doc_id": "2009.11616", "revision_depth": 1, "before_revision": "We introduce texttt N-LTP , an open-source Python Chinese natural language processing toolkit supporting five basic tasks: Chinese word segmentation, part-of-speech tagging, named entity recognition, dependency parsing, and semantic dependency parsing. texttt N-LTP adopts the multi-task framework with the pre-trained model to capture the shared knowledge across all Chinese relevant tasks. In addition, we propose to use knowledge distillation where single-task models teach a multi-task model, helping the multi-task model surpass its single-task teachers. Finally, we provide fundamental tasks API and a visualization tool to make users easier to use and view the processing results directly. To the best of our knowledge, this is the first toolkit to support all Chinese NLP fundamental tasks. Source code, documentation, and pre-trained models are available at URL", "after_revision": "We introduce N-LTP , an open-source Python Chinese natural language processing toolkit supporting five basic tasks: Chinese word segmentation, part-of-speech tagging, named entity recognition, dependency parsing, and semantic dependency parsing. N-LTP adopts the multi-task framework with the pre-trained model to capture the shared knowledge across all Chinese relevant tasks. In addition, we propose to use knowledge distillation where single-task models teach a multi-task model, helping the multi-task model surpass its single-task teachers. Finally, we provide fundamental tasks API and a visualization tool to make users easier to use and view the processing results directly. To the best of our knowledge, this is the first toolkit to support all Chinese NLP fundamental tasks. Source code, documentation, and pre-trained models are available at URL", "edit_actions": [{"type": "D", "before": "texttt", "after": null, "start_char_pos": 13, "end_char_pos": 19, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "texttt", "after": null, "start_char_pos": 253, "end_char_pos": 259, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 391, 559, 696, 798], "domain": "arxiv"}
{"doc_id": "2009.13267", "revision_depth": 1, "before_revision": "The discrepancy between maximum likelihood estimation (MLE) and task measures such as BLEU score has been studied before for autoregressive neural machine translation (NMT) and resulted in alternative training algorithms (Ranzato et al., 2016; Norouzi et al., 2016; Shen et al., 2016; Wu et al., 2018). However, MLE training remains the de facto approach for autoregressive NMT because of its computational efficiency and stability. Despite this mismatch between the training objective and task measure, we notice that the samples drawn from an MLE-based trained NMT support the desired distribution -- there are samples with much higher BLEU score comparing to the beam decoding output. To benefit from this observation, we train an energy-based model to mimic the behavior of the task measure (i.e., the energy-based model assigns lower energy to samples with higher BLEU score), which is resulted in a re-ranking algorithm based on the samples drawn from NMT: energy-based re-ranking (EBR). Our EBR consistently improves the performance of the Transformer-based NMT: +3 BLEU points on Sinhala-English and +2.0 BLEU points on IWSLT'17 French-English  tasks.", "after_revision": "The discrepancy between maximum likelihood estimation (MLE) and task measures such as BLEU score has been studied before for autoregressive neural machine translation (NMT) and resulted in alternative training algorithms (Ranzato et al., 2016; Norouzi et al., 2016; Shen et al., 2016; Wu et al., 2018). However, MLE training remains the de facto approach for autoregressive NMT because of its computational efficiency and stability. Despite this mismatch between the training objective and task measure, we notice that the samples drawn from an MLE-based trained NMT support the desired distribution -- there are samples with much higher BLEU score comparing to the beam decoding output. To benefit from this observation, we train an energy-based model to mimic the behavior of the task measure (i.e., the energy-based model assigns lower energy to samples with higher BLEU score), which is resulted in a re-ranking algorithm based on the samples drawn from NMT: energy-based re-ranking (EBR). Our EBR consistently improves the performance of the Transformer-based NMT: +3 BLEU points on Sinhala-English , +2.0 BLEU points on IWSLT'17 French-English , and +1.7 BLEU points on WMT'19 German-English tasks.", "edit_actions": [{"type": "R", "before": "and", "after": ",", "start_char_pos": 1104, "end_char_pos": 1107, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "A", "before": null, "after": ", and +1.7 BLEU points on WMT'19 German-English", "start_char_pos": 1152, "end_char_pos": 1152, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 243, 265, 284, 302, 432, 687, 993], "domain": "arxiv"}
{"doc_id": "2009.13267", "revision_depth": 2, "before_revision": "The discrepancy between maximum likelihood estimation (MLE) and task measures such as BLEU score has been studied before for autoregressive neural machine translation (NMT) and resulted in alternative training algorithms (Ranzato et al., 2016; Norouzi et al., 2016; Shen et al., 2016; Wu et al., 2018). However, MLE training remains the de facto approach for autoregressive NMT because of its computational efficiency and stability. Despite this mismatch between the training objective and task measure, we notice that the samples drawn from an MLE-based trained NMT support the desired distribution -- there are samples with much higher BLEU score comparing to the beam decoding output. To benefit from this observation, we train an energy-based model to mimic the behavior of the task measure (i.e., the energy-based model assigns lower energy to samples with higher BLEU score), which is resulted in a re-ranking algorithm based on the samples drawn from NMT: energy-based re-ranking (EBR). Our EBR consistently improves the performance of the Transformer-based NMT: + 3 BLEU points on Sinhala-English , + 2.0 BLEU points on IWSLT'17 French-English, and + 1.7 BLEU points on WMT' 19 German-English tasks.", "after_revision": "The discrepancy between maximum likelihood estimation (MLE) and task measures such as BLEU score has been studied before for autoregressive neural machine translation (NMT) and resulted in alternative training algorithms (Ranzato et al., 2016; Norouzi et al., 2016; Shen et al., 2016; Wu et al., 2018). However, MLE training remains the de facto approach for autoregressive NMT because of its computational efficiency and stability. Despite this mismatch between the training objective and task measure, we notice that the samples drawn from an MLE-based trained NMT support the desired distribution -- there are samples with much higher BLEU score comparing to the beam decoding output. To benefit from this observation, we train an energy-based model to mimic the behavior of the task measure (i.e., the energy-based model assigns lower energy to samples with higher BLEU score), which is resulted in a re-ranking algorithm based on the samples drawn from NMT: energy-based re-ranking (EBR). We use both marginal energy models (over target sentence) and joint energy models (over both source and target sentences). Our EBR with the joint energy model consistently improves the performance of the Transformer-based NMT: + 4 BLEU points on IWSLT'14 German-English , + 3.0 BELU points on Sinhala-English, + 1.2 BLEU on WMT' 16 English-German tasks.", "edit_actions": [{"type": "R", "before": "Our EBR", "after": "We use both marginal energy models (over target sentence) and joint energy models (over both source and target sentences). Our EBR with the joint energy model", "start_char_pos": 994, "end_char_pos": 1001, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "3", "after": "4", "start_char_pos": 1072, "end_char_pos": 1073, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Sinhala-English", "after": "IWSLT'14 German-English", "start_char_pos": 1089, "end_char_pos": 1104, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "2.0 BLEU points on IWSLT'17 French-English, and", "after": "3.0 BELU points on Sinhala-English,", "start_char_pos": 1109, "end_char_pos": 1156, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "1.7 BLEU points", "after": "1.2 BLEU", "start_char_pos": 1159, "end_char_pos": 1174, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "19 German-English", "after": "16 English-German", "start_char_pos": 1183, "end_char_pos": 1200, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}], "sents_char_pos": [0, 243, 265, 284, 302, 432, 687, 993], "domain": "arxiv"}
{"doc_id": "2010.01061", "revision_depth": 1, "before_revision": "For natural language processing (NLP) taskssuch as sentiment or topic classification, currently prevailing approaches heavily rely on pretraining large self-supervised models on massive external data resources. However, this methodology is being critiqued for: exceptional compute and pretraining data requirements ; diminishing returns on both large and small datasets; and importantly, favourable evaluation settings that overestimate performance differences. The core belief behind current methodology, coined `the bitter lesson' by R. Sutton, is that `compute scale-up beats data and compute-efficient algorithms', neglecting that progress in compute hardware scale-up is based almost entirely on the miniaturisation of resource consumption. We thus approach pretrainingfrom a miniaturisation perspective, such as not to require massive external data sources and models, or learned translations from continuous input embeddings to discrete labels. To minimise overly favourable evaluation, we examine learning on a long-tailed, low-resource, multi-label text classification dataset with noisy, highly sparse labels and many rare concepts. To this end, we propose a novel `dataset-internal' contrastive autoencoding approach to self-supervised pretraining and demonstrate marked improvements in zero-shot, few-shot and solely supervised learning performance; even under an unfavorable low-resource scenario, and without defaulting to large-scale external datasets for self-supervision. We also find empirical evidence that zero and few-shot learning markedly benefit from adding more ` dataset-internal' , self-supervised training signals, which is of practical importance when retrieving or computing on large external sources of such signals is infeasible .", "after_revision": "For natural language processing 'text-to-text' tasks, the prevailing approaches heavily rely on pretraining large self-supervised models on massive external data sources, which incurs exceptional pretraining data requirements and a diminished ability to pretrain over small datasets. However, fundamental pretraining method capabilities like few to zero-shot learning or preserving minority concept (long-tail) prediction performance along with accordingly designed evaluation scenarios remain open challenges. We thus propose Contrastive Label-Embedding Self-Supervision (CLESS) pretraining, which enables pretraining from multiple magnitudes smaller, 'task internal' data only, while still strongly improving fully supervised, long-tail, few-shot and self-supervised zero-shot learning abilities. Accordingly, we analyse improvements in learning dynamics over baselines on a challenging long-tailed, low-resource, multi-label text classification scenario with noisy, highly sparse labels and many minority concepts. We find that long-tailed zero and few-shot learning markedly benefit from increasing ' dataset-internal' self-supervised pretraining signals, to help reduce the reliance on large external sources .", "edit_actions": [{"type": "R", "before": "(NLP) taskssuch as sentiment or topic classification, currently", "after": "'text-to-text' tasks, the", "start_char_pos": 32, "end_char_pos": 95, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "resources. However, this methodology is being critiqued for: exceptional compute and", "after": "sources, which incurs exceptional", "start_char_pos": 200, "end_char_pos": 284, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "; diminishing returns on both large and small datasets; and importantly, favourable evaluation settings that overestimate performance differences. The core belief behind current methodology, coined `the bitter lesson' by R. Sutton, is that `compute scale-up beats data and compute-efficient algorithms', neglecting that progress in compute hardware scale-up is based almost entirely on the miniaturisation of resource consumption. We thus approach pretrainingfrom a miniaturisation perspective, such as not to require massive external data sources and models, or learned translations from continuous input embeddings to discrete labels. To minimise overly favourable evaluation, we examine learning on a", "after": "and a diminished ability to pretrain over small datasets. However, fundamental pretraining method capabilities like few to zero-shot learning or preserving minority concept (long-tail) prediction performance along with accordingly designed evaluation scenarios remain open challenges. We thus propose Contrastive Label-Embedding Self-Supervision (CLESS) pretraining, which enables pretraining from multiple magnitudes smaller, 'task internal' data only, while still strongly improving fully supervised, long-tail, few-shot and self-supervised zero-shot learning abilities. Accordingly, we analyse improvements in learning dynamics over baselines on a challenging", "start_char_pos": 315, "end_char_pos": 1018, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "dataset", "after": "scenario", "start_char_pos": 1078, "end_char_pos": 1085, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "rare concepts. To this end, we propose a novel `dataset-internal' contrastive autoencoding approach to self-supervised pretraining and demonstrate marked improvements in zero-shot, few-shot and solely supervised learning performance; even under an unfavorable low-resource scenario, and without defaulting to large-scale external datasets for self-supervision. We also find empirical evidence that", "after": "minority concepts. We find that long-tailed", "start_char_pos": 1128, "end_char_pos": 1525, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "adding more `", "after": "increasing '", "start_char_pos": 1575, "end_char_pos": 1588, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1607, "end_char_pos": 1608, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "training signals, which is of practical importance when retrieving or computing", "after": "pretraining signals, to help reduce the reliance", "start_char_pos": 1625, "end_char_pos": 1704, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "of such signals is infeasible", "after": null, "start_char_pos": 1731, "end_char_pos": 1760, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 210, 316, 370, 461, 745, 951, 1142, 1361, 1488], "domain": "arxiv"}
{"doc_id": "2010.01061", "revision_depth": 2, "before_revision": "For natural language processing ' text-to-text' tasks, the prevailing approaches heavily rely on pretraining large self-supervised models on massive external datasources, which incurs exceptional pretraining data requirements and a diminished ability to pretrain over small datasets. However, fundamental pretraining method capabilities like few to zero-shot learningor preserving minority concept ( long-tail ) prediction performance along with accordingly designed evaluation scenarios remain open challenges . We thus propose Contrastive Label-Embedding Self-Supervision (CLESS) pretraining , which enables pretraining from multiple magnitudes smaller, 'task internal' data only, while still strongly improving fully supervised, long-tail , few-shot and self-supervised zero-shot learning abilities. Accordingly, we analyse improvements in learning dynamics over baselines on a challenging long-tailed, low-resource, multi-label text classification scenario with noisy, highly sparse labels and many minority concepts . We find that long-tailed zero and few-shot learning markedly benefit from increasing 'dataset-internal' self-supervised pretraining signals, to help reduce the reliance on large external sources .", "after_revision": "For natural language processing ` text-to-text' tasks, the prevailing approaches heavily rely on pretraining large self-supervised models on increasingly larger `task-external' data. Transfer learning from high-resource pretraining works well, but research has focused on settings with very large data and compute requirements, while the potential of efficient low-resource learning, without large `task-external' pretraining, remains under-explored. In this work, we evaluate against three core challenges for resource efficient learning. Namely, we analyze: (1) pretraining data (X) efficiency; (2) zero to few-shot label (Y) efficiency; and (3) long-tail generalization, since long-tail preservation has been linked to algorithmic fairness and because data in the tail is limited by definition. To address these challenges, we propose a data and compute efficient self-supervised, contrastive text encoder, pretrained on 60MB of `task-internal' text data, and compare it to RoBERTa, which was pretrained on 160GB of `task-external' text . We find our method outperforms RoBERTa, while pretraining and fine-tuning in a 1/5th of RoBERTa's fine-tuning time .", "edit_actions": [{"type": "R", "before": "'", "after": "`", "start_char_pos": 32, "end_char_pos": 33, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "massive external datasources, which incurs exceptional pretraining", "after": "increasingly larger `task-external' data. Transfer learning from high-resource pretraining works well, but research has focused on settings with very large data and compute requirements, while the potential of efficient low-resource learning, without large `task-external' pretraining, remains under-explored. In this work, we evaluate against three core challenges for resource efficient learning. Namely, we analyze: (1) pretraining", "start_char_pos": 141, "end_char_pos": 207, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "requirements and a diminished ability to pretrain over small datasets. However, fundamental pretraining method capabilities like few to zero-shot learningor preserving minority concept (", "after": "(X) efficiency; (2) zero to few-shot label (Y) efficiency; and (3)", "start_char_pos": 213, "end_char_pos": 399, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": ") prediction performance along with accordingly designed evaluation scenarios remain open challenges . We thus propose Contrastive Label-Embedding Self-Supervision (CLESS) pretraining , which enables pretraining from multiple magnitudes smaller, 'task internal' data only, while still strongly improving fully supervised,", "after": "generalization, since", "start_char_pos": 410, "end_char_pos": 731, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ", few-shot and self-supervised zero-shot learning abilities. Accordingly, we analyse improvements in learning dynamics over baselines on a challenging long-tailed, low-resource, multi-label text classification scenario with noisy, highly sparse labels and many minority concepts", "after": "preservation has been linked to algorithmic fairness and because data in the tail is limited by definition. To address these challenges, we propose a data and compute efficient self-supervised, contrastive text encoder, pretrained on 60MB of `task-internal' text data, and compare it to RoBERTa, which was pretrained on 160GB of `task-external' text", "start_char_pos": 742, "end_char_pos": 1020, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "that long-tailed zero and few-shot learning markedly benefit from increasing 'dataset-internal' self-supervised pretraining signals, to help reduce the reliance on large external sources", "after": "our method outperforms RoBERTa, while pretraining and fine-tuning in a 1/5th of RoBERTa's fine-tuning time", "start_char_pos": 1031, "end_char_pos": 1217, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 283, 512, 802, 1022], "domain": "arxiv"}
{"doc_id": "2010.12008", "revision_depth": 1, "before_revision": "We propose a simple method to generate large amounts of multilingual question and answer pairs by a single generative model. These synthetic samples are then applied to augment the available gold multilingual ones to improve the performance of multilingual QA models on target languages. Our approach only requires existence of automatically translated samples from Englishto the target domain , thus removing the need for human annotations in the target languages  . Experimental results show our proposed approach achieves significant gains in a number of multilingual datasets .", "after_revision": "We propose a simple method to generate multilingual question and answer pairs on a large scale through the use of a single generative model. These synthetic samples can be used to improve the zero-shot performance of multilingual QA models on target languages. Our proposed multi-task training of the generative model only requires the training samples in English , thus removing the need for labeled samples in the target languages , making it applicable to far more languages than those with labeled data . Experimental results show our proposed approach achieves significant gains on several multilingual QA benchmarks, reducing the gap between zero-shot and supervised performance of QA models on various languages .", "edit_actions": [{"type": "D", "before": "large amounts of", "after": null, "start_char_pos": 39, "end_char_pos": 55, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "by a", "after": "on a large scale through the use of a", "start_char_pos": 95, "end_char_pos": 99, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "are then applied to augment the available gold multilingual ones to improve the", "after": "can be used to improve the zero-shot", "start_char_pos": 149, "end_char_pos": 228, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "approach only requires existence of automatically translated samples from Englishto the target domain", "after": "proposed multi-task training of the generative model only requires the training samples in English", "start_char_pos": 292, "end_char_pos": 393, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "human annotations", "after": "labeled samples", "start_char_pos": 423, "end_char_pos": 440, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": ", making it applicable to far more languages than those with labeled data", "start_char_pos": 465, "end_char_pos": 465, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in a number of multilingual datasets", "after": "on several multilingual QA benchmarks, reducing the gap between zero-shot and supervised performance of QA models on various languages", "start_char_pos": 543, "end_char_pos": 579, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}], "sents_char_pos": [0, 124, 287], "domain": "arxiv"}
{"doc_id": "2010.12008", "revision_depth": 2, "before_revision": "We propose a simple method to generate multilingual question and answer pairs on a large scale through the use of a single generative model. These synthetic samples can be used to improve the zero-shot performance of multilingual QA models on target languages. Our proposed multi-task training of the generative model only requires the  training samples in English, thus removing the need for labeled samples in the target languages, making it applicable to far more languages than those with labeled data.  Experimental results show our proposed approach achieves significant gains on several multilingual QA benchmarks , reducing the gap between zero-shot and supervised performance of  QA models on various languages.", "after_revision": "We propose a simple method to generate multilingual question and answer pairs on a large scale through the use of a single generative model. These synthetic samples can be used to improve the zero-shot performance of multilingual QA models on target languages. Our proposed multi-task training of the generative model only requires the labeled training samples in English, thus removing the need for such samples in the target languages, making it applicable to far more languages than those with labeled data. Human evaluations indicate the majority of such samples are grammatically correct and sensible. Experimental results show our proposed approach can achieve large gains on the XQuAD dataset , reducing the gap between zero-shot and supervised performance of smaller QA models on various languages.", "edit_actions": [{"type": "A", "before": null, "after": "labeled", "start_char_pos": 336, "end_char_pos": 336, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "labeled", "after": "such", "start_char_pos": 393, "end_char_pos": 400, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Human evaluations indicate the majority of such samples are grammatically correct and sensible.", "start_char_pos": 507, "end_char_pos": 507, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "achieves significant gains on several multilingual QA benchmarks", "after": "can achieve large gains on the XQuAD dataset", "start_char_pos": 556, "end_char_pos": 620, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "smaller", "start_char_pos": 688, "end_char_pos": 688, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 140, 260, 506], "domain": "arxiv"}
{"doc_id": "2010.12789", "revision_depth": 1, "before_revision": "Natural language is one of the ways information is encoded and it has highly abstracted and conceptualized the information. This paper disassembles the information represented by natural language , analyzes the classification coding system of attribute information  and the abstraction relation between attribute information and entities in the real world , constructs the storage model of information, and simulate the attribute information precessing process in one of the attribute spaces, interprets how the relations which represented by \"Be\", \"Of\", \"Have\", and so on are embodied in the information storage data structures and the corresponding data reading modes , reclassifies the sentences types from the perspective of task types and data reading modes. Then, simulated the understanding process (the information processing process) on a dialogue example. Finally, the author summarizes the basic conditions of understanding and gives out the definition of understanding from a personal point of view. The study in this paper provides a practical, theoretical basis and research methods for NLU.It also can be applied in large-scale, multi-type information processing in the artificial intelligence (AI) area .", "after_revision": "First of all, please URLet all you knew about the lexical classification, then let's jump to the conclusion. This paper reclassified lexical chunks into data chunks, structure chunks, and pointer chunks. Almost all data chunks are information sets. According to the difference of the set structures, data chunks can be further divided into attribute chunks and entity chunks. According to the different abstraction level and method, attribute chunks can be further divided into basic attribute chunks, extended attribute chunks, and advanced attribute chunks. All of the above classification principles are structural and functionalbased discrimination, instead of artificially divide lexical chunks into a noun, adjective, pronouns, and so on. Now, let's back to the normal study process. The author believes natural language is one of the ways information is encoded and it has highly abstracted and conceptualized the information. Therefore the study begins with disassembling the information represented by natural language and then discovered the classification coding system of attribute information , and the abstraction relations between attribute information and entities in the real world . To have a clear and better discussion, the author constructed corresponding data storage models, and extract three kinds of data reading modes on those data storage models, they are the defining reading mode which is driven by the structural word: be, the set reading mode which is driven by the structural word: have, and the process reading mode which is driven by verbs. Sentences output by the above data reading modes can be further divided into the data description task, the data verification task, and the data search task, according to task types represented by these sentences .. .", "edit_actions": [{"type": "R", "before": "Natural", "after": "First of all, please URLet all you knew about the lexical classification, then let's jump to the conclusion. This paper reclassified lexical chunks into data chunks, structure chunks, and pointer chunks. Almost all data chunks are information sets. According to the difference of the set structures, data chunks can be further divided into attribute chunks and entity chunks. According to the different abstraction level and method, attribute chunks can be further divided into basic attribute chunks, extended attribute chunks, and advanced attribute chunks. All of the above classification principles are structural and functionalbased discrimination, instead of artificially divide lexical chunks into a noun, adjective, pronouns, and so on. Now, let's back to the normal study process. The author believes natural", "start_char_pos": 0, "end_char_pos": 7, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "This paper disassembles the", "after": "Therefore the study begins with disassembling the", "start_char_pos": 124, "end_char_pos": 151, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": ", analyzes", "after": "and then discovered", "start_char_pos": 196, "end_char_pos": 206, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 265, "end_char_pos": 265, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "relation", "after": "relations", "start_char_pos": 286, "end_char_pos": 294, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": ", constructs the storage model of information, and simulate the attribute information precessing process in one of", "after": ". To have a clear and better discussion,", "start_char_pos": 356, "end_char_pos": 470, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "attribute spaces, interprets how the relations which represented by \"Be\", \"Of\", \"Have\", and so on are embodied in the information storage data structures and the corresponding", "after": "author constructed corresponding data storage models, and extract three kinds of data reading modes on those data storage models, they are the defining reading mode which is driven by the structural word: be, the set reading mode which is driven by the structural word: have, and the process reading mode which is driven by verbs. Sentences output by the above", "start_char_pos": 475, "end_char_pos": 650, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": ", reclassifies the sentences types from the perspective of task types and data reading modes. Then, simulated the understanding process (the information processing process) on a dialogue example. Finally, the author summarizes the basic conditions of understanding and gives out the definition of understanding from a personal point of view. The study in this paper provides a practical, theoretical basis and research methods for NLU.It also can be applied in large-scale, multi-type information processing in the artificial intelligence (AI) area", "after": "can be further divided into the data description task, the data verification task, and the data search task, according to task types represented by these sentences ..", "start_char_pos": 670, "end_char_pos": 1218, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 123, 197, 357, 763, 865, 1011, 1105], "domain": "arxiv"}
{"doc_id": "2010.12789", "revision_depth": 2, "before_revision": "First of all, please URLet all you knew about the lexical classification, then let's jump to the conclusion. This paper reclassified lexical chunks into data chunks, structure chunks, and pointer chunks. Almost all data chunks are information sets. According to the difference of the set structures, data chunks can be further divided into attribute chunks and entity chunks. According to the different abstraction level and method, attribute chunks can be further divided into basic attribute chunks, extended attribute chunks, and advanced attribute chunks. All of the above classification principles are structural and functionalbased discrimination, instead of artificially divide lexical chunks into a noun, adjective, pronouns, and so on. Now, let's back to the normal study process. The author believes natural language is one of the ways information is encoded and it has highly abstracted and conceptualized the information . Therefore the study begins with disassembling the information represented by natural language and then discovered the classification coding system of attribute information, and the abstraction relations between attribute information and entities in the real world. To have a clear and better discussion , the author constructed corresponding data storage models, and extract three kinds of data reading modes on those data storage models, they are the defining reading mode which is driven by the structural word: be, the set reading mode which is driven by the structural word: have, and the process reading mode which is driven by verbs. Sentences output by the above data reading modes can be further divided into the data description task, the data verification task, and the data search task, according to task types represented by these sentences .. .", "after_revision": "We must recognize that natural language is a way of information encoding, and it encodes not only the information but also the procedures for how information is processed. To understand natural language, the same as we conceive and design computer languages, the first step is to separate information (or data) and the processing procedures of information (or data). In natural language, some processing procedures of data are encoded directly as the structure chunk and the pointer chunk (this paper has reclassified lexical chunks as the data chunk, structure chunk, and the pointer chunk); some processing procedures of data imply in sentences structures; some requests of processing procedures are expressed by information senders and processed by information receivers. For the data parts, the classification encoding system of attribute information and the URLanization architecture (including constitutional structures of information sets and the hierarchy between the information sets) were discussed. In section 2, the theoretical part elaborated in section 2 has been verified in examples and proofed that the studies in this paper have achieved the goal of enabling machines to understand the information conveyed in the dialogue. In section 4 , the author summarizes the basic conditions of \"Understanding\", rethinks what \"Understanding\" is and how to proceed. The study in this paper provides a practical, theoretical basis and research methods for NLU. It also can be applied in large-scale and multi-type information processing in the artificial intelligence (AI) area .", "edit_actions": [{"type": "R", "before": "First of all, please URLet all you knew about the lexical classification, then let's jump to the conclusion. This paper", "after": "We must recognize that natural language is a way of information encoding, and it encodes not only the information but also the procedures for how information is processed. To understand natural language, the same as we conceive and design computer languages, the first step is to separate information (or data) and the processing procedures of information (or data). In natural language, some processing procedures of data are encoded directly as the structure chunk and the pointer chunk (this paper has", "start_char_pos": 0, "end_char_pos": 119, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "into data chunks, structure chunks, and pointer chunks. Almost all data chunks are", "after": "as the data chunk, structure chunk, and the pointer chunk); some processing procedures of data imply in sentences structures; some requests of processing procedures are expressed by information senders and processed by information receivers. For the data parts, the classification encoding system of attribute information and the URLanization architecture (including constitutional structures of", "start_char_pos": 148, "end_char_pos": 230, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "sets. According to the difference of the set structures, data chunks can be further divided into attribute chunks and entity chunks. According to the different abstraction level and method, attribute chunks can be further divided into basic attribute chunks, extended attribute chunks, and advanced attribute chunks. All of the above classification principles are structural and functionalbased discrimination, instead of artificially divide lexical chunks into a noun, adjective, pronouns, and so on. Now, let's back to the normal study process. The author believes natural language is one of the ways information is encoded and it has highly abstracted and conceptualized the", "after": "sets and the hierarchy between the information sets) were discussed. In section 2, the theoretical part elaborated in section 2 has been verified in examples and proofed that the studies in this paper have achieved the goal of enabling machines to understand the", "start_char_pos": 243, "end_char_pos": 920, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". Therefore the study begins with disassembling the information represented by natural language and then discovered the classification coding system of attribute information, and the abstraction relations between attribute information and entities in the real world. To have a clear and better discussion", "after": "conveyed in the dialogue. In section 4", "start_char_pos": 933, "end_char_pos": 1237, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "constructed corresponding data storage models, and extract three kinds of data reading modes on those data storage models, they are the defining reading mode which is driven by the structural word: be, the set reading mode which is driven by the structural word: have, and the process reading mode which is driven by verbs. Sentences output by the above data reading modes can be further divided into the data description task, the data verification task, and the data search task, according to task types represented by these sentences ..", "after": "summarizes the basic conditions of \"Understanding\", rethinks what \"Understanding\" is and how to proceed. The study in this paper provides a practical, theoretical basis and research methods for NLU. It also can be applied in large-scale and multi-type information processing in the artificial intelligence (AI) area", "start_char_pos": 1251, "end_char_pos": 1790, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 108, 203, 248, 375, 559, 744, 789, 934, 1199, 1574], "domain": "arxiv"}
{"doc_id": "2010.12872", "revision_depth": 1, "before_revision": "Symbolic knowledge (e.g., entities, relations, and facts in a knowledge graph) has become an increasingly popular component of neural-symbolic models applied to machine learning tasks, such as question answering and recommender systems. Besides improving downstream performance, these symbolic structures (and their associated attention weights) are often used to help explain the model's predictions and provide \" insights \" to practitioners . In this paper, we question the faithfulness of such symbolic explanations . We demonstrate that, through a learned strategy (or even simple heuristics), one can produce deceptively perturbed symbolic structures which maintain the downstream performance of the original structure while significantly deviating from the original semantics . In particular, we train a reinforcement learning policy to manipulate relation types or edge connections in a knowledge graph, such that the resulting downstream performance is maximally preserved. Across multiple models and tasks, our approach drastically alters knowledge graphs with little to no drop in performance. These results raise doubts about the faithfulness of explanations provided by learned symbolic structures and the reliability of current neural-symbolic modelsin leveraging symbolic knowledge .", "after_revision": "Knowledge graphs (KGs) have helped neural-symbolic models improve performance on various knowledge-intensive tasks, like question answering and item recommendation. By using attention over the KG, such models can also \" explain \" which KG information was most relevant for making a given prediction . In this paper, we question whether these models are really behaving as we expect . We demonstrate that, through a reinforcement learning policy (or even simple heuristics), one can produce deceptively perturbed KGs which maintain the downstream performance of the original KG while significantly deviating from the original semantics and structure. Our findings raise doubts about KG-augmented models' ability to leverage KG information and provide plausible explanations .", "edit_actions": [{"type": "R", "before": "Symbolic knowledge (e.g., entities, relations, and facts in a knowledge graph) has become an increasingly popular component of", "after": "Knowledge graphs (KGs) have helped", "start_char_pos": 0, "end_char_pos": 126, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "applied to machine learning tasks, such as", "after": "improve performance on various knowledge-intensive tasks, like", "start_char_pos": 150, "end_char_pos": 192, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "recommender systems. Besides improving downstream performance, these symbolic structures (and their associated attention weights) are often used to help explain the model's predictions and provide", "after": "item recommendation. By using attention over the KG, such models can also", "start_char_pos": 216, "end_char_pos": 412, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "insights", "after": "explain", "start_char_pos": 415, "end_char_pos": 423, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "to practitioners", "after": "which KG information was most relevant for making a given prediction", "start_char_pos": 426, "end_char_pos": 442, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the faithfulness of such symbolic explanations", "after": "whether these models are really behaving as we expect", "start_char_pos": 472, "end_char_pos": 518, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "learned strategy", "after": "reinforcement learning policy", "start_char_pos": 552, "end_char_pos": 568, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "symbolic structures", "after": "KGs", "start_char_pos": 636, "end_char_pos": 655, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "structure", "after": "KG", "start_char_pos": 714, "end_char_pos": 723, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}, {"type": "R", "before": ". In particular, we train a reinforcement learning policy to manipulate relation types or edge connections in a knowledge graph, such that the resulting downstream performance is maximally preserved. Across multiple models and tasks, our approach drastically alters knowledge graphs with little to no drop in performance. These results", "after": "and structure. Our findings", "start_char_pos": 782, "end_char_pos": 1117, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the faithfulness of explanations provided by learned symbolic structures and the reliability of current neural-symbolic modelsin leveraging symbolic knowledge", "after": "KG-augmented models' ability to leverage KG information and provide plausible explanations", "start_char_pos": 1137, "end_char_pos": 1295, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 236, 444, 520, 783, 981, 1103], "domain": "arxiv"}
{"doc_id": "2010.12872", "revision_depth": 2, "before_revision": "Knowledge graphs (KGs) have helped neural-symbolic models improve performance on various knowledge-intensive tasks, like question answering and item recommendation. By using attention over the KG, such  models can also \"explain\" which KG information was most relevant for making a given prediction. In this paper, we question whether these models are really behaving as we expect. We demonstrate that, through a reinforcement learning policy (or even simple heuristics), one can produce deceptively perturbed KGs  which maintain the downstream performance of the original KG while significantly deviating from the original  semantics and structure. Our findings raise doubts about KG-augmented models' ability to leverage KG information and provide plausible explanations.", "after_revision": "Knowledge graphs (KGs) have helped neural models improve performance on various knowledge-intensive tasks, like question answering and item recommendation. By using attention over the KG, such KG-augmented models can also \"explain\" which KG information was most relevant for making a given prediction. In this paper, we question whether these models are really behaving as we expect. We show that, through a reinforcement learning policy (or even simple heuristics), one can produce deceptively perturbed KGs , which maintain the downstream performance of the original KG while significantly deviating from the original KG's semantics and structure. Our findings raise doubts about KG-augmented models' ability to reason about KG information and give sensible explanations.", "edit_actions": [{"type": "R", "before": "neural-symbolic", "after": "neural", "start_char_pos": 35, "end_char_pos": 50, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "KG-augmented", "start_char_pos": 202, "end_char_pos": 202, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "demonstrate", "after": "show", "start_char_pos": 384, "end_char_pos": 395, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 513, "end_char_pos": 513, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "KG's", "start_char_pos": 623, "end_char_pos": 623, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "leverage", "after": "reason about", "start_char_pos": 713, "end_char_pos": 721, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "provide plausible", "after": "give sensible", "start_char_pos": 741, "end_char_pos": 758, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 164, 298, 380, 648], "domain": "arxiv"}
{"doc_id": "2010.12873", "revision_depth": 1, "before_revision": "Recently, neural-symbolic architectures have achieved success on commonsense reasoning through effectively encoding relational structures retrieved from external knowledge graphs (KGs) and obtained state-of-the-art results in tasks such as (commonsense ) question answering and natural language inference . However, these methods rely on quality and contextualized knowledge structures (i.e., fact triples) that are retrieved at the pre-processing stage but overlook challenges caused by incompleteness of a KG, limited expressiveness of its relations, and retrieved facts irrelevant to the reasoning context. In this paper, we present a novel neural-symbolic model, named Hybrid Graph Network (HGN), which jointly generates feature representations for new triples (as a complement to existing edges in the KG) , determines the relevance of the triples to the reasoning context, and learns graph module parameters for encoding the relational information. Our model learns a compact graph structure(comprising both extracted and generated edges) through filtering edges that are unhelpful to the reasoning process. We show marked improvement on three commonsense reasoning benchmarks and demonstrate the superiority of the learned graph structures with user studies .", "after_revision": "Recently, neural-symbolic models have achieved noteworthy success in leveraging knowledge graphs (KGs) for commonsense reasoning tasks, like question answering (QA) . However, fact sparsity, inherent in human-annotated KGs, can hinder such models from retrieving task-relevant knowledge. To address these issues, we propose Hybrid Graph Network (HGN), a neural-symbolic model that reasons over both extracted (human-labeled) and generated facts within the same learned graph structure. Given a KG subgraph of extracted facts, HGN is jointly trained to generate complementary facts, encode relational information in the resulting \"hybrid\" subgraph, and filter out task-irrelevant facts. We demonstrate HGN's ability to produce contextually pertinent subgraphs by showing considerable performance gains across four commonsense reasoning benchmarks and a user study of fact validness and helpfulness .", "edit_actions": [{"type": "R", "before": "architectures have achieved success on commonsense reasoning through effectively encoding relational structures retrieved from external", "after": "models have achieved noteworthy success in leveraging", "start_char_pos": 26, "end_char_pos": 161, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "and obtained state-of-the-art results in tasks such as (commonsense ) question answering and natural language inference", "after": "for commonsense reasoning tasks, like question answering (QA)", "start_char_pos": 185, "end_char_pos": 304, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "these methods rely on quality and contextualized knowledge structures (i.e., fact triples) that are retrieved at the pre-processing stage but overlook challenges caused by incompleteness of a KG, limited expressiveness of its relations, and retrieved facts irrelevant to the reasoning context. In this paper, we present a novel neural-symbolic model, named", "after": "fact sparsity, inherent in human-annotated KGs, can hinder such models from retrieving task-relevant knowledge. To address these issues, we propose", "start_char_pos": 316, "end_char_pos": 672, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "which jointly generates feature representations for new triples (as a complement to existing edges in the KG) , determines the relevance of the triples to the reasoning context, and learns graph module parameters for encoding the relational information. Our model learns a compact graph structure(comprising both extracted and generated edges) through filtering edges that are unhelpful to the reasoning process. We show marked improvement on three", "after": "a neural-symbolic model that reasons over both extracted (human-labeled) and generated facts within the same learned graph structure. Given a KG subgraph of extracted facts, HGN is jointly trained to generate complementary facts, encode relational information in the resulting \"hybrid\" subgraph, and filter out task-irrelevant facts. We demonstrate HGN's ability to produce contextually pertinent subgraphs by showing considerable performance gains across four", "start_char_pos": 701, "end_char_pos": 1149, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "demonstrate the superiority of the learned graph structures with user studies", "after": "a user study of fact validness and helpfulness", "start_char_pos": 1187, "end_char_pos": 1264, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 306, 609, 954, 1113], "domain": "arxiv"}
{"doc_id": "2011.00416", "revision_depth": 2, "before_revision": "Text style transfer (TST) is an important task in natural language generation (NLG), which aims to control certain attributes in the generated text, such as politeness, emotion, humor, and many others. It has a long history in the field of natural language processing (NLP), but recently it has gained significant attention thanks to the promising performance brought by deep learning models. In this paper, we present a systematic survey of the research done on neural text style transfer . We have collected, summarized, and discussed nearly 70 representative articles since the first neural text style transfer work in 2017. Overall, we have covered the task formulation, existing datasets and subtasks, evaluation metrics, and methods on parallel and non-parallel data. We also provide discussions  a variety of important topics regarding TST, which can shed light on new development in this field . Our curated paper list is at URL", "after_revision": "Text style transfer (TST) is an important task in natural language generation (NLG), which aims to control certain attributes in the generated text, such as politeness, emotion, humor, and many others. It has a long history in the field of natural language processing (NLP), and recently has re-gained significant attention thanks to the promising performance brought by deep neural models. In this paper, we present a systematic survey of the research on neural text style transfer , spanning over 100 representative articles since the first neural text style transfer work in 2017. We discuss the task formulation, existing datasets and subtasks, evaluation , as well as the rich methodologies in the presence of parallel and non-parallel data. We also provide discussions on a variety of important topics regarding the future development of TST . Our curated paper list is at URL", "edit_actions": [{"type": "R", "before": "but recently it has gained", "after": "and recently has re-gained", "start_char_pos": 275, "end_char_pos": 301, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "learning", "after": "neural", "start_char_pos": 376, "end_char_pos": 384, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "done", "after": null, "start_char_pos": 455, "end_char_pos": 459, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": ". We have collected, summarized, and discussed nearly 70", "after": ", spanning over 100", "start_char_pos": 490, "end_char_pos": 546, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Overall, we have covered", "after": "We discuss", "start_char_pos": 628, "end_char_pos": 652, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "metrics, and methods on", "after": ", as well as the rich methodologies in the presence of", "start_char_pos": 718, "end_char_pos": 741, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "on", "start_char_pos": 802, "end_char_pos": 802, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "TST, which can shed light on new development in this field", "after": "the future development of TST", "start_char_pos": 843, "end_char_pos": 901, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 201, 392, 491, 627, 773, 903], "domain": "arxiv"}
{"doc_id": "2011.02593", "revision_depth": 1, "before_revision": "Neural sequence models can generate highly fluent sentences but recent studies have also shown that they are also prone to hallucinate additional content not supported by the input, which can cause a lack of trust in the model. To better assess the faithfulness of the machine outputs, we propose a new task to predict whether each token in the output sequence is hallucinated conditioned on the source input, and collect new manually annotated evaluation sets for this task. We also introduce a novel method for learning to model hallucination detection, based on pretrained language models fine tuned on synthetic data that includes automatically inserted hallucinations. Experiments on machine translation and abstract text summarization demonstrate the effectiveness of our proposed approach -- we obtain an average F1 of around 0.6 across all the benchmark datasets  and achieve significant improvements in sentence-level hallucination scoring compared to baseline methods. We also release our annotated data and code for future researchat URL", "after_revision": "Neural sequence models can generate highly fluent sentences but recent studies have also shown that they are also prone to hallucinate additional content not supported by the input, which can cause a lack of trust in the model. To better assess the faithfulness of the machine outputs, we propose a new task to predict whether each token in the output sequence is hallucinated conditioned on the source input, and collect new manually annotated evaluation sets for this task. We also introduce a novel method for learning to model hallucination detection, based on pretrained language models fine tuned on synthetic data that includes automatically inserted hallucinations. Experiments on machine translation and abstract text summarization demonstrate the effectiveness of our proposed approach -- we obtain an average F1 of around 60 across all the benchmark datasets . Furthermore, we demonstrate how to use the token-level hallucination labels to define a fine-grained loss over the target sequence in the low-resource machine translation and achieve significant improvements over strong baseline methods. We will release our annotated data and code to support future research.", "edit_actions": [{"type": "R", "before": "0.6", "after": "60", "start_char_pos": 833, "end_char_pos": 836, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ". Furthermore, we demonstrate how to use the token-level hallucination labels to define a fine-grained loss over the target sequence in the low-resource machine translation", "start_char_pos": 871, "end_char_pos": 871, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in sentence-level hallucination scoring compared to", "after": "over strong", "start_char_pos": 909, "end_char_pos": 960, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "also", "after": "will", "start_char_pos": 982, "end_char_pos": 986, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "for future researchat URL", "after": "to support future research.", "start_char_pos": 1023, "end_char_pos": 1048, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 227, 475, 673, 978], "domain": "arxiv"}
{"doc_id": "2011.04393", "revision_depth": 2, "before_revision": "Recently, contextualized word embeddings outperform static word embeddings on many NLP tasks. However, we still do not know much about the mechanism inside these representations. Do they have any common patterns? If so, where do these patterns come from? We find that almost all the contextualized word vectors of BERT and RoBERTa have a commonpattern. For BERT, the 557^{th neuron-level method to analyze where these \"tails\" come from. We find that these \"tails\" are closely related to the positional information . We also investigate what will happen if we \"cutting the tails\" (zero-out). Our results show that \"tails\" are the major cause of anisotropy of vector space. After \"cutting the tails\", a word's different vectors are more similar to each other. The internal representations have a better ability to distinguish a word 's different senseswith the word-in-context (WiC) dataset. The performance on the word sense disambiguation task is better for BERT and unchanged for RoBERTa. We can also better induce phrase grammar from the vector space. These suggest that \"tails\" are less related to the sense and syntax information in vectors. These findings provide insights into the inner workings of contextualized word vectors .", "after_revision": "In this work, we demonstrate that the contextualized word vectors derived from pretrained masked language model-based encoders share a common, perhaps undesirable pattern across layers. Namely, we find cases of persistent outlier neurons within BERT and RoBERTa's hidden state vectors that consistently bear the smallest or largest values in said vectors. In an attempt to investigate the source of this information, we introduce a neuron-level analysis method, which reveals that the outliers are closely related to information captured by positional embeddings . We also pre-train the RoBERTa-base models from scratch and find that the outliers disappear without using positional embeddings. These outliers, we find, are the major cause of anisotropy of encoders' raw vector spaces, and clipping them leads to increased similarity across vectors. We demonstrate this in practice by showing that clipped vectors can more accurately distinguish word senses, as well as lead to better sentence embeddings when mean pooling. In three supervised tasks, we find that clipping does not affect the performance .", "edit_actions": [{"type": "R", "before": "Recently, contextualized word embeddings outperform static word embeddings on many NLP tasks. However, we still do not know much about the mechanism inside these representations. Do they have any common patterns? If so, where do these patterns come from? We find that almost all", "after": "In this work, we demonstrate that", "start_char_pos": 0, "end_char_pos": 278, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "of BERT and RoBERTa have a commonpattern. For BERT, the 557^{th", "after": "derived from pretrained masked language model-based encoders share a common, perhaps undesirable pattern across layers. Namely, we find cases of persistent outlier neurons within BERT and RoBERTa's hidden state vectors that consistently bear the smallest or largest values in said vectors. In an attempt to investigate the source of this information, we introduce a", "start_char_pos": 311, "end_char_pos": 374, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "method to analyze where these \"tails\" come from. We find that these \"tails\"", "after": "analysis method, which reveals that the outliers", "start_char_pos": 388, "end_char_pos": 463, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the positional information", "after": "information captured by positional embeddings", "start_char_pos": 487, "end_char_pos": 513, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "investigate what will happen if we \"cutting the tails\" (zero-out). Our results show that \"tails\"", "after": "pre-train the RoBERTa-base models from scratch and find that the outliers disappear without using positional embeddings. These outliers, we find,", "start_char_pos": 524, "end_char_pos": 620, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "vector space. After \"cutting the tails\", a word's different vectors are more similar to each other. The internal representations have a better ability to distinguish a word 's different senseswith the word-in-context (WiC) dataset. The performance on the word sense disambiguation task is better for BERT and unchanged for RoBERTa. We can also better induce phrase grammar from the vector space. These suggest that \"tails\" are less related to the sense and syntax information in vectors. These findings provide insights into the inner workings of contextualized word vectors", "after": "encoders' raw vector spaces, and clipping them leads to increased similarity across vectors. We demonstrate this in practice by showing that clipped vectors can more accurately distinguish word senses, as well as lead to better sentence embeddings when mean pooling. In three supervised tasks, we find that clipping does not affect the performance", "start_char_pos": 658, "end_char_pos": 1232, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 93, 178, 212, 254, 352, 436, 515, 590, 671, 757, 889, 989, 1053, 1145]}
{"doc_id": "2011.06174", "revision_depth": 1, "before_revision": "Discovering precise and specific rules from knowledge graphs is regarded as an essential challenge, which can improve the performances of many downstream tasks and even provide new ways to approach some Natural Language Processing research topics. In this paper, we provide a fundamental theory for knowledge graph reasoning based on  ending anchored rules. Our theory provides precise reasons answering why or why not a triple is correct. Then, we implement our theory by what we called the EARDict model. Results show that the EARDict model achieves new state-of-the-art performances on benchmark knowledge graph completion tasks, including a Hits@10 score of 80.38 percent on WN18RR.", "after_revision": "Discovering precise and specific rules from knowledge graphs is regarded as an essential challenge, which can improve the performances of many downstream tasks and even provide new ways to approach some Natural Language Processing research topics. In this paper, we provide a fundamental theory for knowledge graph reasoning based on the ending anchored rules. Our theory provides precise reasons explaining why or why not a triple is correct. Then, we implement our theory by what we call the EARDict model. Results show that our EARDict model significantly outperforms all the benchmark models on two large datasets of knowledge graph completion , including achieving a Hits@10 score of 96.6 percent on WN18RR.", "edit_actions": [{"type": "A", "before": null, "after": "the", "start_char_pos": 334, "end_char_pos": 334, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "answering", "after": "explaining", "start_char_pos": 394, "end_char_pos": 403, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "called", "after": "call", "start_char_pos": 481, "end_char_pos": 487, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "the EARDict model achieves new state-of-the-art performances on benchmark", "after": "our EARDict model significantly outperforms all the benchmark models on two large datasets of", "start_char_pos": 525, "end_char_pos": 598, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "tasks, including", "after": ", including achieving", "start_char_pos": 626, "end_char_pos": 642, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "80.38", "after": "96.6", "start_char_pos": 662, "end_char_pos": 667, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 247, 357, 439, 506], "domain": "arxiv"}
{"doc_id": "2011.06174", "revision_depth": 2, "before_revision": "Discovering precise and specific rules from knowledge graphs is regarded as an essential challenge, which can improve the performances of many downstream tasks and even provide new ways to approach some Natural Language Processing research topics. In this paper, we provide a fundamental theory for knowledge graph reasoning based on the ending anchored rules. Our theory provides precise reasons explaining why or why not a triple is correct. Then, we implement our theory by what we call the EARDict model. Results show that our EARDict model significantly outperforms all the benchmark models on two large datasets of knowledge graph completion , including achieving a Hits@10 score of 96.6 percent on WN18RR.", "after_revision": "Discovering precise and specific rules from knowledge graphs is regarded as an essential challenge, which can improve the performances of many downstream tasks and even provide new ways to approach some Natural Language Processing research topics. In this paper, we provide a fundamental theory for knowledge graph reasoning based on the ending anchored rules. Our theory provides precise reasons explaining why or why not a triple is correct. Then, we implement our theory by what we call the EARDict model. Results show that our EARDict model significantly outperforms all the benchmark models on three large datasets of knowledge graph completion . Especially, our model achieves a Hits@10 score of 96.6 percent on WN18RR.", "edit_actions": [{"type": "R", "before": "two", "after": "three", "start_char_pos": 599, "end_char_pos": 602, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", including achieving", "after": ". Especially, our model achieves", "start_char_pos": 648, "end_char_pos": 669, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 247, 360, 443, 508], "domain": "arxiv"}
{"doc_id": "2011.10896", "revision_depth": 1, "before_revision": "Hardware-agnostic programming with high performance portability will be the bedrock for realizing the ubiquitous adoption of emerging accelerator technologies in future heterogeneous high-performance computing (HPC) systems, which is the key to achieving the next level of HPC performance on an expanding accelerator landscape. In this paper, we present HALO 1.0, an open-ended extensible multi-agent software framework , that implements a set of proposed hardware-agnostic accelerator orchestration (HALO) principles and a novel compute-centric message passing interface (C^2MPI) specification for enabling the portable and performance-optimized execution of hardware-agnostic application  codes across heterogeneous accelerator resources. The experiment results of evaluating eight widely used HPC subroutines based on Intel Xeon E5-2620 v4 CPUs, Intel Arria 10 GX FPGAs, and NVIDIA GeForce RTX 2080 Ti GPUs show that HALO 1.0 allows the same hardware-agnostic application codes of the HPC kernels, without any change, to run across all the computing devices with a consistently maximum performance portability score of 1.0, which is 2x-861,883x higher than the OpenCL-based solution that suffers from an unstably low performance portability score  .", "after_revision": "Hardware-agnostic programming with high performance portability will be the bedrock for realizing the ubiquitous adoption of emerging accelerator technologies in future heterogeneous high-performance computing (HPC) systems, which is the key to achieving the next level of HPC performance on an expanding accelerator landscape. In this paper, we present HALO 1.0, an open-ended extensible multi-agent software framework that implements a set of proposed hardware-agnostic accelerator orchestration (HALO) principles and a novel compute-centric message passing interface (C^2MPI) specification for enabling the portable and performance-optimized execution of hardware-agnostic application host codes across heterogeneous accelerator resources. The experiment results of evaluating eight widely used HPC subroutines based on Intel Xeon E5-2620 v4 CPUs, Intel Arria 10 GX FPGAs, and NVIDIA GeForce RTX 2080 Ti GPUs show that HALO 1.0 allows for a unified control flow for the host program to run across all the computing devices with a consistently maximum performance portability score of 1.0, which is 2x-861,883x higher than the OpenCL-based solution that suffers from an unstably low performance portability score . of the documentation of their work .", "edit_actions": [{"type": "D", "before": ",", "after": null, "start_char_pos": 420, "end_char_pos": 421, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "host", "start_char_pos": 690, "end_char_pos": 690, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "the same hardware-agnostic application codes of the HPC kernels, without any change,", "after": "for a unified control flow for the host program", "start_char_pos": 936, "end_char_pos": 1020, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ". of the documentation of their work", "start_char_pos": 1250, "end_char_pos": 1250, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}], "sents_char_pos": [0, 327, 740, 848], "domain": "arxiv"}
{"doc_id": "2011.11465", "revision_depth": 1, "before_revision": "Many online comments on social media platforms are hateful , humorous, or sarcastic. The sarcastic nature of these comments (especially the short ones) alters their actual implied sentiments , which leads to misinterpretations by the existing sentiment analysis models. A lot of research has already been done to detect sarcasm in the text using user-based, topical, and conversational information but not much work has been done to use inter-sentence contextual information for detecting the same. This paper proposes a new state-of-the-art deep learning architecture that uses a novel Bidirectional Inter-Sentence Contextual Attention mechanism(Bi-ISCA) to capture inter-sentence dependencies for detecting sarcasm in the user-generated short text using only the conversational context . The proposed deep learning model demonstrates the capability to capture explicit, implicit, and contextual incongruous words phrases responsible for invoking sarcasm . Bi-ISCA generates state-of-the-art results on two widely used benchmark datasets for the sarcasm detection task (Reddit and Twitter). To the best of our knowledge, none of the existing state-of-the-art models use an inter-sentence contextual attention mechanism to detect sarcasm in the user-generated short text using only conversational context .", "after_revision": "Sentiment analysis of social media comments is very important for review analysis. Many online reviews are sarcastic , humorous, or hateful. This sarcastic nature of these short texts change the actual sentiments of the review as predicted by a machine learning model that attempts to detect sentiment alone. Thus, having a model that is explicitly aware of these features should help it perform better on reviews that are characterized by them. Several research has already been done in this field. This paper deals with sarcasm detection on reddit comments. Several machine learning and deep learning algorithms have been applied for the same but each of these models only take into account the initial text instead of the conversation which serves as a better measure to determine sarcasm. The other shortcoming these papers have is they rely on word embedding for representing comments and thus do not take into account the problem of polysemy(A word can have multiple meanings based on the context in which it appears). These existing modules were able to solve the problem of capturing inter sentence contextual information but not the intra sentence contextual information. So we propose a novel architecture which solves the problem of sarcasm detection by capturing intra sentence contextual information using a novel contextual attention mechanism . The proposed model solves the problem of polysemy also by using context enriched language modules like ELMO and BERT in its first component. This model comprises a total of three major components which takes into account inter sentence, intra sentence contextual information and at last use a convolutional neural network for capturing global contextual information for sarcasm detection. The proposed model was able to generate decent results and cleared showed potential to perform state of the art if trained on a larger dataset .", "edit_actions": [{"type": "R", "before": "Many online comments on social media platforms are hateful", "after": "Sentiment analysis of social media comments is very important for review analysis. Many online reviews are sarcastic", "start_char_pos": 0, "end_char_pos": 58, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "sarcastic. The", "after": "hateful. This", "start_char_pos": 74, "end_char_pos": 88, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "comments (especially the short ones) alters their actual implied sentiments , which leads to misinterpretations by the existing sentiment analysis models. A lot of", "after": "short texts change the actual sentiments of the review as predicted by a machine learning model that attempts to detect sentiment alone. Thus, having a model that is explicitly aware of these features should help it perform better on reviews that are characterized by them. Several", "start_char_pos": 115, "end_char_pos": 278, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "to detect sarcasm in the text using user-based, topical, and conversational information but not much work has been done to use inter-sentence contextual information for detecting the same. This paper proposes a new state-of-the-art deep learning architecture that uses a novel Bidirectional Inter-Sentence Contextual Attention mechanism(Bi-ISCA) to capture inter-sentence dependencies for detecting sarcasm in the user-generated short text using only the conversational context", "after": "in this field. This paper deals with sarcasm detection on reddit comments. Several machine learning and deep learning algorithms have been applied for the same but each of these models only take into account the initial text instead of the conversation which serves as a better measure to determine sarcasm. The other shortcoming these papers have is they rely on word embedding for representing comments and thus do not take into account the problem of polysemy(A word can have multiple meanings based on the context in which it appears). These existing modules were able to solve the problem of capturing inter sentence contextual information but not the intra sentence contextual information. So we propose a novel architecture which solves the problem of sarcasm detection by capturing intra sentence contextual information using a novel contextual attention mechanism", "start_char_pos": 310, "end_char_pos": 787, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "deep learning model demonstrates the capability to capture explicit, implicit, and contextual incongruous words", "after": null, "start_char_pos": 803, "end_char_pos": 914, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "phrases responsible for invoking sarcasm . Bi-ISCA generates state-of-the-art results on two widely used benchmark datasets for the sarcasm detection task (Reddit and Twitter). To the best of our knowledge, none of the existing state-of-the-art models use an inter-sentence contextual attention mechanism to detect sarcasm in the user-generated short text using only conversational context", "after": "model solves the problem of polysemy also by using context enriched language modules like ELMO and BERT in its first component. This model comprises a total of three major components which takes into account inter sentence, intra sentence contextual information and at last use a convolutional neural network for capturing global contextual information for sarcasm detection. The proposed model was able to generate decent results and cleared showed potential to perform state of the art if trained on a larger dataset", "start_char_pos": 915, "end_char_pos": 1304, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 84, 269, 498, 789, 1091], "domain": "arxiv"}
{"doc_id": "2011.11465", "revision_depth": 2, "before_revision": "Sentiment analysis of social media comments is very important for review analysis. Many online reviews are sarcastic , humorous, or hateful. This sarcastic nature of these short texts change the actual sentiments of the review as predicted by a machine learning model that attempts to detect sentiment alone. Thus, having a model that is explicitly aware of these features should help it perform better on reviews that are characterized by them. Several research has already been done in this field. This paper deals with sarcasm detection on reddit comments. Several machine learning and deep learning algorithms have been applied for the same but each of these models only take into account the initial text instead of the conversation which serves as a better measure to determine sarcasm . The other shortcoming these papers have is they rely on word embedding for representing comments and thus do not take into account the problem of polysemy(A word can have multiple meanings based on the context in which it appears). These existing modules were able to solve the problem of capturing inter sentence contextual information but not the intra sentence contextual information . So we propose a novel architecture which solves the problem of sarcasm detection by capturing intra sentence contextual information using a novel contextual attention mechanism . The proposed model solves the problem of polysemy also by using context enriched language modules like ELMO and BERT in its first component. This model comprises a total of three major components which takes into account inter sentence, intra sentence contextual information and at last use a convolutional neural network for capturing global contextual information for sarcasm detection. The proposed model was able to generate decent results and cleared showed potential to perform state of the art if trained on a larger dataset  .", "after_revision": "Many online comments on social media platforms are hateful , humorous, or sarcastic. The sarcastic nature of these comments (especially the short ones) alters their actual implied sentiments, which leads to misinterpretations by the existing sentiment analysis models. A lot of research has already been done to detect sarcasm in the text using user-based, topical, and conversational information but not much work has been done to use inter-sentence contextual information for detecting the same. This paper proposes a new state-of-the-art deep learning architecture that uses a novel Bidirectional Inter-Sentence Contextual Attention mechanism (Bi-ISCA) to capture inter-sentence dependencies for detecting sarcasm in the user-generated short text using only the conversational context . The proposed deep learning model demonstrates the capability to capture explicit, implicit, and contextual incongruous words phrases responsible for invoking sarcasm. Bi-ISCA generates state-of-the-art results on two widely used benchmark datasets for the sarcasm detection task (Reddit and Twitter). To the best of our knowledge, none of the existing state-of-the-art models use an inter-sentence contextual attention mechanism to detect sarcasm in the user-generated short text using only conversational context .", "edit_actions": [{"type": "R", "before": "Sentiment analysis of social media comments is very important for review analysis. Many online reviews are sarcastic", "after": "Many online comments on social media platforms are hateful", "start_char_pos": 0, "end_char_pos": 116, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "hateful. This", "after": "sarcastic. The", "start_char_pos": 132, "end_char_pos": 145, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "short texts change the actual sentiments of the review as predicted by a machine learning model that attempts to detect sentiment alone. Thus, having a model that is explicitly aware of these features should help it perform better on reviews that are characterized by them. Several", "after": "comments (especially the short ones) alters their actual implied sentiments, which leads to misinterpretations by the existing sentiment analysis models. A lot of", "start_char_pos": 172, "end_char_pos": 453, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "in this field. This paper deals with sarcasm detection on reddit comments. Several machine learning and deep learning algorithms have been applied for the same but each of these models only take into account the initial text instead of the conversation which serves as a better measure to determine sarcasm . The other shortcoming these papers have is they rely on word embedding for representing comments and thus do not take into account the problem of polysemy(A word can have multiple meanings based on the context in which it appears). These existing modules were able to solve the problem of capturing inter sentence contextual", "after": "to detect sarcasm in the text using user-based, topical, and conversational", "start_char_pos": 485, "end_char_pos": 1118, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "the intra sentence contextual information", "after": "much work has been done to use inter-sentence contextual information for detecting the same. This paper proposes a new state-of-the-art deep learning architecture that uses a novel Bidirectional Inter-Sentence Contextual Attention mechanism (Bi-ISCA) to capture inter-sentence dependencies for detecting sarcasm in the user-generated short text using only the conversational context", "start_char_pos": 1139, "end_char_pos": 1180, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "So we propose a novel architecture which solves the problem of sarcasm detection by capturing intra sentence contextual information using a novel contextual attention mechanism . The proposed model solves the problem of polysemy also by using context enriched language modules like ELMO and BERT in its first component. This model comprises a total of three major components which takes into account inter sentence, intra sentence contextual information and at last use a convolutional neural network for capturing global contextual information for sarcasm detection. The proposed model was able to generate decent results and cleared showed potential to perform state of the art if trained on a larger dataset", "after": "The proposed deep learning model demonstrates the capability to capture explicit, implicit, and contextual incongruous words", "start_char_pos": 1183, "end_char_pos": 1893, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "phrases responsible for invoking sarcasm. Bi-ISCA generates state-of-the-art results on two widely used benchmark datasets for the sarcasm detection task (Reddit and Twitter). To the best of our knowledge, none of the existing state-of-the-art models use an inter-sentence contextual attention mechanism to detect sarcasm in the user-generated short text using only conversational context", "start_char_pos": 1894, "end_char_pos": 1894, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 82, 140, 308, 445, 499, 559, 793, 1025, 1182, 1361, 1502, 1750], "domain": "arxiv"}
{"doc_id": "2012.08987", "revision_depth": 1, "before_revision": "Discovering new intents is a crucial task in a dialogue system . Most existing methods are limited in transferring the prior knowledge from known intents to new intents. These methods also have difficulties in providing high-quality supervised signals to learn clustering-friendly features for grouping unlabeled intents. In this work, we propose an effective method (Deep Aligned Clustering) to discover new intents with the aid of limited known intent data. Firstly, we leverage a few labeled known intent samples as prior knowledge to pre-train the model. Then, we perform k-means to produce cluster assignments as pseudo-labels. Moreover, we propose an alignment strategy to tackle the label inconsistency  during clustering assignments. Finally, we learn the intent representations under the supervision of the aligned pseudo-labels. With an unknown number of new intents, we predict the number of intent categories by eliminating low-confidence intent-wise clusters. Extensive experiments on two benchmark datasets show that our method is more robust and achieves substantial improvements over the state-of-the-art methods.( Code available at URL", "after_revision": "Discovering new intents is a crucial task in dialogue systems . Most existing methods are limited in transferring the prior knowledge from known intents to new intents. These methods also have difficulties in providing high-quality supervised signals to learn clustering-friendly features for grouping unlabeled intents. In this work, we propose an effective method (Deep Aligned Clustering) to discover new intents with the aid of limited known intent data. Firstly, we leverage a few labeled known intent samples as prior knowledge to pre-train the model. Then, we perform k-means to produce cluster assignments as pseudo-labels. Moreover, we propose an alignment strategy to tackle the label inconsistency problem during clustering assignments. Finally, we learn the intent representations under the supervision of the aligned pseudo-labels. With an unknown number of new intents, we predict the number of intent categories by eliminating low-confidence intent-wise clusters. Extensive experiments on two benchmark datasets show that our method is more robust and achieves substantial improvements over the state-of-the-art methods.( The code is available at URL", "edit_actions": [{"type": "R", "before": "a dialogue system", "after": "dialogue systems", "start_char_pos": 45, "end_char_pos": 62, "major_intent": "style", "raw_intents": ["style", "style", "fluency"]}, {"type": "A", "before": null, "after": "problem", "start_char_pos": 710, "end_char_pos": 710, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Code", "after": "The code is", "start_char_pos": 1131, "end_char_pos": 1135, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 64, 169, 321, 459, 558, 632, 741, 838, 972, 1129], "domain": "arxiv"}
{"doc_id": "2012.15859", "revision_depth": 1, "before_revision": "Natural Language Processing (NLP) systems learn harmful societal biases that cause them to extend and proliferate inequality widely, as they are deployed in more and more situations. To address and combat this, the NLP community has come to rely on a variety of metrics to identify and quantify bias in black-box models , which are used to monitor model behaviour and to guide efforts at debiasing. Some of these metrics are intrinsic, and are measured in word embedding spaces, and some are extrinsic, which measure the bias present downstream in the tasks that the word embeddings are plugged into. This research examines whether intrinsic metrics (which are easy to measure) correlate well to extrinsic metrics (which reflect real world bias) . We measure both intrinsic and extrinsic bias across hundreds of trained models covering different tasks and experimental conditions and find that there is no reliable correlation between these metrics that holds in more than extremely specific settings . We advise that efforts to debias embedding spaces be always also paired with measurement of downstream model bias, and suggest that that community direct more effort into making downstream measurement simpler and easier .", "after_revision": "Natural Language Processing (NLP) systems learn harmful societal biases that cause them to widely proliferate inequality as they are deployed in more and more situations. To address and combat this, the NLP community relies on a variety of metrics to identify and quantify bias in black-box models and to guide efforts at debiasing. Some of these metrics are intrinsic, and are measured in word embedding spaces, and some are extrinsic, which measure the bias present downstream in the tasks that the word embeddings are plugged into. This research examines whether easy-to-measure intrinsic metrics correlate well to real world extrinsic metrics . We measure both intrinsic and extrinsic bias across hundreds of trained models covering different tasks and experimental conditions and find that there is no reliable correlation between these metrics that holds in all scenarios across tasks and languages . We advise that efforts to debias embedding spaces be always also paired with measurement of downstream model bias, and suggest that that community increase effort into making downstream measurement more feasible via creation of additional challenge sets and annotated test data. We additionally release code, a new intrinsic metric, and an annotated test set for gender bias for hatespeech .", "edit_actions": [{"type": "R", "before": "extend and proliferate inequality widely,", "after": "widely proliferate inequality", "start_char_pos": 91, "end_char_pos": 132, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "has come to rely", "after": "relies", "start_char_pos": 229, "end_char_pos": 245, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "D", "before": ", which are used to monitor model behaviour", "after": null, "start_char_pos": 320, "end_char_pos": 363, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "intrinsic metrics (which are easy to measure)", "after": "easy-to-measure intrinsic metrics", "start_char_pos": 632, "end_char_pos": 677, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "extrinsic metrics (which reflect real world bias)", "after": "real world extrinsic metrics", "start_char_pos": 696, "end_char_pos": 745, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "more than extremely specific settings", "after": "all scenarios across tasks and languages", "start_char_pos": 963, "end_char_pos": 1000, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "direct more", "after": "increase", "start_char_pos": 1150, "end_char_pos": 1161, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "simpler and easier", "after": "more feasible via creation of additional challenge sets and annotated test data. We additionally release code, a new intrinsic metric, and an annotated test set for gender bias for hatespeech", "start_char_pos": 1204, "end_char_pos": 1222, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 182, 398, 600, 747, 1002], "domain": "arxiv"}
{"doc_id": "2101.01321", "revision_depth": 2, "before_revision": "Transformer based models, like BERT and RoBERTa, have achieved state-of-the-art results in many Natural Language Processing tasks. However, their memory footprint, inference latency, and power consumption are prohibitive for efficient inference at the edge, and even at the data center. While quantization can be a viable solution for this, previous work on quantizing Transformer based models use floating-point arithmetic during inference, which cannot efficiently utilize integer-only logical units such as the recent Turing Tensor Cores, or traditional integer-only ARM processors. In this work, we propose I-BERT, a novel quantization scheme for Transformer based models that quantizes the entire inference with integer-only arithmetic. Based on lightweight integer-only approximation methods for nonlinear operations, e.g., GELU, Softmax, and Layer Normalization, I-BERT performs an end-to-end integer-only BERT inference without any floating point calculation. We evaluate our approach on GLUE downstream tasks using RoBERTa-Base/Large. We show that for both cases, I-BERT achieves similar (and slightly higher) accuracy as compared to the full-precision baseline. Furthermore, our preliminary implementation of I-BERT shows a speedup of 2.4 - 4.0 x for INT8 inference on a T4 GPU system as compared to FP32 inference. The framework has been developed in PyTorch and has been open-sourced.", "after_revision": "Transformer based models, like BERT and RoBERTa, have achieved state-of-the-art results in many Natural Language Processing tasks. However, their memory footprint, inference latency, and power consumption are prohibitive efficient inference at the edge, and even at the data center. While quantization can be a viable solution for this, previous work on quantizing Transformer based models use floating-point arithmetic during inference, which cannot efficiently utilize integer-only logical units such as the recent Turing Tensor Cores, or traditional integer-only ARM processors. In this work, we propose I-BERT, a novel quantization scheme for Transformer based models that quantizes the entire inference with integer-only arithmetic. Based on lightweight integer-only approximation methods for nonlinear operations, e.g., GELU, Softmax, and Layer Normalization, I-BERT performs an end-to-end integer-only BERT inference without any floating point calculation. We evaluate our approach on GLUE downstream tasks using RoBERTa-Base/Large. We show that for both cases, I-BERT achieves similar (and slightly higher) accuracy as compared to the full-precision baseline. Furthermore, our preliminary implementation of I-BERT shows a speedup of 2.4 -4.0 x for INT8 inference on a T4 GPU system as compared to FP32 inference. The framework has been developed in PyTorch and has been open-sourced.", "edit_actions": [{"type": "D", "before": "for", "after": null, "start_char_pos": 221, "end_char_pos": 224, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "fluency"]}, {"type": "R", "before": "- 4.0", "after": "-4.0", "start_char_pos": 1249, "end_char_pos": 1254, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 130, 286, 585, 741, 967, 1043, 1171, 1325], "domain": "arxiv"}
{"doc_id": "2102.07594", "revision_depth": 1, "before_revision": "Attention-based encoder-decoder (AED) models have achieved promising performance in speech recognition. However, because the decoder predicts text tokens (such as characters or words) in an autoregressive manner, it is difficult for an AED model to predict all tokens in parallel. This makes the inference speed relatively slow. We believe that because the encoder already captures the whole speech utterance, which has the token-level relationship implicitly, we can predict a token without explicitly autoregressive language modeling. When the prediction of a token does not rely on other tokens, the parallel prediction of all tokens in the sequence is realizable. Based on this idea, we propose a non-autoregressive speech recognition model called LASO (Listen Attentively, and Spell Once). The model consists of an encoder, a decoder, and a position dependent summarizer (PDS). The three modules are based on basic attention blocks. The encoder extracts high-level representations from the speech. The PDS uses positional encodings corresponding to tokensto convert the acoustic representations into token-level representations. The decoder further captures token-level relationships with the self-attention mechanism. At last, the probability distribution on the vocabulary is computed for each token position. Therefore, speech recognition is re-formulated as a position-wise classification problem. Further, we propose a cross-modal transfer learning method to refine semantics from a large-scale pre-trained language model BERT for improving the performance .", "after_revision": "Attention-based encoder-decoder (AED) models have achieved promising performance in speech recognition. However, because the decoder predicts text tokens (such as characters or words) in an autoregressive manner, it is difficult for an AED model to predict all tokens in parallel. This makes the inference speed relatively slow. In contrast, we propose an end-to-end non-autoregressive speech recognition model called LASO (Listen Attentively, and Spell Once). The model aggregates encoded speech features into the hidden representations corresponding to each token with attention mechanisms. Thus, the model can capture the token relations by self-attention on the aggregated hidden representations from the whole speech signal rather than autoregressive modeling on tokens. Without explicitly autoregressive language modeling, this model predicts all tokens in the sequence in parallel so that the inference is efficient. Moreover, we propose a cross-modal transfer learning method to use a text-modal language model to improve the performance of speech-modal LASO by aligning token semantics. We conduct experiments on two scales of public Chinese speech datasets AISHELL-1 and AISHELL-2. Experimental results show that our proposed model achieves a speedup of about 50\\times and competitive performance, compared with the autoregressive transformer models. And the cross-modal knowledge transferring from the text-modal model can improve the performance of the speech-modal model .", "edit_actions": [{"type": "R", "before": "We believe that because the encoder already captures the whole speech utterance, which has the token-level relationship implicitly, we can predict a token without explicitly autoregressive language modeling. When the prediction of a token does not rely on other tokens, the parallel prediction of all tokens in the sequence is realizable. Based on this idea, we propose a", "after": "In contrast, we propose an end-to-end", "start_char_pos": 329, "end_char_pos": 700, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "consists of an encoder, a decoder, and a position dependent summarizer (PDS). The three modules are based on basic attention blocks. The encoder extracts high-level", "after": "aggregates encoded speech features into the hidden representations corresponding to each token with attention mechanisms. Thus, the model can capture the token relations by self-attention on the aggregated hidden", "start_char_pos": 805, "end_char_pos": 969, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "speech. The PDS uses positional encodings corresponding to tokensto convert the acoustic representations into token-level representations. The decoder further captures token-level relationships with the self-attention mechanism. At last, the probability distribution on the vocabulary is computed for each token position. Therefore, speech recognition is re-formulated as a position-wise classification problem. Further,", "after": "whole speech signal rather than autoregressive modeling on tokens. Without explicitly autoregressive language modeling, this model predicts all tokens in the sequence in parallel so that the inference is efficient. Moreover,", "start_char_pos": 995, "end_char_pos": 1415, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "refine semantics from a large-scale pre-trained language model BERT for improving the performance", "after": "use a text-modal language model to improve the performance of speech-modal LASO by aligning token semantics. We conduct experiments on two scales of public Chinese speech datasets AISHELL-1 and AISHELL-2. Experimental results show that our proposed model achieves a speedup of about 50\\times and competitive performance, compared with the autoregressive transformer models. And the cross-modal knowledge transferring from the text-modal model can improve the performance of the speech-modal model", "start_char_pos": 1469, "end_char_pos": 1566, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 103, 280, 328, 536, 667, 794, 882, 937, 1002, 1133, 1223, 1316, 1406]}
{"doc_id": "2102.07594", "revision_depth": 2, "before_revision": "Attention-based encoder-decoder (AED) models have achieved promising performance in speech recognition. However, because the decoder predicts text tokens (such as characters or words) in an autoregressive manner, it is difficult for an AED model to predict all tokens in parallel. This makes the inference speed relatively slow. In contrast, we propose an end-to-end non-autoregressive speech recognition model called LASO (Listen Attentively, and Spell Once). The model aggregates encoded speech features into the hidden representations corresponding to each token with attention mechanisms. Thus, the model can capture the token relations by self-attention on the aggregated hidden representations from the whole speechsignal rather than autoregressive modeling on tokens . Without explicitly autoregressive language modeling, this model predicts all tokens in the sequence in parallel so that the inference is efficient. Moreover, we propose a cross-modal transfer learning method to use a text-modal language model to improve the performanceof speech-modal LASO by aligning token semantics. We conduct experiments on two scales of public Chinese speech datasets AISHELL-1 and AISHELL-2. Experimental results show that our proposed model achieves a speedup of about 50\\times and competitive performance, compared with the autoregressive transformer models. And the cross-modal knowledge transferring from the text-modal model can improve the performance of the speech-modal model .", "after_revision": "Attention-based encoder-decoder (AED) models have achieved promising performance in speech recognition. However, because the decoder predicts text tokens (such as characters or words) in an autoregressive manner, it is difficult for an AED model to predict all tokens in parallel. This makes the inference speed relatively slow. We believe that because the encoder already captures the whole speech utterance, which has the token-level relationship implicitly, we can predict a token without explicitly autoregressive language modeling. When the prediction of a token does not rely on other tokens, the parallel prediction of all tokens in the sequence is realizable. Based on this idea, we propose a non-autoregressive speech recognition model called LASO (Listen Attentively, and Spell Once). The model consists of an encoder, a decoder, and a position dependent summarizer (PDS). The three modules are based on basic attention blocks. The encoder extracts high-level representations from the speech. The PDS uses positional encodings corresponding to tokens to convert the acoustic representations into token-level representations. The decoder further captures token-level relationships with the self-attention mechanism. At last, the probability distribution on the vocabulary is computed for each token position. Therefore, speech recognition is re-formulated as a position-wise classification problem. Further, we propose a cross-modal transfer learning method to refine semantics from a large-scale pre-trained language model BERT for improving the performance .", "edit_actions": [{"type": "R", "before": "In contrast, we propose an end-to-end", "after": "We believe that because the encoder already captures the whole speech utterance, which has the token-level relationship implicitly, we can predict a token without explicitly autoregressive language modeling. When the prediction of a token does not rely on other tokens, the parallel prediction of all tokens in the sequence is realizable. Based on this idea, we propose a", "start_char_pos": 329, "end_char_pos": 366, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "aggregates encoded speech features into the hidden representations corresponding to each token with attention mechanisms. Thus, the model can capture the token relations by self-attention on the aggregated hidden", "after": "consists of an encoder, a decoder, and a position dependent summarizer (PDS). The three modules are based on basic attention blocks. The encoder extracts high-level", "start_char_pos": 471, "end_char_pos": 683, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "whole speechsignal rather than autoregressive modeling on tokens . Without explicitly autoregressive language modeling, this model predicts all tokens in the sequence in parallel so that the inference is efficient. Moreover,", "after": "speech. The PDS uses positional encodings corresponding to tokens to convert the acoustic representations into token-level representations. The decoder further captures token-level relationships with the self-attention mechanism. At last, the probability distribution on the vocabulary is computed for each token position. Therefore, speech recognition is re-formulated as a position-wise classification problem. Further,", "start_char_pos": 709, "end_char_pos": 933, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "use a text-modal language model to improve the performanceof speech-modal LASO by aligning token semantics. We conduct experiments on two scales of public Chinese speech datasets AISHELL-1 and AISHELL-2. Experimental results show that our proposed model achieves a speedup of about 50\\times and competitive performance, compared with the autoregressive transformer models. And the cross-modal knowledge transferring from the text-modal model can improve the performance of the speech-modal model", "after": "refine semantics from a large-scale pre-trained language model BERT for improving the performance", "start_char_pos": 987, "end_char_pos": 1482, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 103, 280, 328, 460, 592, 775, 923, 1094, 1190, 1359]}
{"doc_id": "2103.06874", "revision_depth": 2, "before_revision": "Pipelined NLP systems have largely been superseded by end-to-end neural modeling, yet nearly all commonly-used models still require an explicit tokenization step. While recent tokenization approaches based on data-derived subword lexicons are less brittle than manually engineered tokenizers, these techniques are not equally suited to all languages, and the use of any fixed vocabulary may limit a model's ability to adapt. In this paper, we present CANINE, a neural encoder that operates directly on character sequences, without explicit tokenization or vocabulary, and a pre-training strategy with soft inductive biases in place of hard token boundaries . To use its finer-grained input effectively and efficiently, CANINE combines downsampling, which reduces the input sequence length, with a deep transformer stack, which encodes context. CANINE outperforms a comparable mBERT model by >= 1 F1 on TyDi QA, a challenging multilingual benchmark, despite having 28\\% fewer model parameters.", "after_revision": "Pipelined NLP systems have largely been superseded by end-to-end neural modeling, yet nearly all commonly-used models still require an explicit tokenization step. While recent tokenization approaches based on data-derived subword lexicons are less brittle than manually engineered tokenizers, these techniques are not equally suited to all languages, and the use of any fixed vocabulary may limit a model's ability to adapt. In this paper, we present CANINE, a neural encoder that operates directly on character sequences, without explicit tokenization or vocabulary, and a pre-training strategy that operates either directly on characters or optionally uses subwords as a soft inductive bias . To use its finer-grained input effectively and efficiently, CANINE combines downsampling, which reduces the input sequence length, with a deep transformer stack, which encodes context. CANINE outperforms a comparable mBERT model by 2.8 F1 on TyDi QA, a challenging multilingual benchmark, despite having 28\\% fewer model parameters.", "edit_actions": [{"type": "R", "before": "with soft inductive biases in place of hard token boundaries", "after": "that operates either directly on characters or optionally uses subwords as a soft inductive bias", "start_char_pos": 596, "end_char_pos": 656, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ">= 1", "after": "2.8", "start_char_pos": 891, "end_char_pos": 895, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 162, 424, 658, 843], "domain": "arxiv"}
{"doc_id": "2103.06922", "revision_depth": 1, "before_revision": "Recent studies indicate that NLU models are prone to rely on shortcut features for prediction  . As a result, these models could potentially fail to generalize to real-world out-of-distribution scenarios . In this work, we show that the shortcut learning behavior can be explained by the long-tailed phenomenon . There are two findings: 1) Trained NLU models have strong preference for features located at the head of the long-tailed distribution, and 2) Shortcut features are picked up during very early few iterations of the model training. These two observations are further employed to formulate a measurement which can quantify the shortcut degree of each training sample. Based on this shortcut measurement, we propose a shortcut mitigation framework  , to suppress the model from making overconfident predictions for samples with large shortcut degree. Experimental results on three NLU benchmarks demonstrate that our long-tailed distribution explanation accurately reflects the shortcut learning behavior of NLU models. Experimental analysis further indicates that our method can improve the generalization accuracy on OOD data, while preserving the accuracy on in distribution test data.", "after_revision": "Recent studies indicate that NLU models are prone to rely on shortcut features for prediction , without achieving true language understanding . As a result, these models fail to generalize to real-world out-of-distribution data . In this work, we show that the words in the NLU training set can be modeled as a long-tailed distribution . There are two findings: 1) NLU models have strong preference for features located at the head of the long-tailed distribution, and 2) Shortcut features are picked up during very early few iterations of the model training. These two observations are further employed to formulate a measurement which can quantify the shortcut degree of each training sample. Based on this shortcut measurement, we propose a shortcut mitigation framework LGTR , to suppress the model from making overconfident predictions for samples with large shortcut degree. Experimental results on three NLU benchmarks demonstrate that our long-tailed distribution explanation accurately reflects the shortcut learning behavior of NLU models. Experimental analysis further indicates that LGTR can improve the generalization accuracy on OOD data, while preserving the accuracy on in-distribution data.", "edit_actions": [{"type": "A", "before": null, "after": ", without achieving true language understanding", "start_char_pos": 94, "end_char_pos": 94, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "D", "before": "could potentially", "after": null, "start_char_pos": 123, "end_char_pos": 140, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "scenarios", "after": "data", "start_char_pos": 194, "end_char_pos": 203, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "shortcut learning behavior can be explained by the", "after": "words in the NLU training set can be modeled as a", "start_char_pos": 237, "end_char_pos": 287, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "phenomenon", "after": "distribution", "start_char_pos": 300, "end_char_pos": 310, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "Trained", "after": null, "start_char_pos": 340, "end_char_pos": 347, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "LGTR", "start_char_pos": 757, "end_char_pos": 757, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "our method", "after": "LGTR", "start_char_pos": 1074, "end_char_pos": 1084, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "in distribution test", "after": "in-distribution", "start_char_pos": 1171, "end_char_pos": 1191, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 96, 205, 312, 542, 677, 859, 1028], "domain": "arxiv"}
{"doc_id": "2103.09474", "revision_depth": 1, "before_revision": "Previous works on expressive text-to-speech (TTS) have a limitation on robustness and speed when training and inferring. Such drawbacks mostly come from autoregressive decoding, which makes the succeeding step vulnerable to preceding error. To overcome this weakness , we propose STYLER, a novel expressive text-to-speech model with parallelized architecture. Expelling autoregressive decoding and introducing speech decomposition for encoding enables speech synthesis more robust even with high style transfer performance . Moreover, our novel noise modeling approach from audio using domain adversarial training and Residual Decoding enabled style transferwithout transferring noise . Our experiments prove the naturalness and expressiveness of our model from comparison with other parallel TTS models . Together we investigate our model 's robustness and speed by comparison with the expressive TTS model with autoregressive decoding .", "after_revision": "Previous works on neural text-to-speech (TTS) have been tackled on limited speed in training and inference time, robustness for difficult synthesis conditions, expressiveness, and controllability. Although several approaches resolve some limitations, none of them has resolved all weaknesses at once. In this paper , we propose STYLER, an expressive and controllable text-to-speech model with robust speech synthesis and high speed. Excluding autoregressive decoding and introducing a novel audio-text aligning method called Mel Calibrator leads speech synthesis more robust on long, unseen data. Disentangled style factor modeling under supervision enlarges the controllability of synthesizing speech with fruitful expressivity . Moreover, our novel noise modeling pipeline using domain adversarial training and Residual Decoding enables noise-robust style transfer, decomposing the noise without any additional label. Our extensive and various experiments demonstrate STYLER's effectiveness in the aspects of speed, robustness, expressiveness, and controllability by comparison with existing neural TTS models and ablation studies. Synthesis samples of our model and experiment results are provided via our demo page .", "edit_actions": [{"type": "R", "before": "expressive", "after": "neural", "start_char_pos": 18, "end_char_pos": 28, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "a limitation on robustness and speed when training and inferring. Such drawbacks mostly come from autoregressive decoding, which makes the succeeding step vulnerable to preceding error. To overcome this weakness", "after": "been tackled on limited speed in training and inference time, robustness for difficult synthesis conditions, expressiveness, and controllability. Although several approaches resolve some limitations, none of them has resolved all weaknesses at once. In this paper", "start_char_pos": 55, "end_char_pos": 266, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "a novel expressive", "after": "an expressive and controllable", "start_char_pos": 288, "end_char_pos": 306, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "parallelized architecture. Expelling", "after": "robust speech synthesis and high speed. Excluding", "start_char_pos": 333, "end_char_pos": 369, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "speech decomposition for encoding enables speech", "after": "a novel audio-text aligning method called Mel Calibrator leads speech", "start_char_pos": 410, "end_char_pos": 458, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "even with high style transfer performance", "after": "on long, unseen data. Disentangled style factor modeling under supervision enlarges the controllability of synthesizing speech with fruitful expressivity", "start_char_pos": 481, "end_char_pos": 522, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "approach from audio", "after": "pipeline", "start_char_pos": 560, "end_char_pos": 579, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "enabled style transferwithout transferring noise . Our experiments prove the naturalness and expressiveness of our model from comparison with other parallel TTS models . Together we investigate our model 's robustness and speed by comparison with the expressive TTS model with autoregressive decoding", "after": "enables noise-robust style transfer, decomposing the noise without any additional label. Our extensive and various experiments demonstrate STYLER's effectiveness in the aspects of speed, robustness, expressiveness, and controllability by comparison with existing neural TTS models and ablation studies. Synthesis samples of our model and experiment results are provided via our demo page", "start_char_pos": 636, "end_char_pos": 936, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 120, 240, 359, 524, 686, 805], "domain": "arxiv"}
{"doc_id": "2103.09474", "revision_depth": 2, "before_revision": "Previous works on neural text-to-speech (TTS) have been tackled on limited speed in training and inference time, robustness for difficult synthesis conditions, expressiveness, and controllability. Although several approaches resolve some limitations, none of them has resolved all weaknesses at once. In this paper, we propose STYLER, an expressive and controllable text-to-speech model with robust speech synthesisand high speed. Excluding autoregressive decoding and introducing a novel audio-text aligning method called Mel Calibrator leads speech synthesis more robust on long, unseen data. Disentangled style factor modeling under supervision enlarges the controllability of synthesizing speech with fruitful expressivity. Moreover, our novel noise modeling pipeline using domain adversarial training and Residual Decoding enables noise-robust style transfer, decomposing the noise without any additional label. Our extensive and various experiments demonstrate STYLER's effectiveness in the aspects of speed , robustness, expressiveness, and controllability by comparison with existing neural TTS models and ablation studies . Synthesis samples of our model and experiment results are provided via our demo page  .", "after_revision": "Previous works on neural text-to-speech (TTS) have been addressed on limited speed in training and inference time, robustness for difficult synthesis conditions, expressiveness, and controllability. Although several approaches resolve some limitations, there has been no attempt to solve all weaknesses at once. In this paper, we propose STYLER, an expressive and controllable TTS framework with high-speed and robust synthesis. Our novel audio-text aligning method called Mel Calibrator and excluding autoregressive decoding enable rapid training and inference and robust synthesis on unseen data. Also, disentangled style factor modeling under supervision enlarges the controllability in synthesizing process leading to expressive TTS. On top of it, a novel noise modeling pipeline using domain adversarial training and Residual Decoding empowers noise-robust style transfer, decomposing the noise without any additional label. Various experiments demonstrate that STYLER is more effective in speed and robustness than expressive TTS with autoregressive decoding and more expressive and controllable than reading style non-autoregressive TTS . Synthesis samples and experiment results are provided via our demo page , and code is available publicly .", "edit_actions": [{"type": "R", "before": "tackled", "after": "addressed", "start_char_pos": 56, "end_char_pos": 63, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "none of them has resolved", "after": "there has been no attempt to solve", "start_char_pos": 251, "end_char_pos": 276, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "text-to-speech model with robust speech synthesisand high speed. Excluding autoregressive decoding and introducing a", "after": "TTS framework with high-speed and robust synthesis. Our", "start_char_pos": 366, "end_char_pos": 482, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "leads speech synthesis more robust on long,", "after": "and excluding autoregressive decoding enable rapid training and inference and robust synthesis on", "start_char_pos": 538, "end_char_pos": 581, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "Disentangled", "after": "Also, disentangled", "start_char_pos": 595, "end_char_pos": 607, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "of synthesizing speech with fruitful expressivity. Moreover, our", "after": "in synthesizing process leading to expressive TTS. On top of it, a", "start_char_pos": 677, "end_char_pos": 741, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "enables", "after": "empowers", "start_char_pos": 828, "end_char_pos": 835, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Our extensive and various experiments demonstrate STYLER's effectiveness in the aspects of speed , robustness, expressiveness, and controllability by comparison with existing neural TTS models and ablation studies", "after": "Various experiments demonstrate that STYLER is more effective in speed and robustness than expressive TTS with autoregressive decoding and more expressive and controllable than reading style non-autoregressive TTS", "start_char_pos": 917, "end_char_pos": 1130, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "D", "before": "of our model", "after": null, "start_char_pos": 1151, "end_char_pos": 1163, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ", and code is available publicly", "start_char_pos": 1218, "end_char_pos": 1218, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 196, 300, 430, 594, 727, 916], "domain": "arxiv"}
{"doc_id": "2103.14972", "revision_depth": 4, "before_revision": "The understanding of an offense is subjective and people may have different opinions about the offensiveness of a comment. Also , offenses and hate speech may occur through sarcasm, which hides the real intention of the comment and makes the decision of the annotators more confusing. Therefore, provide a well-structured annotation process is crucial to a better understanding of hate speech and offensive language phenomena, as well as supply better performance for machine learning classifiers. In this paper, we describe a corpus annotation process , which was guided by a linguist, and a hate speech skilled to support the identification of hate speech and offensive language on social media. In addition, we provide the first robust corpus of this kind for the Brazilian Portuguese language. The corpus was collected from Instagram posts of political personalities and manually annotated, being composed by 7,000 documents annotated according to three different layers: a binary classification (offensive versus non-offensive language), the level of offense (highly offensive, moderately offensive, and slightly offensive messages), and the identification regarding the target of the discriminatory content (xenophobia, racism, homophobia, sexism, religious intolerance, partyism, apology to the dictatorship, antisemitism, and fat phobia ). Each comment was annotated by three different annotators , and achieved high inter-annotator agreement. The new proposed annotation approach is also language and domain-independent (although it has been applied for Brazilian Portuguese ) .", "after_revision": "The understanding of an offense is subjective and people may have different opinions about the offensiveness of a comment. Moreover , offenses and hate speech may occur through sarcasm, which hides the real intention of the comment and makes the decision of the annotators more confusing. Therefore, providing a well-structured annotation process is crucial to a better understanding of hate speech and offensive language phenomena, as well as supplying better performance for machine learning classifiers. In this paper, we describe a corpus annotation process proposed by a linguist, a hate speech specialist, and machine learning engineers in order to support the identification of hate speech and offensive language on social media. In addition, we provide the first robust dataset of this kind for the Brazilian Portuguese language. The corpus was collected from Instagram posts of political personalities and manually annotated, being composed by 7,000 annotated documents according to three different layers: a binary classification (offensive versus non-offensive language), the level of offense (highly offensive, moderately offensive, and slightly offensive messages), and the identification regarding the target of the discriminatory content (xenophobia, racism, homophobia, sexism, religious intolerance, partyism, apology to the dictatorship, antisemitism, and fatphobia ). Each comment was annotated by three different annotators and achieved high inter-annotator agreement. The proposed annotation approach is also language and domain-independent nevertheless it is currently customized for Brazilian Portuguese .", "edit_actions": [{"type": "R", "before": "Also", "after": "Moreover", "start_char_pos": 123, "end_char_pos": 127, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "provide", "after": "providing", "start_char_pos": 296, "end_char_pos": 303, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "supply", "after": "supplying", "start_char_pos": 438, "end_char_pos": 444, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", which was guided", "after": "proposed", "start_char_pos": 553, "end_char_pos": 571, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "and", "after": null, "start_char_pos": 587, "end_char_pos": 590, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "skilled", "after": "specialist, and machine learning engineers in order", "start_char_pos": 605, "end_char_pos": 612, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "corpus", "after": "dataset", "start_char_pos": 739, "end_char_pos": 745, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "documents annotated", "after": "annotated documents", "start_char_pos": 919, "end_char_pos": 938, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "fat phobia", "after": "fatphobia", "start_char_pos": 1334, "end_char_pos": 1344, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1405, "end_char_pos": 1406, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "D", "before": "new", "after": null, "start_char_pos": 1456, "end_char_pos": 1459, "major_intent": "coherence", "raw_intents": ["coherence", "style", "coherence"]}, {"type": "R", "before": "(although it has been applied", "after": "nevertheless it is currently customized", "start_char_pos": 1529, "end_char_pos": 1558, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": ")", "after": null, "start_char_pos": 1584, "end_char_pos": 1585, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 122, 284, 497, 697, 797, 975, 1347, 1451], "domain": "arxiv"}
{"doc_id": "2104.12265", "revision_depth": 2, "before_revision": "This paper presents a new approach for offensive language and hate speech detection on social media. Our approach incorporate an offensive lexicon composed by implicit and explicit offensive and swearing expressions annotated with binary classes: context-dependent and context-independent offensive. Due to the severity of the hate speech and offensive comments in Brazil, and the lack of research in Portuguese, Brazilian Portuguese is the language used to validate our method. However, the proposal may be applied to any other language or domain. Based on the obtained results, the proposed approach showed high performance results overcoming the current baselines for European and Brazilian Portuguese.", "after_revision": "This paper provides a new approach for offensive language and hate speech detection on social media. Our approach incorporates an offensive lexicon composed of implicit and explicit offensive and swearing expressions annotated with binary classes: context-dependent and context-independent offensive. Due to the severity of the hate speech and offensive comments in Brazil, and the lack of research in Portuguese, Brazilian Portuguese is the language used to validate the proposed method. Nevertheless, our proposal may be applied to any other language or domain. Based on the obtained results, the proposed approach showed high-performance overcoming the current baselines for European and Brazilian Portuguese.", "edit_actions": [{"type": "R", "before": "presents", "after": "provides", "start_char_pos": 11, "end_char_pos": 19, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "incorporate", "after": "incorporates", "start_char_pos": 114, "end_char_pos": 125, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "by", "after": "of", "start_char_pos": 156, "end_char_pos": 158, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "our method. However, the", "after": "the proposed method. Nevertheless, our", "start_char_pos": 467, "end_char_pos": 491, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "high performance results", "after": "high-performance", "start_char_pos": 609, "end_char_pos": 633, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 100, 299, 478, 548], "domain": "arxiv"}
{"doc_id": "2105.01542", "revision_depth": 1, "before_revision": "Machine reading comprehension (MRC) is a sub-field in natural language processing or computational linguistics. MRC aims to help computers understand unstructured texts and then answer questions related to them. In this paper, we present a new Vietnamese corpus for conversational machine reading comprehension ( ViCoQA ), consisting of 10,000 questions with answers over 2,000 conversations about health news articles. We analyze ViCoQA in depth with different linguistic aspects. Then, we evaluate several baseline models about dialogue and reading comprehension on the ViCoQA corpus. The best model obtains an F1 score of 45.27\\%, which is 30.91 points behind human performance (76.18\\%), indicating that there is ample room for improvement.", "after_revision": "Machine reading comprehension (MRC) is a sub-field in natural language processing or computational linguistics. MRC aims to help computers understand unstructured texts and then answer questions related to them. In this paper, we present a new Vietnamese corpus for conversational machine reading comprehension ( UIT-ViCoQA ), consisting of 10,000 questions with answers over 2,000 conversations about health news articles. We analyze UIT-ViCoQA in depth with different linguistic aspects. Then, we evaluate several baseline models about dialogue and reading comprehension on the UIT-ViCoQA corpus. The best model obtains an F1 score of 45.27\\%, which is 30.91 points behind human performance (76.18\\%), indicating that there is ample room for improvement.", "edit_actions": [{"type": "R", "before": "ViCoQA", "after": "UIT-ViCoQA", "start_char_pos": 313, "end_char_pos": 319, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "ViCoQA", "after": "UIT-ViCoQA", "start_char_pos": 431, "end_char_pos": 437, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "ViCoQA", "after": "UIT-ViCoQA", "start_char_pos": 572, "end_char_pos": 578, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 111, 211, 419, 481, 586], "domain": "arxiv"}
{"doc_id": "2105.01542", "revision_depth": 2, "before_revision": "Machine reading comprehension (MRC) is a sub-field in natural language processing or computational linguistics. MRC aims to help computers understand unstructured texts and then answer questions related to them. In this paper, we present a new Vietnamese corpus for conversational machine reading comprehension (UIT-ViCoQA), consisting of 10,000 questions with answers  over 2,000 conversations about health news articles. We analyze UIT-ViCoQA in depth with different linguistic aspects. Then, we evaluate several baseline models about dialogue and reading comprehension on the UIT-ViCoQA corpus. The best model obtains an F1 score of 45.27\\%, which is 30.91 points behind human performance (76.18\\%), indicating that there is ample room for improvement.", "after_revision": "Machine reading comprehension (MRC) is a sub-field in natural language processing which aims to help computers understand unstructured texts and then answer questions related to them. In practice, conversation is an essential way to communicate and transfer information. To help machines understand conversation texts, we present UIT-ViCoQA - a new corpus for conversational machine reading comprehension in the Vietnamese language. This corpus consists of 10,000 questions with answers to over 2,000 conversations about health news articles. Then, we evaluate several baseline approaches for conversational machine comprehension on the UIT-ViCoQA corpus. The best model obtains an F1 score of 45.27\\%, which is 30.91 points behind human performance (76.18\\%), indicating that there is ample room for improvement.", "edit_actions": [{"type": "R", "before": "or computational linguistics. MRC", "after": "which", "start_char_pos": 82, "end_char_pos": 115, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "this paper, we present a new Vietnamese", "after": "practice, conversation is an essential way to communicate and transfer information. To help machines understand conversation texts, we present UIT-ViCoQA - a new", "start_char_pos": 215, "end_char_pos": 254, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "R", "before": "(UIT-ViCoQA), consisting", "after": "in the Vietnamese language. This corpus consists", "start_char_pos": 311, "end_char_pos": 335, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "to", "start_char_pos": 369, "end_char_pos": 369, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "D", "before": "We analyze UIT-ViCoQA in depth with different linguistic aspects.", "after": null, "start_char_pos": 423, "end_char_pos": 488, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "models about dialogue and reading", "after": "approaches for conversational machine", "start_char_pos": 524, "end_char_pos": 557, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 111, 211, 422, 488, 597], "domain": "arxiv"}
{"doc_id": "10010", "revision_depth": 1, "before_revision": "During a meeting of the Democratic Party of Serbia today , the Prime Minister Vojislav Kotunica said that he will not agree to Kosovo's independence  and that the solution for Kosovo is wide autonomy. \"When I say autonomy, that's more than autonomy Kosovo had prior to 1999, but it is still just autonomy and not independence,\" Kotunica said. After the war in 1999, Kosovo was put under UNMIK's administration, but it still officialy remains a province in Serbia. The province's Albanian majority wants full independence from Serbia. He also reaffirmed his stance on participation of Serbs in Kosovo's institutions: \"For now, there is no room for Serbian representatives in institutions of Kosovo.\" The Serbian government had advised Serbs in Kosovo not to vote in the local parliamentary elections held last October. Kotunica said that the degree of participation depends on advancement in protection of human rights of the Serbs remaining in the province.", "after_revision": "During a meeting of the Democratic Party of Serbia on Saturday , the Prime Minister Vojislav Kotunica said that he will not agree to Kosovo's independence , and that the solution for Kosovo is wide autonomy. \"When I say autonomy, that's more than autonomy Kosovo had prior to 1999, but it is still just autonomy and not independence,\" Kotunica said. After the war in 1999, Kosovo was put under UNMIK's administration, but it still officially remains a province in Serbia. The province's Albanian majority wants full independence from Serbia. Kotunica also reaffirmed his stance on participation of Serbs in Kosovo's institutions: \"For now, there is no room for Serbian representatives in institutions of Kosovo.\" The Serbian government had advised Serbs in Kosovo not to vote in the local parliamentary elections held last October. Kotunica said that the degree of participation depends on advancement in protection of human rights of the Serbs remaining in the province.", "edit_actions": [{"type": "R", "before": "today", "after": "on Saturday", "start_char_pos": 51, "end_char_pos": 56, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 149, "end_char_pos": 149, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "officialy", "after": "officially", "start_char_pos": 424, "end_char_pos": 433, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "He", "after": "Kotunica", "start_char_pos": 534, "end_char_pos": 536, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 200, 342, 463, 533, 698, 817], "domain": "news"}
{"doc_id": "10044", "revision_depth": 3, "before_revision": "Five of the sites sued are registered in the U.S., and the sixth site is registered in Spain. The sites sued are Shuntv.net, Zonatracker.com, Btefnet.net, Scifi-classics.net, Cddvdheaven.co.uk, and Bragginrights.biz  According to research by Envisional, recent surveys report downloads of TV programmes are up by 150\\% in a year. Most of the downloads, 70\\%, used BitTorrent sites. With smaller size than movies and broadband access, popular shows often appear within hours of airing on TV. CEO of the MPAA , Dan Glickman told the BBC  \"Since we began shutting these sites down, the time that it takes to download a file on BitTorrent has increased exponentially  which means the experience of downloading copyrighted films and TV shows is not what it used to be. We intend to make it even worse. Protecting the television industry is essential.\"", "after_revision": "Five of the sites sued are registered in the U.S., and the sixth site is registered in Spain. The sites sued are Shuntv.net, Zonatracker.com, Btefnet.net, Scifi-classics.net, Cddvdheaven.co.uk, and Bragginrights.biz . According to research by Envisional, recent surveys report downloads of TV programmes are up by 150\\% in a year. Most of the downloads, 70\\%, used BitTorrent sites. With the increasing ubiquity of broadband Internet access, popular television shows, which are smaller downloads than movies, often appear within hours of airing on TV. MPAA CEO Dan Glickman told the BBC , \"Since we began shutting these sites down, the time that it takes to download a file on BitTorrent has increased exponentially , which means the experience of downloading copyrighted films and TV shows is not what it used to be. We intend to make it even worse. Protecting the television industry is essential.\"", "edit_actions": [{"type": "A", "before": null, "after": ".", "start_char_pos": 216, "end_char_pos": 216, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "smaller size than movies and broadband", "after": "the increasing ubiquity of broadband Internet", "start_char_pos": 387, "end_char_pos": 425, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "shows", "after": "television shows, which are smaller downloads than movies,", "start_char_pos": 442, "end_char_pos": 447, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "CEO of the MPAA ,", "after": "MPAA CEO", "start_char_pos": 491, "end_char_pos": 508, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 535, "end_char_pos": 535, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 663, "end_char_pos": 663, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 93, 329, 381, 490, 763, 796], "domain": "news"}
{"doc_id": "10050", "revision_depth": 1, "before_revision": "Elections for local governments were held in Croatia today . The polls were closed at 19:00 local time (UTC+2). The results are not in yet, but exit polls and preliminary results were published by GONG and State Election Committee. According to GONG, an NGO observing the elections, Milan Bandi\u0107 will be able to form a government in the city of Zagreb, since his list has won around 46\\%. This would give them 27 of 51 seats in the capital city. Bandi\u0107 is a head of the coalition list formed by Social Democrat Party of Croatia (SDP), Croatian Peasant Party (HSS), and Croatian Pensioner Party (HUS). The preliminary results from the State Election Committee show 40.90\\% for Bandi\u0107 at 23:30 local time. Branimir Glava\u0161's, who was recently expelled from Croatian Democratic Union (HDZ), list seems to be in lead in Osijek according to the exit polls, with 27.24\\% (which would equal 9 seats in the local parliament) , Liberal Party has 18.47\\% (6 seats)  and Croatian Party of Justice is third with 14.5\\% (4 seats). According to GONG, Social Democrat Party has a lead in Split with 29.02\\% of votes (10 seats), while Croatian Democratic Union is following with 18.54\\% (6 seats).", "after_revision": "Elections for local governments were held in Croatia Sunday . The polls were closed at 19:00 local time (UTC+2). The results are not in yet, but exit polls and preliminary results were published by GONG and State Election Committee. According to GONG, an NGO observing the elections, Milan Bandi\u0107 , can form a government in the city of Zagreb, since his list has won around 46\\%. This would give the coalition 27 of 51 seats in the capital city. Bandi\u0107 is a head of the coalition list formed by Social Democrat Party of Croatia (SDP), Croatian Peasant Party (HSS), and Croatian Pensioner Party (HUS). The preliminary results from the State Election Committee show 40.90\\% for Bandi\u0107 at 23:30 local time. Branimir Glava\u0161's, who was recently expelled from Croatian Democratic Union (HDZ), list seems to be in lead in Osijek according to the exit polls, with 27.24\\% (which would equal 9 seats in the local parliament) . Liberal Party has 18.47\\% (6 seats) , and Croatian Party of Justice is third with 14.5\\% (4 seats). According to GONG, Social Democrat Party has a lead in split with 29.02\\% of votes (10 seats), while Croatian Democratic Union is following with 18.54\\% (6 seats).", "edit_actions": [{"type": "R", "before": "today", "after": "Sunday", "start_char_pos": 53, "end_char_pos": 58, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "will be able to", "after": ", can", "start_char_pos": 296, "end_char_pos": 311, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "them", "after": "the coalition", "start_char_pos": 405, "end_char_pos": 409, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ",", "after": ".", "start_char_pos": 916, "end_char_pos": 917, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 954, "end_char_pos": 954, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Split", "after": "split", "start_char_pos": 1072, "end_char_pos": 1077, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 60, 111, 231, 388, 445, 600, 703, 1016], "domain": "news"}
{"doc_id": "100524", "revision_depth": 1, "before_revision": "360px|Executives at the press conference of 2008 APRICOT (from left to right): Philip Smith, Gaurab Raj Upadhaya (Chairman of Asia & Pacific Internet Association), Tony Teng, and Feipei Lai. The twelfth-annual Asia Pacific Regional Internet Conference on Operational Technologies (a.k.a APRICOT) , returned to Taiwan this year at  Taipei Howard Plaza Hotel ; the first appearance since the 2003 conference held by the Taiwan Network Information Center (TWNIC) at the Taipei International Convention Center and Grand Hyatt Taipei. As Wikinews Journalist Rico Shen reported on the recent \"Edison Chen photo scandal\" incident, he commented: Workshops with varied topics and different technology levels took place from February 20 to 24, while several main seminars and speeches for industry, governmental, and academic executives ran from February 25 to 29. Several industry experts such as Wilfred Kwan (Chief Technology Officer of AsiaNetCom), Chung-laung Liu ( ISOC Taiwan Chapter Chair  ), and Maemura Akinori (EC Chair of APNIC) will give several speeches related to the Internet industry at the conference.", "after_revision": "360px|Executives at the press conference of 2008 APRICOT (from left to right): Philip Smith, Gaurab Raj Upadhaya (Chairman of Asia & Pacific Internet Association), Tony Teng, and Feipei Lai. The twelfth-annual Asia Pacific Regional Internet Conference on Operational Technologies (a.k.a APRICOT) returned to Taiwan this year at the Taipei Howard Plaza Hotel . This is the first appearance since the 2003 conference held by the Taiwan Network Information Center (TWNIC) at the Taipei International Convention Center and Grand Hyatt Taipei. When Wikinews journalist Rico Shen reported on the recent \"Edison Chen photo scandal\" incident, Tony Tien-lai Teng commented: Workshops with varied topics and different technology levels took place from February 20 to 24, while several main seminars and speeches for industry, governmental, and academic executives ran from February 25 to 29. Several industry experts such as Wilfred Kwan (Chief Technology Officer of AsiaNetCom), Chung-laung Liu ( Taiwan Chapter Chair of ISOC ), and Maemura Akinori (EC Chair of APNIC) will give several speeches related to the Internet industry at the conference.", "edit_actions": [{"type": "D", "before": ",", "after": null, "start_char_pos": 296, "end_char_pos": 297, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 330, "end_char_pos": 330, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": ";", "after": ". This is", "start_char_pos": 357, "end_char_pos": 358, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "As Wikinews Journalist", "after": "When Wikinews journalist", "start_char_pos": 530, "end_char_pos": 552, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "he", "after": "Tony Tien-lai Teng", "start_char_pos": 624, "end_char_pos": 626, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "ISOC", "after": null, "start_char_pos": 961, "end_char_pos": 965, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "of ISOC", "start_char_pos": 987, "end_char_pos": 987, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 190, 358, 529, 854], "domain": "news"}
{"doc_id": "100534", "revision_depth": 1, "before_revision": "There is also a Tree of Life Web Project, and a Catalogue of Life, the latter being the most successful; it is a compilation of 1,008,965 species, from 47 taxonomical databases. ZipcodeZoo offers  over 3 million web pages describing species of plants and animals . Pages contain 258, 753 photostaken by 1, 369 photographers, 1, 104 sound recordings, and definitions of 234 , 888 terms. 70 , 725 Large photos can be zoomed and panned . ARKive collects media of species, and the All Species Foundation was a failed early attempt at a web catalog.", "after_revision": "There is also a Tree of Life Web Project, and a Catalogue of Life, the latter being the most successful; it is a compilation of 1,008,965 species, from 47 taxonomical databases. ZipcodeZoo offers descriptions of over 3 million species of plants and animals , with 250 , 000 photos, maps, pronunciations, and definitions in 13 languages . ARKive collects media of species, and the All Species Foundation was a failed early attempt at a web catalog.", "edit_actions": [{"type": "A", "before": null, "after": "descriptions of", "start_char_pos": 196, "end_char_pos": 196, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "web pages describing", "after": null, "start_char_pos": 212, "end_char_pos": 232, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": ". Pages contain 258, 753 photostaken by 1, 369 photographers, 1, 104 sound recordings, and definitions of 234", "after": null, "start_char_pos": 263, "end_char_pos": 372, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "888 terms. 70", "after": "with 250", "start_char_pos": 375, "end_char_pos": 388, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "725 Large photos can be zoomed and panned", "after": "000 photos, maps, pronunciations, and definitions in 13 languages", "start_char_pos": 391, "end_char_pos": 432, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 104, 177, 385], "domain": "news"}
{"doc_id": "100809", "revision_depth": 1, "before_revision": "Oil in productionCrude oil prices in New York rose to a new record of  $102.59 per barrel on Thursday, although the figure increased even more during after hours trading. In less than a month, prices have risen $10, leading the figures to above the record highs set during the 1980s (taking inflation into account) . Information provided by the International Energy Agency has said that the previous record was $102.53, with the figures being adjusted according to inflation levels. The weak dollar is seen as a major cause of this rise . Congressman Ron Paul of Texas, pointed out to Federal Reserve chairman Ben Bernanke in a committee meeting this week that despite the  price of oil 's rapid ascent , it had remained flat when compared to the price of gold.  Increasing demand for oil has also been cited as the cause for this increase. Violence in Nigeria earlier this year has led to a drop the country's production by almost a quarter. The most recent information produced by the Energy Information Administration has shown an increase in gasoline prices for all but one of the areas surveyed. A graph to show the increase in gasoline pricesThere have also been suggestions that reports of a fire at a National Gas Terminal may have contributed to the rising oil price. Time Evans from Citigroup Futures has stated he believes that this fire at the UK natural gas terminal is creating a strong push in the European market, and that is translating here [the US].\"", "after_revision": "An oil refineryCrude oil prices in New York rose to a new record of US $102.59 per barrel on Thursday, although the figure increased even more during after hours trading. In less than a month, prices have risen $10, leading the inflation adjusted prices above the record highs set during the 1980s . Information provided by the International Energy Agency has said that the previous record was $102.53, with the figures being adjusted according to inflation levels. While the weak dollar is seen as a major cause of this rise , Congressman Ron Paul of Texas, pointed out to Federal Reserve chairman Ben Bernanke in a committee meeting this week that despite the rapid ascent the of price of oil , it had remained flat when compared to the price of gold. A graph showing recent increases in gasoline prices Increasing demand for oil has also been cited as the cause for this increase. Violence in Nigeria earlier this year has led to a drop the country's production by almost a quarter. The most recent information produced by the Energy Information Administration has shown an increase in gasoline prices for all but one of the areas surveyed. There have also been suggestions that reports of a fire at a National Gas Terminal may have contributed to the rising oil price. Time Evans from Citigroup Futures said he believes \" that this fire at the UK natural gas terminal is creating a strong push in the European market, and that is translating here [the US].\"", "edit_actions": [{"type": "R", "before": "Oil in productionCrude oil", "after": "An oil refineryCrude oil", "start_char_pos": 0, "end_char_pos": 26, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "A", "before": null, "after": "US", "start_char_pos": 70, "end_char_pos": 70, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "fluency", "meaning-changed"]}, {"type": "R", "before": "figures to", "after": "inflation adjusted prices", "start_char_pos": 228, "end_char_pos": 238, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "D", "before": "(taking inflation into account)", "after": null, "start_char_pos": 283, "end_char_pos": 314, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "The", "after": "While the", "start_char_pos": 483, "end_char_pos": 486, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ".", "after": ",", "start_char_pos": 537, "end_char_pos": 538, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "rapid ascent the of", "start_char_pos": 673, "end_char_pos": 673, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "'s rapid ascent", "after": null, "start_char_pos": 687, "end_char_pos": 702, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "A graph showing recent increases in gasoline prices", "start_char_pos": 762, "end_char_pos": 762, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "A graph to show the increase in gasoline pricesThere", "after": "There", "start_char_pos": 1101, "end_char_pos": 1153, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "has stated he believes", "after": "said he believes \"", "start_char_pos": 1311, "end_char_pos": 1333, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 170, 316, 482, 761, 840, 942, 1100, 1276], "domain": "news"}
{"doc_id": "100868", "revision_depth": 1, "before_revision": "The H5N1 Avian Flu Virus has been found in a dead wild Canadian Goose in Abbotsbury Swannery in Dorset, England. This is the 11th case of the virus turning up in wild birds. The goose was found on February 25, 2008. 10 other cases of the virus have appeared in dead birds, all Mute Swans from the same area.", "after_revision": "The H5N1 Avian Flu virus has been found in a dead wild Canadian Goose in Abbotsbury Swannery in Dorset, England. This is the eleventh case of the virus turning up in wild birds. The goose was discovered on February 25, 2008. Ten other cases of the virus have appeared in dead birds, all Mute Swans from the same area.", "edit_actions": [{"type": "R", "before": "Virus", "after": "virus", "start_char_pos": 19, "end_char_pos": 24, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "11th", "after": "eleventh", "start_char_pos": 125, "end_char_pos": 129, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "found", "after": "discovered", "start_char_pos": 188, "end_char_pos": 193, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "10", "after": "Ten", "start_char_pos": 216, "end_char_pos": 218, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}], "sents_char_pos": [0, 112, 173, 215], "domain": "news"}
{"doc_id": "100873", "revision_depth": 1, "before_revision": "Interpol has released an \"Orange Notice\" for the Southeast Asian Jemaah Islamiyah terrorist group leader Mas Selamat bin Kastari, who escaped from a detention center Wednesday. After the Orange Notice was released, Mas Selamat's picture and fingerprints were released to Interpol's 186 member countries. He remains at large.", "after_revision": "Interpol has issued an \"Orange Notice\" for the leader of southeast Asian Jemaah Islamiyah , Mas Selamat bin Kastari, who escaped from a detention center Wednesday. After the \"Orange Notice\" was released, Mas Selamat's picture and fingerprints were released to Interpol's 186 member countries. He remains at large.", "edit_actions": [{"type": "R", "before": "released", "after": "issued", "start_char_pos": 13, "end_char_pos": 21, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Southeast", "after": "leader of southeast", "start_char_pos": 49, "end_char_pos": 58, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "terrorist group leader", "after": ",", "start_char_pos": 82, "end_char_pos": 104, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "Orange Notice", "after": "\"Orange Notice\"", "start_char_pos": 187, "end_char_pos": 200, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 176, 303], "domain": "news"}
{"doc_id": "101134", "revision_depth": 2, "before_revision": "240px|The starting of 42K and 21K classes. The 2008 Taipei County Jin Shi International Marathon which was established in 1979 and suspended from 1986 to 2001 before being transformed from \"Jin Shan\" to \"Jin Shi\" with its racing scale expanded, crossed through 3 townships including Wanli, Jinshan, and Shihmen was held last Sunday ( Mar. 2 in Taipei Time). Not only professional runners from Asian  European countries participated in this race,  several enterprises, governmental and academic teams all supported this race by participating in 6K Fun Run classes. Even though the Central Weather Bureau warned of an impending dust storm, good weather graced this race with rain-free conditions. Another challenge for the runners in the 21K and 42K classes was cold winds due to the race route partially including coast line. 180px|Rik Ceulemans, a Belgian marathon runner, finally won Men's 42K Champion of this race. Finally, Belgian marathon runner Rik Ceulemans, men's record holder of this race, extended his record with 2H18m13s to win the men's champion , and Yu-fang Hsu, women's champion of 2007 ING Taipei Marathon, won her title at women's class with 2H53m39s. Coincidently, Wen-chien Wu and Wan-ling Wu, former champions of this race, all won the 2nd place in their own gender class. As a recap for the race, not only runners gave their best performances in this race, but several foreign teams - especially from Okinawa, Japan - played a great role adding a cultural exchange aspect to the event. Bros Sports, the URLanizer, supervised by Taipei County Government in this race, some senior managers from Bros Sports URLanizations from the sports industry could arrange the best integration of several marathon races in Taiwan in the future even though Bros Sports promoted a special attempt this year with 3 marathon races including Jin Shi, Dajia Mazu, and UMC Hsinchu stages.", "after_revision": "240px|The start of 42K and 21K classes. The 2008 Taipei County Jin Shi International Marathon which was established in 1979 and suspended from 1986 to 2001 before being transformed from \"Jin Shan\" to \"Jin Shi\" with its racing scale expanded, crossed through three townships including Wanli, Jinshan, and Shihmen was held last Sunday ( March 2 in Taipei Time). Not only professional runners from Asian and European countries participated in this race, but several enterprises, governmental and academic teams also supported this race by participating in 6K Fun Run classes. Even though the Central Weather Bureau warned of an impending dust storm, good weather graced this race with rain-free conditions. Another challenge for the runners in the 21K and 42K classes was cold winds due to the race route partially including coast line. 180px|Rik Ceulemans, a Belgian marathon runner, finally won Men's 42K Champion of this race. Finally, Belgian marathon runner Rik Ceulemans, men's record holder of this race, extended his record with 2H18m13s to win the men's championship , and Yu-fang Hsu, women's champion of 2007 ING Taipei Marathon, won her title at women's class with 2H53m39s. Coincidently, Wen-chien Wu and Wan-ling Wu, former champions of this race, all won the second place in their own gender class. As a recap for the race, not only runners gave their best performances in this race, but several foreign teams especially from Okinawa, Japan played a great role adding a cultural exchange aspect to the event. Bros Sports, the URLanizer, supervised by Taipei County Government in this race, some senior managers from Bros Sports URLanizations from the sports industry could arrange the best integration of several marathon races in Taiwan in the future even though Bros Sports promoted a special attempt this year with three marathon races including Jin Shi, Dajia Mazu, and UMC Hsinchu stages.", "edit_actions": [{"type": "R", "before": "starting", "after": "start", "start_char_pos": 10, "end_char_pos": 18, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "3", "after": "three", "start_char_pos": 261, "end_char_pos": 262, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "Mar.", "after": "March", "start_char_pos": 334, "end_char_pos": 338, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "and", "start_char_pos": 399, "end_char_pos": 399, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "but", "start_char_pos": 446, "end_char_pos": 446, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "all", "after": "also", "start_char_pos": 500, "end_char_pos": 503, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "champion", "after": "championship", "start_char_pos": 1051, "end_char_pos": 1059, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "2nd", "after": "second", "start_char_pos": 1258, "end_char_pos": 1261, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "D", "before": "-", "after": null, "start_char_pos": 1406, "end_char_pos": 1407, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "D", "before": "-", "after": null, "start_char_pos": 1439, "end_char_pos": 1440, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "3", "after": "three", "start_char_pos": 1818, "end_char_pos": 1819, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}], "sents_char_pos": [0, 42, 357, 563, 694, 824, 917, 1170, 1294, 1508], "domain": "news"}
{"doc_id": "10138", "revision_depth": 1, "before_revision": "The BBC announced yesterday (16 May 2005) a trial of their new Interactive Media Player (iMP). Five thousand participants from across the UK were selected for this trial. Participants in the trial will be able to download selected BBC television and radio programmes. Approximately 190 hours of TV shows and 310 hours of radio programmes will be available for download. Some feature films and local programming will also be available. Ashley Highfield , BBC director of new media and technology has said this program \"Could just be the iTunes for the broadcast industry, enabling our audience to access our TV and radio programmes on their terms - anytime, any place, any how.\" The BBC is funded by a mandatory licence fee on all televisions in the UK . The fee is 126.50 a year for colour TV and 42 a year for monochrome TV.", "after_revision": "The BBC announced on Monday a trial of their new Interactive Media Player (iMP). Five thousand participants from across the U.K. were selected for this trial. Participants in the trial will be able to download selected BBC television and radio programmes. Approximately 190 hours of TV shows and 310 hours of radio programmes will be available for download. Some feature films and local programming will also be available. Ashley Highfield is the BBC's director of new media and technology . Highfield has said that this program \"Could just be the iTunes for the broadcast industry, enabling our audience to access our TV and radio programmes on their terms - anytime, any place, any how.\" The BBC is funded by a mandatory licence fee on all televisions in the U.K . The fee is 126.50 a year for colour TV and 42 a year for monochrome TV.", "edit_actions": [{"type": "R", "before": "yesterday (16 May 2005)", "after": "on Monday", "start_char_pos": 18, "end_char_pos": 41, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "UK", "after": "U.K.", "start_char_pos": 138, "end_char_pos": 140, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": ", BBC", "after": "is the BBC's", "start_char_pos": 452, "end_char_pos": 457, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "has said", "after": ". Highfield has said that", "start_char_pos": 495, "end_char_pos": 503, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "UK", "after": "U.K", "start_char_pos": 749, "end_char_pos": 751, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 94, 170, 267, 369, 434, 677], "domain": "news"}
{"doc_id": "101803", "revision_depth": 1, "before_revision": "Eliot Spitzer (D), Governator of New York New York Governator Eliot Spitzer has announced that he has had gender reassignment surgery , effective Monday. Spitzer was recently discovered to have been a client of a prostitution ring on numerous occasions. In a press conference, Governator Spitzer said: \"I am deeply sorry that I did not live up to what was expected of me. To every New Yorker, and to all those who believed in what I tried to stand for, I sincerely apologize. I will try once again outside of politics to serve the common Good .\" Public opinion polls in recent days have shown that 68\\% of New Yorkers wanted Spitzer to reassign and many politicians had also called for his assassination . As a result Spitzer will be succeeded by Lieutenant Governator David Paterson, who will become New York's first African American governator . It is believed he will be the first illegally blind governator in the United States. Spitzer remains a superdelegate for Senator Hillary Clinton's presidential campaign until he officially steps down as governator . New York Governator Eliot Spitzer and Senator Hillary Rodham Clinton.", "after_revision": "Eliot Spitzer (D), Governor of New York New York Governor Eliot Spitzer has announced that he has resigned , effective Monday. Spitzer was recently discovered to have been a client of a prostitution ring on numerous occasions. In a press conference, Governor Spitzer said: \"I am deeply sorry that I did not live up to what was expected of me. To every New Yorker, and to all those who believed in what I tried to stand for, I sincerely apologize. I will try once again outside of politics to serve the common good .\" Public opinion polls in recent days have shown that 68\\% of New Yorkers wanted Spitzer to resign and many politicians had also called for his resignation . As a result Spitzer will be succeeded by Lieutenant Governor David Paterson, who will become New York's first African American governor . It is believed he will be the first legally blind governor in the United States. Spitzer remains a superdelegate for Senator Hillary Clinton's presidential campaign until he officially steps down as governor . New York Governor Eliot Spitzer and Senator Hillary Rodham Clinton.", "edit_actions": [{"type": "R", "before": "Governator", "after": "Governor", "start_char_pos": 19, "end_char_pos": 29, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Governator", "after": "Governor", "start_char_pos": 51, "end_char_pos": 61, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "had gender reassignment surgery", "after": "resigned", "start_char_pos": 102, "end_char_pos": 133, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Governator", "after": "Governor", "start_char_pos": 277, "end_char_pos": 287, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Good", "after": "good", "start_char_pos": 538, "end_char_pos": 542, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "reassign", "after": "resign", "start_char_pos": 636, "end_char_pos": 644, "major_intent": "fluency", "raw_intents": ["meaning-changed", "fluency", "fluency"]}, {"type": "R", "before": "assassination", "after": "resignation", "start_char_pos": 690, "end_char_pos": 703, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Governator", "after": "Governor", "start_char_pos": 758, "end_char_pos": 768, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "governator", "after": "governor", "start_char_pos": 835, "end_char_pos": 845, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "illegally blind governator", "after": "legally blind governor", "start_char_pos": 884, "end_char_pos": 910, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "governator", "after": "governor", "start_char_pos": 1051, "end_char_pos": 1061, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Governator", "after": "Governor", "start_char_pos": 1073, "end_char_pos": 1083, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 153, 253, 371, 475, 705, 847, 932, 1063], "domain": "news"}
{"doc_id": "102520", "revision_depth": 1, "before_revision": "Diane Abbott: \"Post offices are central spaces within a community.\" Yesterday, the U.K. Labour Party successfully prevented a bill restricting the closure of post offices from being passed, although the leading Labour party only had a majority of 20 due to 19 of it' s MP's voting against the party line. One of these was Diane Abbott, who told Wikinews why she voted for the bill. Mrs. Abbott told Wikinewsthat \"Post offices are central spaces within a community.\" She added that they provide invaluable services and a point of contact for vulnerable people. She also claimed they were more important in places like her constituency which has a large number of elderly people. She said that \"they are particularly important [in] places in areas like Hackney where there are fewer bank branches and a large elderly population who rely on the post office to collect their pensions.\" Mrs. Abbot also said that she was \"appalled\" to hear of the Post Office closures in her constituency. The Shadow Minister for Business, Enterprise & Regulatory Reform Alan Duncan of the Conservative Party was also contacted by Wikinews , he directed Wikinews to his website to for more information on his opinion. On his website he says that it is \"simply not good enough\" to allow post offices to close.", "after_revision": "Diane Abbott: \"Post offices are central spaces within a community.\" Yesterday, the U.K. Labour Party successfully prevented a bill restricting the closure of post offices from being passed, although the governing Labour party only had a majority of 20 due to 19 of its MPs' voting against the party line. One of these was Diane Abbott, who told Wikinews why she voted for the bill. She told Wikinews: \"Post offices are central spaces within a community.\" She added that they provided invaluable services and a point of contact for vulnerable people. She also claimed they were more important in places like her constituency which has a large number of elderly people. She added: \"they are particularly important [in] places in areas like Hackney where there are fewer bank branches and a large elderly population who rely on the post office to collect their pensions.\" She was \"appalled\" to hear of the Post Office closures in her constituency. The Shadow Minister for Business, Enterprise & Regulatory Reform Alan Duncan of the Conservative Party was also contacted by Wikinews . He directed Wikinews to his website for more information on his opinion. On his website he says that it is \"simply not good enough\" to allow post offices to close.", "edit_actions": [{"type": "R", "before": "leading", "after": "governing", "start_char_pos": 203, "end_char_pos": 210, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "it' s MP's", "after": "its MPs'", "start_char_pos": 263, "end_char_pos": 273, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Mrs. Abbott told Wikinewsthat", "after": "She told Wikinews:", "start_char_pos": 382, "end_char_pos": 411, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "provide", "after": "provided", "start_char_pos": 486, "end_char_pos": 493, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "said that", "after": "added:", "start_char_pos": 682, "end_char_pos": 691, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "Mrs. Abbot also said that she", "after": "She", "start_char_pos": 882, "end_char_pos": 911, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ", he", "after": ". He", "start_char_pos": 1118, "end_char_pos": 1122, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "coherence"]}, {"type": "D", "before": "to", "after": null, "start_char_pos": 1156, "end_char_pos": 1158, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 67, 304, 381, 465, 559, 677, 886, 983, 1195], "domain": "news"}
{"doc_id": "102922", "revision_depth": 1, "before_revision": "Nicholas Sarkozy  On the first day of his state visit to Great Britain, in a Presidential speech French President Nicolas Sarkozy praised the British nation and called for an \"entente amicale\" between Great Britain and France, in place of the existing, more formal entente cordiale agreement. Speaking, unusually, to both Houses of Parliament, Mr Sarkozy went on to praise the help Britain gave to France during both World Wars and said that \"France would never URLet\" and that it owed Britain a debt of gratitude. As well as praising Great Britain, Mr Sarkozy also promised to propose sending more French troops to Afghanistan at the Bucharest Summit, to be held later this year, although he did not specify the number of troops he was planning to send. During his two-day state visit, Mr Sarkozy will meet with British Prime Minister Gordon Brown and they are expected to discuss many different issues, from global finance to Afghanistan.", "after_revision": "Nicholas Sarkozy in 2007. On the first day of his state visit to the United Kingdom, French President Nicolas Sarkozy gave a speech praising the British nation and called for an \"entente amicale\" between Great Britain and France, in place of the existing, more formal entente cordiale agreement. Speaking, unusually, to both Houses of Parliament, Sarkozy went on to praise the help Britain gave to France during both World Wars and said that \"France would never URLet\" and that it owed Britain a debt of gratitude. As well as praising Great Britain, Sarkozy also promised to propose sending more French troops to Afghanistan at the Bucharest Summit, to be held later this year, although he did not specify the number of troops he was planning to send. During his two-day state visit, Sarkozy will meet with British Prime Minister Gordon Brown and they are expected to discuss many different issues, from global finance to Afghanistan.", "edit_actions": [{"type": "A", "before": null, "after": "in 2007.", "start_char_pos": 17, "end_char_pos": 17, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Great Britain, in a Presidential speech", "after": "the United Kingdom,", "start_char_pos": 57, "end_char_pos": 96, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "praised", "after": "gave a speech praising", "start_char_pos": 130, "end_char_pos": 137, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "Mr", "after": null, "start_char_pos": 344, "end_char_pos": 346, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "D", "before": "Mr", "after": null, "start_char_pos": 550, "end_char_pos": 552, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "D", "before": "Mr", "after": null, "start_char_pos": 787, "end_char_pos": 789, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}], "sents_char_pos": [0, 292, 514, 754], "domain": "news"}
{"doc_id": "10439", "revision_depth": 1, "before_revision": "Italian police officers and guardians are going on trial for alleged abuse of anti-globalization protesters during the G8 meeting in Genoa in 2001. Reports made by over 250 participants in the event detail verbal and physical abuse. The police allegedly took detainees to a holding center outside the city, where they say they were verbally and physically abused,  threatened with rape , and spat at . In one alleged incident, the guards forced a woman's head down a toilet. Another allegation is that asphyxiating gas was sprayed at protesters in their cells. Over 60 people were injured while being taken into custody, three critically. One 23-year old protester was shot dead. Twenty-five of the 250 were brought up on minor charges of looting and ransacking. The police alleged that 93 protesters were in possession of dangerous weapons and resisted arrest, yet Genoa prosecutors  dropped all charges against them. Meanwhile, the police face accusations of planting evidence , and fabricating charges against the demonstrators. Unfortunately, due to delays in bringing the authorities to justice, the 5-year statute of limitations has expired on the major charges .", "after_revision": "Italian police officers and prison guards have been ordered to stand trial for the alleged abuse of anti-globalization protesters during the G8 summit in Genoa in 2001. Over 250 people people who attended the event have detailed verbal and physical abuse. The police took detainees to a holding center outside the city, where they say they were verbally and physically abused, spat at and threatened with rape . In one alleged incident, the guards forced a woman's head down a toilet. Another allegation is that asphyxiating gas was sprayed at protesters in their cells. Over 60 people were injured while being taken into custody, three critically. One 23-year old protester was shot dead. Twenty-five demostrators are standing trial on minor charges of looting and ransacking. The police alleged that 93 protesters were in possession of dangerous weapons and resisted arrest, yet Genoa prosecutors have dropped all charges against them. Meanwhile, the police have been charged with planting evidence and fabricating charges against the demonstrators. Unfortunately, due to delays in bringing the authorities to justice, the five-year statute of limitations is expectd to expire on most of the more serious charges before rulings are given .", "edit_actions": [{"type": "R", "before": "guardians are going on trial for", "after": "prison guards have been ordered to stand trial for the", "start_char_pos": 28, "end_char_pos": 60, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "meeting", "after": "summit", "start_char_pos": 122, "end_char_pos": 129, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Reports made by over", "after": "Over", "start_char_pos": 148, "end_char_pos": 168, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "participants in the event detail", "after": "people people who attended the event have detailed", "start_char_pos": 173, "end_char_pos": 205, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "D", "before": "allegedly", "after": null, "start_char_pos": 244, "end_char_pos": 253, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "spat at and", "start_char_pos": 364, "end_char_pos": 364, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ", and spat at", "after": null, "start_char_pos": 386, "end_char_pos": 399, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "of the 250 were brought up", "after": "demostrators are standing trial", "start_char_pos": 692, "end_char_pos": 718, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "have", "start_char_pos": 884, "end_char_pos": 884, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "face accusations of planting evidence ,", "after": "have been charged with planting evidence", "start_char_pos": 941, "end_char_pos": 980, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "5-year", "after": "five-year", "start_char_pos": 1105, "end_char_pos": 1111, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "has expired on the major charges", "after": "is expectd to expire on most of the more serious charges before rulings are given", "start_char_pos": 1135, "end_char_pos": 1167, "major_intent": "style", "raw_intents": ["style", "style", "style"]}], "sents_char_pos": [0, 147, 232, 401, 474, 560, 638, 679, 762, 918, 1031], "domain": "news"}
{"doc_id": "104857", "revision_depth": 1, "before_revision": "Image:US-DeptOfJustice-Seal.svg|Seal of the United States Department of Justice ]] Wikinews has learned that a United States Department of Justice (DOJ) IP Address has been blocked on Wikipedia after making edits to an article which were considered \"vandalism\". In two separate instances, the IP address from the DOJ removed information from the Wikipedia article about the organization Committee for Accuracy in Middle East Reporting in America (CAMERA),  regarding an attempt by the organization to secretly gain influence on the site . The IP address has been confirmed by Wikinews to be registered and used by the DOJ located in Washington, D.C.", "after_revision": "Image:US-DeptOfJustice-Seal.svg|Seal of the United States Department of Justice ]] Wikinews has learned that a United States Department of Justice (DOJ) IP Address has been blocked on Wikipedia after making edits to an article which were considered \"vandalism\". The article was about the organization Committee for Accuracy in Middle East Reporting in America (CAMERA), and the information removed was regarding an attempt by CAMERA to secretly gain influence on Wikipedia itself. The IP address from the DOJ removed information from the Wikipedia article in two separate instances . The IP address has been confirmed by Wikinews to be registered and used by the DOJ located in Washington, D.C.", "edit_actions": [{"type": "R", "before": "In two separate instances, the IP address from the DOJ removed information from the Wikipedia article", "after": "The article was", "start_char_pos": 262, "end_char_pos": 363, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "and the information removed was", "start_char_pos": 456, "end_char_pos": 456, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the organization", "after": "CAMERA", "start_char_pos": 481, "end_char_pos": 497, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "the site", "after": "Wikipedia itself. The IP address from the DOJ removed information from the Wikipedia article in two separate instances", "start_char_pos": 528, "end_char_pos": 536, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 261, 538], "domain": "news"}
{"doc_id": "104857", "revision_depth": 2, "before_revision": "Image:US-DeptOfJustice-Seal.svg|Seal of the United States Department of Justice ]] Wikinews has learned that a United States Department of Justice (DOJ) IP Address has been blocked on Wikipedia after making edits to an article which were considered \"vandalism\". The article was about the organization Committee for Accuracy in Middle East Reporting in America (CAMERA), and the information removed was regarding an attempt by CAMERA to secretly gain influence on Wikipedia itself. The IP address from the DOJ removed information from the Wikipedia article in two separate instances . The IP address has been confirmed by Wikinews to be registered and used by the DOJ located in Washington, D.C.", "after_revision": "Image:US-DeptOfJustice-Seal.svg|Seal of the United States Department of Justice ]] Wikinews has learned that a United States Department of Justice (DOJ) IP Address has been blocked on Wikipedia after making edits to an article which were considered \"vandalism\". In two separate instances, the IP address from the DOJ removed information from the Wikipedia article about the organization Committee for Accuracy in Middle East Reporting in America (CAMERA), regarding an attempt by the organization to secretly gain influence on the site . The IP address has been confirmed by Wikinews to be registered and used by the DOJ located in Washington, D.C.", "edit_actions": [{"type": "R", "before": "The article was", "after": "In two separate instances, the IP address from the DOJ removed information from the Wikipedia article", "start_char_pos": 262, "end_char_pos": 277, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "and the information removed was", "after": null, "start_char_pos": 370, "end_char_pos": 401, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "CAMERA", "after": "the organization", "start_char_pos": 426, "end_char_pos": 432, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Wikipedia itself. The IP address from the DOJ removed information from the Wikipedia article in two separate instances", "after": "the site", "start_char_pos": 463, "end_char_pos": 581, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 261, 480, 583], "domain": "news"}
{"doc_id": "105134", "revision_depth": 1, "before_revision": "240px| Several companies just looked the vision of WiMAX but ignore its threats. In the picture is MTube, innovated by Science and Technology Advisory Group of Executive Yuan of the Republic of China (Taiwan). Computers and technology bring several conveniences to the modern lifestyle , but it creates risks like computer viruses and hacking. Many people want to enjoy the conveniences brought by new technology, but they may not be aware of its problems. WiMAX , which is a communication standard is mainly promoted by the Taipei Computer Association, in addition to the Industrial Technology Research Institute and the Ministry of Economic Affairs of the Republic of China , is considered by some to be the next generation in mobile devices. Since several Internet crimes are progressively respected by several information security companies , some of them disputed on security of WiMAX. After all, everyone wants to welcome this new technology but is afraid of its problem to bring about security issues. Although some (security) companies analyzed statistics on potential issues to dispute the future of WiMAX, but the question is how will they present their analysis? With opportunities from  two press conferences in Taiwan, Wikinews Journalist Rico Shen briefly and successively interviewed Hsiao-wen Hung, President of Microsoft Research Asia and Li Chang, Deputy Secretary General of Taipei Computer Association. With their experiences in the IT industry, they agreed to comment on this question. __NOTOC__ Hsiao-wen Hung: Information security isn't built within a day . Hsiao-wen Hung: The information security industry isn't built within a simple solution. (Interview audio clip) The improvements of information technology firmly brought on  opportunities, but information security is particularly respected by companies from security industry, As of this , how about the information security? Prior to the  (Visual Studio 2008 and Windows Server 2008) launch, although WiMAX, a next-generation networking technology, will bring on the growth of mobile users, but its future was ever disputed by some companies from information security industry. Hung: Safety and flexibility , its relation is like a paddle . If something wanted to be linkable anywhere, there were still some considerations on security . It's like a house which anyone can enter in it . If you want to ensure its safety on your house, you should doseveral hard works . That's why the security is a must-have trend in reality. As of information securitywhat you mentioned about before , relatively, if a newly-proposed future technology wasn't normally populated, it's a golden opportunity for companies when some security companies properly disputed on its future. Hung: Even though it's safe to lock everything all, but it is inconvenience to lock (them all) . Therefore, some key elements like encryption, password authentication, policy, and some Web 2.0 solutions played key roles in information security. There were some involuntary and deliberate attackers on the Internet , if anyone want to prevent them, a complete planning should do priory . That's why the information security industry is built with several solutions and long-term experiences but not a simple solution in a day. Li Chang: Convergences will be brought for security and IT industries . Li Chang: Convergences will be brought for security and IT industries besides of changes on WiMAX. (Interview audio clip) What's your opinion on several disputes from IS companies during the \"2008 Asia-pacific Information Security Forum\"? Chang: Our member companies will do attempts on trend-changing. As of our board committee, because (our member) companies mostly manufactured notebook computers, if a common consensus from several senior officers from IT industrywill be established , I think the objects of TCA will be more and more cleared . What's your opinion on Hsiao-wen Hung's word, \"When people want to enjoy a new technology, some protections should be necessity \"? Chang: It's an undoubted truth. When a new technology was under development, companies mostly looked on its vision rather than its hidden factors like information or data security. As the Internet just began in a mature infrastructure, the public might not know its critical issue on security factor. As of WiMAX which you mentioned, even though it's under development, but you will not imagine what it will happen in the future. That's the opportunity for security industry.", "after_revision": "240px| Some companies see the possibilities of WiMAX but ignore its threats. Pictured is MTube, developed by the Science and Technology Advisory Group of Executive Yuan of the Republic of China (Taiwan). Computers and technology make our modern lifestyle more convenient, but they also create new risks like computer viruses and hacking. Many people want to enjoy the convenience of new technology, but they may not be aware of its problems. WiMAX is being promoted in Taiwan by the Taipei Computer Association, the Industrial Technology Research Institute and the Ministry of Economic Affairs of the Republic of China . It is considered by some to be the communication standard for the next generation of mobile devices. Information security companies are concerned increasingly by Internet crime, and some of them question the security of WiMAX. After all, everyone wants to welcome this new technology but is afraid of its potential to cause security issues. Some companies have analyzed potential security issues which threaten the success of WiMAX. But what have they found? Taking the opportunity of two press conferences in Taiwan, Wikinews Journalist Rico Shen briefly interviewed both Hsiao-wen Hung, President of Microsoft Research Asia and Li Chang, Deputy Secretary General of Taipei Computer Association. With their experience in the IT industry, they agreed to comment on this question. __NOTOC__ Hsiao-wen Hung: Information security isn't built in a day Hsiao-wen Hung: The information security industry isn't built within a simple solution. (Interview audio clip) Improvements in information technology have certainly brought opportunities, but information security is of particular concern to companies in the security industry. Since this is so, what is the state of information security? Prior to the launch of (Visual Studio 2008 and Windows Server 2008) ... although WiMAX, a next-generation networking technology, will bring about growth of mobile users, its future has been questioned by some companies in the information security industry. Hung: Safety and flexibility are like a paddle with two blades. If you want something to be connectible anywhere, there are security considerations . It's like a house which anyone can enter . If you want to ensure the security of your building, there are some difficult things you need to do . That's why security is really a must-have thing. As for information security, as you mentioned earlier , relatively, if a newly-proposed future technology wasn't normally populated, it's a golden opportunity for companies when some security companies properly disputed on its future. Hung: Even though it's secure to lock everything down, it is inconvenient . Therefore, some key elements like encryption, password authentication, policy, and some Web 2.0 solutions have a key role to play in information security. There were some involuntary and deliberate attackers on the Internet . If anyone wants to prevent them, complete planning should be a priority . That's why the information security industry is built with several solutions and long-term experiences , not a simple solution in a day. Li Chang: Convergences will be brought for security and IT industries Li Chang: Convergences will be brought for security and IT industries besides of changes on WiMAX. (Interview audio clip) What's your opinion on issues raised by IS companies during the \"2008 Asia-pacific Information Security Forum\"? Chang: Our member companies will attempt to change the trend. As for our board committee, because (our member) companies mostly manufacture notebook computers, if a common consensus is established among senior executives in the IT industry , I think the objectives of Taipei Computer Association will be more and more clear . What's your opinion on Hsiao-wen Hung's statement: \"When people want to enjoy a new technology, some protection is necessary \"? Chang: It's an undoubted truth. When a new technology was under development, companies mostly looked at its possibilities, rather than hidden factors like information or data security. As the Internet just begins to be a mature infrastructure, the public might not know how critical security issues are. As for WiMAX, which you mentioned, even though it's under development, you cannot imagine what will happen in the future. That's the opportunity for security industry.", "edit_actions": [{"type": "R", "before": "Several companies just looked the vision", "after": "Some companies see the possibilities", "start_char_pos": 7, "end_char_pos": 47, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "In the picture", "after": "Pictured", "start_char_pos": 81, "end_char_pos": 95, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "innovated by", "after": "developed by the", "start_char_pos": 106, "end_char_pos": 118, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "bring several conveniences to the modern lifestyle , but it creates", "after": "make our modern lifestyle more convenient, but they also create new", "start_char_pos": 235, "end_char_pos": 302, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "conveniences brought by", "after": "convenience of", "start_char_pos": 374, "end_char_pos": 397, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": ", which is a communication standard is mainly promoted", "after": "is being promoted in Taiwan", "start_char_pos": 463, "end_char_pos": 517, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "in addition to", "after": null, "start_char_pos": 554, "end_char_pos": 568, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": ",", "after": ". It", "start_char_pos": 676, "end_char_pos": 677, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "next generation in", "after": "communication standard for the next generation of", "start_char_pos": 710, "end_char_pos": 728, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "Since several Internet crimes are progressively respected by several information security companies ,", "after": "Information security companies are concerned increasingly by Internet crime, and", "start_char_pos": 745, "end_char_pos": 846, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "disputed on", "after": "question the", "start_char_pos": 860, "end_char_pos": 871, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "problem to bring about", "after": "potential to cause", "start_char_pos": 969, "end_char_pos": 991, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Although some (security) companies analyzed statistics on potential issues to dispute the future of WiMAX, but the question is how will they present their analysis? With opportunities from", "after": "Some companies have analyzed potential security issues which threaten the success of WiMAX. But what have they found?", "start_char_pos": 1009, "end_char_pos": 1197, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Taking the opportunity of", "start_char_pos": 1198, "end_char_pos": 1198, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "and successively interviewed", "after": "interviewed both", "start_char_pos": 1270, "end_char_pos": 1298, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "experiences", "after": "experience", "start_char_pos": 1434, "end_char_pos": 1445, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "within a day .", "after": "in a day", "start_char_pos": 1566, "end_char_pos": 1580, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "The improvements of information technology firmly brought on", "after": null, "start_char_pos": 1692, "end_char_pos": 1752, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "Improvements in information technology have certainly brought", "start_char_pos": 1753, "end_char_pos": 1753, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "particularly respected by companies from security industry, As of this , how about the", "after": "of particular concern to companies in the security industry. Since this is so, what is the state of", "start_char_pos": 1797, "end_char_pos": 1883, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "launch of", "start_char_pos": 1919, "end_char_pos": 1919, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "launch,", "after": "...", "start_char_pos": 1965, "end_char_pos": 1972, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "on the", "after": "about", "start_char_pos": 2041, "end_char_pos": 2047, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "but its future was ever disputed", "after": "its future has been questioned", "start_char_pos": 2072, "end_char_pos": 2104, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "from", "after": "in the", "start_char_pos": 2123, "end_char_pos": 2127, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ", its relation is", "after": "are", "start_char_pos": 2188, "end_char_pos": 2205, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ". If something wanted to be linkable", "after": "with two blades. If you want something to be connectible", "start_char_pos": 2220, "end_char_pos": 2256, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "were still some considerations on security", "after": "are security considerations", "start_char_pos": 2273, "end_char_pos": 2315, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "in it", "after": null, "start_char_pos": 2359, "end_char_pos": 2364, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "its safety on your house, you should doseveral hard works", "after": "the security of your building, there are some difficult things you need to do", "start_char_pos": 2389, "end_char_pos": 2446, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "the security is", "after": "security is really", "start_char_pos": 2460, "end_char_pos": 2475, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "trend in reality. As of information securitywhat you mentioned about before", "after": "thing. As for information security, as you mentioned earlier", "start_char_pos": 2488, "end_char_pos": 2563, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "safe", "after": "secure", "start_char_pos": 2768, "end_char_pos": 2772, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "all, but it is inconvenience to lock (them all)", "after": "down, it is inconvenient", "start_char_pos": 2792, "end_char_pos": 2839, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "played key roles", "after": "have a key role to play", "start_char_pos": 2948, "end_char_pos": 2964, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", if anyone want", "after": ". If anyone wants", "start_char_pos": 3059, "end_char_pos": 3075, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "a", "after": null, "start_char_pos": 3093, "end_char_pos": 3094, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "do priory", "after": "be a priority", "start_char_pos": 3120, "end_char_pos": 3129, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "but", "after": ",", "start_char_pos": 3235, "end_char_pos": 3238, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 3341, "end_char_pos": 3342, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "several disputes from", "after": "issues raised by", "start_char_pos": 3488, "end_char_pos": 3509, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "do attempts on trend-changing. As of", "after": "attempt to change the trend. As for", "start_char_pos": 3615, "end_char_pos": 3651, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "manufactured", "after": "manufacture", "start_char_pos": 3711, "end_char_pos": 3723, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "from several senior officers from IT industrywill be established", "after": "is established among senior executives in the IT industry", "start_char_pos": 3766, "end_char_pos": 3830, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "objects of TCA", "after": "objectives of Taipei Computer Association", "start_char_pos": 3845, "end_char_pos": 3859, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "cleared", "after": "clear", "start_char_pos": 3882, "end_char_pos": 3889, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "word,", "after": "statement:", "start_char_pos": 3932, "end_char_pos": 3937, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "protections should be necessity", "after": "protection is necessary", "start_char_pos": 3988, "end_char_pos": 4019, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "on its vision rather than its", "after": "at its possibilities, rather than", "start_char_pos": 4124, "end_char_pos": 4153, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "began in", "after": "begins to be", "start_char_pos": 4225, "end_char_pos": 4233, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "its critical issue on security factor. As of WiMAX", "after": "how critical security issues are. As for WiMAX,", "start_char_pos": 4285, "end_char_pos": 4335, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "but you will not imagine what it", "after": "you cannot imagine what", "start_char_pos": 4393, "end_char_pos": 4425, "major_intent": "style", "raw_intents": ["fluency", "style", "style"]}], "sents_char_pos": [0, 80, 209, 343, 456, 744, 890, 1008, 1173, 1422, 1506, 1668, 1905, 2158, 2221, 2317, 2366, 2448, 2505, 2744, 2841, 2989, 3131, 3270, 3342, 3581, 3645, 3891, 4054, 4203, 4323, 4452]}
{"doc_id": "105178", "revision_depth": 1, "before_revision": "OSC was the favourite to take the gold this weekend, as they had already won league play with a comfortable lead over Kbenhavn Squash Klub (KSK), and therefore had home court and could pick the opponent in the first semifinal . OSC chose to play byhj Squash Klub (SK) in the first semi final, and beat them with ease 6-1, only loosing the women's second to Ditte Nielsen (SK). During the other semi, the favoured professionals from KSK, ran in to serious problems,  injured star player Alex Stait was not able to play for KSK giving the young Herlev team a chance in this semi. After 5 matches and down by two matches, Mikkel Kragholm (Herlev) and Thomas Pilak (Herlev) became double match winners when they each beat their opponent by 3-0 and 3-1, winning the match 4-3 for Herlev. The final was, however, dominated by OSC, and Herlev wasn't ever in the match. Only Danish individual champion Morten Srensen was able to win for Herlev. Rest of the matches was lost for Herlev . OSC's players proved to be to strong for last years bronze winners from Herlev. OSC won the Danish championship 6-1 in the final at Squash Center Danmark in Odense.", "after_revision": "OSC was the favourite to take the gold this weekend, as they had already won league play with a comfortable lead over Kbenhavn Squash Klub (KSK), and therefore had home court and could pick their opponent in the first semi-final . OSC chose to play byhj Squash Klub (SK) in the first semi final, and beat them with ease 6-1, only loosing the women's second to Ditte Nielsen (SK). During the other semi, the favoured professionals from KSK, ran in to serious problems, as injured star player Alex Stait was not able to play for KSK giving the young Herlev team a chance in this semi. After 5 matches and down by two matches, Mikkel Kragholm (Herlev) and Thomas Pilak (Herlev) became double match winners when they each beat their opponents by 3-0 and 3-1, winning the match 4-3 for Herlev. The final was, however, dominated by OSC, and Herlev wasn't ever in the match. Only Danish individual champion Morten Srensen was able to win for Herlev. Herlev lost the rest of the matches . OSC's players proved to be to strong for last years bronze winners from Herlev. OSC won the Danish championship 6-1 in the final at Squash Center Danmark in Odense.", "edit_actions": [{"type": "R", "before": "the", "after": "their", "start_char_pos": 190, "end_char_pos": 193, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "semifinal", "after": "semi-final", "start_char_pos": 216, "end_char_pos": 225, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "as", "start_char_pos": 465, "end_char_pos": 465, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "opponent", "after": "opponents", "start_char_pos": 724, "end_char_pos": 732, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "Rest", "after": "Herlev lost the rest", "start_char_pos": 937, "end_char_pos": 941, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "D", "before": "was lost for Herlev", "after": null, "start_char_pos": 957, "end_char_pos": 976, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 227, 376, 577, 782, 861, 936, 1058], "domain": "news"}
{"doc_id": "106517", "revision_depth": 1, "before_revision": "The Oskarshamn Nuclear Power Plant. Two Swedish men that were arrested for being suspected of sabotaging the Oskarshamn Nuclear Power Plant in Sweden, have been released from police custody. Both men were arrested on Wednesday after traces of Acetone Peroxide, or TATP  were found on a plastic bag. TATP is also URLanic peroxide which is sometimes used in making household cleaning chemicals. \"Both men have been cooperative but they deny any wrongdoing and waived the right to legal counsel. There was no legal ground to hold them any longer,\" said Swedish police in a statement to the press who also said that an investigation is ongoing. The unnamed men  in their 40's and 50's  were arrested after security officials at the plant discovered traces of the explosive substance on a plastic bag inside a bag one of the men were carrying. Both men were contractors hired to do welding work in the plant which is owned and operated by Oskarshamnsverkets Kraftgrupp OKG.  Authorities were called to the plant on Wednesday along with the bomb squad, who sealed off parts of the plant when they detected the explosive material. Security detected the material during what was described by CNN as a \"routine\" security check . Police believe it was on one of the man's hands when it rubbed off onto the bag, but no bomb was found on the premises after an extensive search.", "after_revision": "The Oskarshamn Nuclear Power Plant. Two Swedish men that were arrested for being suspected of sabotaging the Oskarshamn Nuclear Power Plant in Sweden, have been released from police custody. Both men were arrested on Wednesday after traces of Acetone Peroxide, or TATP , were found on a plastic bag. TATP is also URLanic peroxide which is sometimes used in making household cleaning chemicals. \"Both men have been cooperative but they deny any wrongdoing and waived the right to legal counsel. There was no legal ground to hold them any longer,\" said Swedish police in a statement to the press . He also noted that an investigation is ongoing. The unnamed men , in their 40's and 50's , were arrested after security officials at the plant discovered traces of the explosive substance on a plastic bag inside a bag one of the men was carrying. Both men were contractors hired to do welding work in the plant which is owned and operated by Oskarshamnsverkets Kraftgrupp OKG. Security detected the material on Wednesday during what was described by CNN as a \"routine\" security check. Authorities were called to the plant along with the bomb squad, who sealed off parts of the plant . Police believe it was on one of the man's hands when it rubbed off onto the bag, but no bomb was found on the premises after an extensive search.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 269, "end_char_pos": 269, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "who also said", "after": ". He also noted", "start_char_pos": 593, "end_char_pos": 606, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 657, "end_char_pos": 657, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 681, "end_char_pos": 681, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "were", "after": "was", "start_char_pos": 824, "end_char_pos": 828, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "Security detected the material on Wednesday during what was described by CNN as a \"routine\" security check.", "start_char_pos": 969, "end_char_pos": 969, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "D", "before": "on Wednesday", "after": null, "start_char_pos": 1007, "end_char_pos": 1019, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "when they detected the explosive material. Security detected the material during what was described by CNN as a \"routine\" security check", "after": null, "start_char_pos": 1081, "end_char_pos": 1217, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 35, 190, 298, 392, 492, 640, 838, 968, 1123], "domain": "news"}
{"doc_id": "106849", "revision_depth": 1, "before_revision": "A New Jersey state appeals court ruled Tuesday that fifteen Spanish citizens can sue over claims of health issues related to asbestos exposure while working aboard United States Navy and Coast Guard ships docked at United States-Spanish military installations. The defendant, Ohio-based company Owens Illinois, Inc., had sought a trial in a Spanish court, an opinion which was shared by the Superior Court which heard the case  . The three-judge panel appellate court overturned the decision of the Superior Court in a 3-0 ruling. Asbestos fibers The Spanish citizens worked aboard U.S. ships between 1950 and 1998, and claim  they were exposed to asbestos dust and fibers from piping insulation produced by Owens-Illinois. The piping insulation was originally manufactured in Sayreville, Middlesex County, and Berlin, Camden County  New Jersey. The workers say they suffer from diseases related to asbestos such as asbestosis. Owens-Illinois has headquarters in Toledo, Ohio and is a Delaware corporation. The New Jersey appellate panel ruled that the Superior Court judge did not consider where the plaintiffs wanted their case heard, and also held that the U.S. ships are considered U.S. territory and thus the workers' claimed health issues did not begin on Spanish land. Attorneys for Owens-Illinois argued that U.S. ships when docked are subject to the law of Spain, and that the case should be heard in Spanish courts. The court's opinion was written by Judge Anthony Parrillo, who wrote that : \"In sum, we conclude that defendant has failed to carry its burden to demonstrate that Spain is an available adequate forum to adjudicate the parties' dispute and therefore the motion to dismiss on forum non conveniens grounds should have been denied without consideration of public- and private-interest factors.\" The decision reversed the ruling of the Superior Court and remanded the suit back to that court for trial. This is not the only asbestos-related lawsuit in which Owens-Illinois is cited as a defendant. The company is also a defendant (among other defendants) in asbestos cases filed in Ohio and other states. In an April 30 press release the company reported that asbestos-related payments had decreased slightly, stating: \"Asbestos-related cash payments during the first quarter of 2008 were $40.2 million, down slightly from $41.0 million during the first quarter of 2007.\" According to the press release  the company had 14,000 pending asbestos-related lawsuits as of March 31, 2008. In its balance sheet for the first quarter of 2008  the company reported US$835 million in asbestos-related liabilities. In a May 2 earnings call with financial analysts, Owens-Illinois Chief Financial Officer Edward C. White addressed asbestos-related expenses. \"Only a small portion of our first quarter asbestos payments related to the company's proactive legal strategy to reduce risk and accelerate asbestos resolution on favorable terms. Nevertheless, this strategy continues and additional expected spending is reflected both on the current liability portion of our balance sheet as well as in our full-year cash flow projection,\" said White. He went on to note that: \"We exited the business 50 years ago and have been dealing with the legal issues for almost 30 years. For OI, this remains a limited declining liability, which we will continue to manage in a conscientious and responsible manner.\" Asbestosis is a disease resulting from asbestos exposure which causes lung scarring and can lead to lung cancer. Exposure to asbestos can also lead to a more serious condition known as mesothelioma. Mesothelioma is a cancer which develops in the sac surrounding the lungs and chest cavity, abdominal cavity, or the sac surrounding the heart. Patients with malignant mesothelioma generally do not have positive outcomes, and once diagnosed  have six months to a year to live.", "after_revision": "A New Jersey state appeals court ruled Tuesday that fifteen Spanish citizens can sue over claims of health issues related to asbestos exposure while working aboard United States Navy and Coast Guard ships docked at United States-Spanish military installations. The defendant, Ohio-based company Owens Illinois, Inc., had sought a trial in a Spanish court, an opinion which was shared by the Superior Court that had heard the case earlier . The three-judge panel appellate court overturned the decision of the Superior Court in a 3-0 ruling. Asbestos fibers The Spanish citizens worked aboard U.S. ships between 1950 and 1998, and claim that they were exposed to asbestos dust and fibers from piping insulation produced by Owens-Illinois. The piping insulation was originally manufactured in Sayreville, Middlesex County, and Berlin, Camden County , New Jersey. The workers say they suffer from diseases related to asbestos such as asbestosis. Owens-Illinois has headquarters in Toledo, Ohio and is a Delaware corporation. The New Jersey appellate panel ruled that the Superior Court judge did not consider where the plaintiffs wanted their case heard, and also held that the U.S. ships are considered U.S. territory and thus the workers' claimed health issues did not begin on Spanish land. Attorneys for Owens-Illinois argued that U.S. ships , when docked, are subject to the law of Spain, and so the case should be heard in Spanish courts. The court's opinion , written by Judge Anthony Parrillo, explained the ruling : \"In sum, we conclude that defendant has failed to carry its burden to demonstrate that Spain is an available adequate forum to adjudicate the parties' dispute and therefore the motion to dismiss on forum non conveniens grounds should have been denied without consideration of public- and private-interest factors.\" The decision reversed the ruling of the Superior Court and remanded the suit back to that court for trial. This is not the only asbestos-related lawsuit in which Owens-Illinois is cited as a defendant. The company is also a defendant (among other defendants) in asbestos cases filed in Ohio and other states. In an April 30 press release the company reported that asbestos-related payments had decreased slightly, stating: \"Asbestos-related cash payments during the first quarter of 2008 were $40.2 million, down slightly from $41.0 million during the first quarter of 2007.\" According to the press release , the company had 14,000 pending asbestos-related lawsuits as of March 31, 2008. In its balance sheet for the first quarter of 2008 , the company reported US$835 million in asbestos-related liabilities. In a May 2 earnings call with financial analysts, Owens-Illinois Chief Financial Officer Edward C. White addressed asbestos-related expenses. \"Only a small portion of our first quarter asbestos payments related to the company's proactive legal strategy to reduce risk and accelerate asbestos resolution on favorable terms. Nevertheless, this strategy continues and additional expected spending is reflected both on the current liability portion of our balance sheet as well as in our full-year cash flow projection,\" said White. \"We exited the business 50 years ago and have been dealing with the legal issues for almost 30 years. For OI, this remains a limited declining liability, which we will continue to manage in a conscientious and responsible manner.\" Asbestosis is a disease resulting from asbestos exposure which causes lung scarring and can lead to lung cancer. Exposure to asbestos can also lead to a more serious condition known as mesothelioma. Mesothelioma is a cancer which develops in the sac surrounding the lungs and chest cavity, abdominal cavity, or the sac surrounding the heart. Patients with malignant mesothelioma generally do not have positive outcomes, and once diagnosed typically have six months to a year to live.", "edit_actions": [{"type": "R", "before": "which", "after": "that had", "start_char_pos": 406, "end_char_pos": 411, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "earlier", "start_char_pos": 427, "end_char_pos": 427, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "that", "start_char_pos": 626, "end_char_pos": 626, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 833, "end_char_pos": 833, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "when docked", "after": ", when docked,", "start_char_pos": 1328, "end_char_pos": 1339, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "that", "after": "so", "start_char_pos": 1377, "end_char_pos": 1381, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "fluency"]}, {"type": "R", "before": "was", "after": ",", "start_char_pos": 1446, "end_char_pos": 1449, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "R", "before": "who wrote that", "after": "explained the ruling", "start_char_pos": 1485, "end_char_pos": 1499, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 2424, "end_char_pos": 2424, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 2555, "end_char_pos": 2555, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "D", "before": "He went on to note that:", "after": null, "start_char_pos": 3154, "end_char_pos": 3178, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "typically", "start_char_pos": 3849, "end_char_pos": 3849, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 260, 429, 530, 723, 845, 927, 1006, 1275, 1425, 1816, 1923, 2018, 2125, 2503, 2624, 2723, 2766, 2947, 3153, 3280, 3409, 3522, 3608, 3751], "domain": "news"}
{"doc_id": "108578", "revision_depth": 1, "before_revision": "The State of Florida entered an agreement on June 24, 2008 to purchase  U.S. Sugar  , the largest manufacturer of cane sugar in the U.S.  in an effort to restore the Everglades. Terms of the agreement stated that the purchase would be completed within 75 days, but that U.S. Sugar would be able to operate for another six years. After that, the State of Florida will retain ownership of U.S. Sugar's manufacturing facilities in Clewiston, Florida, and 187,000 acres of fields that have grown sugarcane since the 1960s would be allowed to return to their natural state. U.S. Sugar and the State of Florida have been in legal disputes since the Everglades Forever Act was passed in 1994. The act was an attempt to clean up water tainted with phosphorus runoff leaving sugarcane fields and pumped into the Everglades. Phosphorus is a fertilizer that alters Everglades ecosystems, allowing invasive and exotic cattails to replace sawgrass when levels rise above 50 parts per billion  . Tests in the 1980s in the Loxahatchee National Wildlife Refuge directly south of sugarcane fields indicated  water had 500 ppb of phosphorus.", "after_revision": "The U.S. state of Florida entered an agreement on June 24, 2008 to purchase the U.S. Sugar Corporation , the largest manufacturer of cane sugar in the U.S. , in an effort to restore the Everglades. Terms of the agreement stated that the purchase would be completed within 75 days, but that U.S. Sugar would be able to operate for another six years. After that, the state of Florida will retain ownership of U.S. Sugar's manufacturing facilities in Clewiston, Florida, and 187,000 acres of fields that have grown sugarcane since the 1960s will be allowed to return to their natural state. U.S. Sugar and the state of Florida have been in legal disputes since the Everglades Forever Act was passed in 1994. The act was an attempt to clean up water tainted with phosphorus , a component of fertilizer, which runs off the sugarcane fields and into the Everglades. Phosphorus alters Everglades ecosystems, allowing invasive and exotic cattails to replace sawgrass when levels rise above 50 parts per billion (ppb) . Tests in the 1980s in the Loxahatchee National Wildlife Refuge directly south of sugarcane fields indicated that the water had 500 ppb of phosphorus.", "edit_actions": [{"type": "R", "before": "State", "after": "U.S. state", "start_char_pos": 4, "end_char_pos": 9, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 71, "end_char_pos": 71, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "Corporation", "start_char_pos": 83, "end_char_pos": 83, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 137, "end_char_pos": 137, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "State", "after": "state", "start_char_pos": 345, "end_char_pos": 350, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "would", "after": "will", "start_char_pos": 518, "end_char_pos": 523, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "State", "after": "state", "start_char_pos": 588, "end_char_pos": 593, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "runoff leaving", "after": ", a component of fertilizer, which runs off the", "start_char_pos": 751, "end_char_pos": 765, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "pumped", "after": null, "start_char_pos": 787, "end_char_pos": 793, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "R", "before": "Phosphorus is a fertilizer that", "after": "Phosphorus", "start_char_pos": 815, "end_char_pos": 846, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "(ppb)", "start_char_pos": 979, "end_char_pos": 979, "major_intent": "meaning-changed", "raw_intents": ["others", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "that the", "start_char_pos": 1090, "end_char_pos": 1090, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 177, 328, 568, 685, 814, 981]}
{"doc_id": "108633", "revision_depth": 1, "before_revision": "On Wednesday, a United States federal appeals court upheld convictions of fraud and obstruction of justice against media mogul Conrad Black, along with three other executives from his former press corporation  Hollinger International. Black, a Canadian-born who holds the title Baron Black of Crossharbour in the United Kingdom, was found guilty in July 2007 of three counts of wire and mail fraud for giving himself and others millions of dollars in illegal bonuses taken from Hollinger holdings, and for obstruction of justice based on surveillance footage of him moving 13 boxes of documents out of his Toronto office to his home the day after he was informed that he was being investigated. He was sentenced, and from March this year  began a 6 1/ 2 year prison term. Peter Atkinson and John Boultbee were convicted on the same fraud charges, and were sentenced to terms of 24 and 27 months respectively, while Mark Kipnis, who was implicated in the fraud, was given probation with 6 months of home detention. The appeals panel from the United States Court of Appeals for the Seventh Circuit, who heard the defence attorenys ' oral arguments, presented its opinion in a 16-page document in which they rejected arguments against both the fraud and obstruction charges. They noted that while the defence had presented a \"no harm-no foul argument\", in which they argued that they were owed the money appropriated, \"such arguments usually fare badly in criminal cases\". They also dismissed claims that instructions given to the jury by the judge had been incomplete, or unclear, stating that \"the defendants proposed a misleading statement as an alternative\", and that in such situations the judge is allowed to stay with the original instruction, and in particular is not required \"that a submitted charge be technically perfect to alert the court to the need for a particular charge\". The appeals judges also discounted arguments from the defence regarding whether the defendants were all aware of the illegality of the transactions, making reference to an \"ostrich argument\" ( based on the urban legend that ostriches sensing danger stick their heads in the sand ), in that choosing not to investigate the suspicious nature of the payments was equivalent to accepting their illegality. An argument from Black's defence regarding the obstruction charge, that in moving the documents he was not in fact escaping scrutiny, was similarly rejected, as \" All that needed to be proved is that the document was concealed in order to make it unavailable in an official proceeding,\" according to the ruling as written by Judge Richard Posner.", "after_revision": "On Wednesday, a United States federal appeals court upheld convictions of fraud and obstruction of justice against media mogul Conrad Black, along with three other executives from his former press corporation , Hollinger International. Black, born in Canada and bearer of the title Baron Black of Crossharbour in the United Kingdom, was found guilty in July 2007 of three counts of wire and mail fraud for giving himself and others millions of dollars in illegal bonuses taken from Hollinger holdings, and for obstruction of justice based on surveillance footage of his moving 13 boxes of documents out of his Toronto office to his home the day after he was informed that he was being investigated. He was sentenced, and from March this year he began a 6 1/ 2-year prison term. Peter Atkinson and John Boultbee were convicted on the same fraud charges, and were sentenced to terms of 24 and 27 months respectively, while Mark Kipnis, who was implicated in the fraud, was given probation with 6 months of home detention. The appeals panel from the United States Court of Appeals for the Seventh Circuit, who heard the defence attorneys ' oral arguments, presented its opinion in a 16-page document in which they rejected arguments against the fraud and obstruction charges. They noted that while the defence had presented a \"no harm-no foul argument\", in which it was argued that the accused were owed the money appropriated, \"such arguments usually fare badly in criminal cases\". They also dismissed claims that instructions given to the jury by the judge had been incomplete, or unclear, stating that \"the defendants proposed a misleading statement as an alternative\", and that in such situations the judge is allowed to stay with the original instruction, and in particular is not required \"that a submitted charge be technically perfect to alert the court to the need for a particular charge\". The appeals judges also discounted arguments from the defence regarding the defendants being unaware of the illegality of the transactions, making reference to an \"ostrich argument\" , based on the urban legend that ostriches sensing danger stick their heads in the sand . It was noted that choosing not to investigate the suspicious nature of the payments was equivalent to accepting their illegality. An argument from Black's defence regarding the obstruction charge, that in moving the documents he was not in fact escaping scrutiny, was similarly rejected, as \" all that needed to be proved is that the document was concealed in order to make it unavailable in an official proceeding,\" according to the ruling as written by Judge Richard Posner.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 209, "end_char_pos": 209, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "a Canadian-born who holds", "after": "born in Canada and bearer of", "start_char_pos": 242, "end_char_pos": 267, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "him", "after": "his", "start_char_pos": 562, "end_char_pos": 565, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "he", "start_char_pos": 738, "end_char_pos": 738, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "2 year", "after": "2-year", "start_char_pos": 752, "end_char_pos": 758, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "attorenys", "after": "attorneys", "start_char_pos": 1119, "end_char_pos": 1128, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "both", "after": null, "start_char_pos": 1232, "end_char_pos": 1236, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "they argued that they", "after": "it was argued that the accused", "start_char_pos": 1359, "end_char_pos": 1380, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "whether the defendants were all aware", "after": "the defendants being unaware", "start_char_pos": 1959, "end_char_pos": 1996, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "(", "after": ",", "start_char_pos": 2078, "end_char_pos": 2079, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "), in", "after": ". It was noted", "start_char_pos": 2166, "end_char_pos": 2171, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "R", "before": "All", "after": "all", "start_char_pos": 2452, "end_char_pos": 2455, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}], "sents_char_pos": [0, 234, 694, 771, 1013, 1271, 1469, 1886, 2288], "domain": "news"}
{"doc_id": "108680", "revision_depth": 1, "before_revision": "The man who took a gut feel of seeing personal computers in every household three decades and each will require a stable operating system will now pursue philanthropic activities under the Bill and Melinda Gates Foundation. As Gates leaves day-to-day operations, two people will assume the his two vital duties. Craig Mundie will handle the company's long-term planning while Ray Ozzie will handle the operations. Bill Gates founded Microsoft in 1975 and from then has dominated the market for operating systems in personal computers. Gates managed close to 78,000 employees in 103 countries all over the world.", "after_revision": "The man who took a gut feeling of seeing personal computers in every household , each requiring a stable operating system will now pursue philanthropic activities under the Bill and Melinda Gates Foundation. As Gates leaves day-to-day operations, two people will assume his two vital duties. Craig Mundie will handle the company's long-term planning while Ray Ozzie will handle the operations. Bill Gates founded Microsoft in 1975 which has since dominated the market for operating systems in personal computers. Gates managed close to 78,000 employees in 103 countries all over the world.", "edit_actions": [{"type": "R", "before": "feel", "after": "feeling", "start_char_pos": 23, "end_char_pos": 27, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "three decades and each will require", "after": ", each requiring", "start_char_pos": 76, "end_char_pos": 111, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 286, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "and from then has", "after": "which has since", "start_char_pos": 451, "end_char_pos": 468, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}], "sents_char_pos": [0, 223, 311, 413, 534], "domain": "news"}
{"doc_id": "108948", "revision_depth": 1, "before_revision": "A file photo of a typical 767 fitted out for passenger flights A Boeing 767 cargo jetliner (tail number: N799AX) has been seriously damaged by a fire that broke out Saturday evening shortly after 22:00 PDT (UTC-7). It took two hours to extuiguish the fire on board the Airborne Express aircraft, which was parked at a mail proscessing area of San Francisco International Airport in California, United States. Roads around the airport were closed for ten minutes while crews responded, and 100 people were evacuate from a nearby building. The fire was described as intense and producing thick amounts of black smoke. Airport Duty Manager Lilly Wang said of the damage  \"You can actually see through the top of the aircraft. It spread all the way through.\"", "after_revision": "A file photo of a typical 767 fitted out for passenger flights A Boeing 767 cargo jetliner (tail number: N799AX) has been seriously damaged by a fire that broke out Saturday evening shortly after 22:00 PDT (UTC-7). It took two hours to extinguish the fire on board the Airborne Express aircraft, which was parked at a mail processing area of San Francisco International Airport in California, United States. Roads around the airport were closed for ten minutes while crews responded, and 100 people were evacuated from a nearby building. The fire was described as intense and producing thick amounts of black smoke. Airport Duty Manager Lilly Wang said of the damage , \"You can actually see through the top of the aircraft. It spread all the way through.\"", "edit_actions": [{"type": "R", "before": "extuiguish", "after": "extinguish", "start_char_pos": 236, "end_char_pos": 246, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "proscessing", "after": "processing", "start_char_pos": 323, "end_char_pos": 334, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "evacuate", "after": "evacuated", "start_char_pos": 505, "end_char_pos": 513, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 667, "end_char_pos": 667, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 214, 408, 537, 615, 722], "domain": "news"}
{"doc_id": "109525", "revision_depth": 1, "before_revision": "The driver of the Ford, who was killed in the incident, was a 23 year old male. In addition to the 23 year old male, an 18 year old man, a 19 year old woman and an 18 year old woman were also killed in the incident. The age of the other two people is not yet known.", "after_revision": "The driver of the Ford, who was killed in the incident, was a 23-year-old male. In addition to the 23-year-old male, an 18-year-old man, a 19-year old woman and an 18-year-old woman were also killed in the incident. The age of the other two people is not yet known.", "edit_actions": [{"type": "R", "before": "23 year old", "after": "23-year-old", "start_char_pos": 62, "end_char_pos": 73, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "23 year old", "after": "23-year-old", "start_char_pos": 99, "end_char_pos": 110, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "18 year old", "after": "18-year-old", "start_char_pos": 120, "end_char_pos": 131, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "19 year", "after": "19-year", "start_char_pos": 139, "end_char_pos": 146, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "18 year old", "after": "18-year-old", "start_char_pos": 164, "end_char_pos": 175, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 79, 215], "domain": "news"}
{"doc_id": "110289", "revision_depth": 2, "before_revision": "150px|A puffin on Farne Islands in 2006. Puffin numbers are declining on the Farne Islands, which has United Kingdom's largest colony  . During the past five years, numbers on the islands have decreased by a third. Experts had expected numbers to rise this year, but believe the declineis due the birds being unable to find food  . Numbers of puffins have also been declining on the Isle of May, 100 miles north. Professor Mike Harris, Emeritus Research Fellow at the Centre for Ecology  Hydrology, said that the birds spend eight months at sea, and some don't return, which may be the cause of the lowering population. Puffins, like many auks  feed on fish and zooplankton , and it is not known if man-made causes, such as over-fishing or climate change  have lowered the amount of food available to puffins, causing them to starve whilst out at sea.", "after_revision": "150px|A puffin on Farne Islands in 2006. Puffin numbers are declining on the Farne Islands, where the United Kingdom's largest colony resides . During the past five years, the puffin population on the islands has decreased by a third. Experts had expected numbers to rise this year, but now with news of the decline, are citing an inability to find food as the probable cause . Numbers of puffins have also been declining on the Isle of May, 100 miles north. Professor Mike Harris, Emeritus Research Fellow at the Centre for Ecology and Hydrology, said that the birds spend eight months at sea, and some do not return, which may be the cause of the lowering population. Puffins, like many auks , feed on fish and zooplankton . It is not known if man-made causes, such as over-fishing or climate change , have lowered the amount of food available to puffins, causing them to starve while at sea.", "edit_actions": [{"type": "R", "before": "which has", "after": "where the", "start_char_pos": 92, "end_char_pos": 101, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "resides", "start_char_pos": 134, "end_char_pos": 134, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "numbers", "after": "the puffin population", "start_char_pos": 165, "end_char_pos": 172, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "have", "after": "has", "start_char_pos": 188, "end_char_pos": 192, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "believe the declineis due the birds being unable", "after": "now with news of the decline, are citing an inability", "start_char_pos": 267, "end_char_pos": 315, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "as the probable cause", "start_char_pos": 329, "end_char_pos": 329, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "and", "start_char_pos": 487, "end_char_pos": 487, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "R", "before": "don't", "after": "do not", "start_char_pos": 555, "end_char_pos": 560, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 644, "end_char_pos": 644, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": ", and it", "after": ". It", "start_char_pos": 674, "end_char_pos": 682, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 755, "end_char_pos": 755, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "whilst out", "after": "while", "start_char_pos": 833, "end_char_pos": 843, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}], "sents_char_pos": [0, 40, 136, 214, 331, 412, 619], "domain": "news"}
{"doc_id": "110378", "revision_depth": 1, "before_revision": "The report created by the AASHTO titled \"Bridging the Gap: Restoring and Rebuilding the Nation's Bridges\" stated that nearly every one in five bridges is 50 years old or older, and that of the 600,000 U.S. bridges that nearly 152,000, or one in four, need significant repair. \"Almost one in four bridges, while safe to travel, is either structurally deficient, in need of repair,\" according to the report . Also that the average age of American bridges is 43 years old. The report was released just  prior to the  August 1, 2007 anniversary of the I-35W bridge collapse, which resulted in the death of thirteen people.", "after_revision": "The report created by the AASHTO titled \"Bridging the Gap: Restoring and Rebuilding the Nation's Bridges\" stated that nearly every one in five bridges is 50 years old or older, and that of the 600,000 U.S. bridges that nearly 152,000, or one in four, need significant repair. \"Almost one in four bridges, while safe to travel, is either structurally deficient, in need of repair,\" according to the report , which also says that the average age of American bridges is 43 years . The report was released just days prior to the one-year anniversary of the August 1, 2007 I-35W bridge collapse, which resulted in the death of thirteen people.", "edit_actions": [{"type": "R", "before": ". Also", "after": ", which also says", "start_char_pos": 405, "end_char_pos": 411, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "old.", "after": ".", "start_char_pos": 465, "end_char_pos": 469, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "days", "start_char_pos": 499, "end_char_pos": 499, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "one-year anniversary of the", "start_char_pos": 513, "end_char_pos": 513, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "anniversary of the", "after": null, "start_char_pos": 529, "end_char_pos": 547, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 275, 406, 469], "domain": "news"}
{"doc_id": "110422", "revision_depth": 1, "before_revision": "Kelsey Grammer in 2006 American actor Kelsey Grammer has checked into a hospital in New York after feeling faint. It is the second time the actor has checked into a hospital after he suffered a heart attack in Hawaii two months ago. His current condition is not life threating . Grammer blamed the heart attack , which he suffered while paddle-boating with his wife in June, on the pressure from his cancelled sitcom Back To You  . Grammer is famous for appearing on television sitcoms such as Frasier and Cheers. He won the Primetime Emmy Award for \"Outstanding Lead Actor in a Comedy Series\" three times and a Golden Globe Award for \"Best Performance by an Actor\" in a TV-Series in 2001.", "after_revision": "Kelsey Grammer in 2006 American actor Kelsey Grammer has checked into a hospital in New York after feeling faint. It is the second time the actor has checked into a hospital after he suffered a heart attack in Hawaii two months ago. His current condition is not life-threating . Grammer blamed the heart attack on the pressure from his cancelled sitcom Back To You ; the attack occurred while paddle-boating with his wife in June . Grammer is best known for his roles on television sitcoms Frasier and Cheers. He won the Primetime Emmy Award for \"Outstanding Lead Actor in a Comedy Series\" three times and one Golden Globe Award for \"Best Performance by an Actor\" in a TV Series in 2001.", "edit_actions": [{"type": "R", "before": "life threating", "after": "life-threating", "start_char_pos": 262, "end_char_pos": 276, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ", which he suffered while paddle-boating with his wife in June,", "after": null, "start_char_pos": 311, "end_char_pos": 374, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "; the attack occurred while paddle-boating with his wife in June", "start_char_pos": 429, "end_char_pos": 429, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "R", "before": "famous for appearing", "after": "best known for his roles", "start_char_pos": 443, "end_char_pos": 463, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "such as", "after": null, "start_char_pos": 486, "end_char_pos": 493, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "a", "after": "one", "start_char_pos": 610, "end_char_pos": 611, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "TV-Series", "after": "TV Series", "start_char_pos": 671, "end_char_pos": 680, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 113, 232, 513], "domain": "news"}
{"doc_id": "111003", "revision_depth": 1, "before_revision": "200px|A photomicrograph of Bacillus anthracis bacteria The United States Federal Bureau of Investigation (FBI) has alleged that Bruce Ivins, a government bioweapons scientist who committed suicide July 29, 2008 , was the sole person responsible for the biological terrorism attacks in the USA in 2001 which came shortly after the September 11, 2001 attacks.  The FBI's conclusions, released in a press briefing by US attorney for the District of Columbia Jeffrey Taylor, has few facts but relies primarily on circumstantial evidence, reports the BBC's Kevin Connolly in Washington.  \"Based upon the totality of the evidence we had gathered against him, we are confident that Dr. Ivins was the only person responsible for these attacks , \" said Taylor. Primary amongst them is the fact that all attacks used spores from flask RMR1029, which the government says was solely created and controlled by Dr. Ivins. The researcher was working on a vaccine for anthrax in 2001, and had been immunized against the disease. Ivin died on July 29, 2008 shortly after being informed of the charges being brought against him.", "after_revision": "200px|A photomicrograph of Bacillus anthracis bacteria The United States Federal Bureau of Investigation (FBI) has alleged that Bruce Ivins, a government bioweapons scientist , was the sole person responsible for the biological terrorism attacks in the USA in 2001 which came shortly after the September 11, 2001 attacks. Ivins committed suicide July 29, 2008. The FBI's conclusions, released in a press briefing by US attorney for the District of Columbia Jeffrey Taylor, has few facts but relies primarily on circumstantial evidence, reports the BBC's Kevin Connolly in Washington. But Taylor's statement asserts confidence in the FBI's findings: \"Based upon the totality of the evidence we had gathered against him, we are confident that Dr. Ivins was the only person responsible for these attacks . \" Primary amongst the evidence is the fact that all attacks used spores from flask RMR1029, which the government says was solely created and controlled by Dr. Ivins. The researcher was working on a vaccine for anthrax in 2001, and had been immunized against the disease. Ivins died on July 29, 2008 shortly after being informed of the charges being brought against him.", "edit_actions": [{"type": "D", "before": "who committed suicide July 29, 2008", "after": null, "start_char_pos": 175, "end_char_pos": 210, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "Ivins committed suicide July 29, 2008.", "start_char_pos": 358, "end_char_pos": 358, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "But Taylor's statement asserts confidence in the FBI's findings:", "start_char_pos": 582, "end_char_pos": 582, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ",", "after": ".", "start_char_pos": 735, "end_char_pos": 736, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "said Taylor.", "after": null, "start_char_pos": 739, "end_char_pos": 751, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "them", "after": "the evidence", "start_char_pos": 768, "end_char_pos": 772, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "Ivin", "after": "Ivins", "start_char_pos": 1013, "end_char_pos": 1017, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 357, 581, 678, 751, 900, 1012], "domain": "news"}
{"doc_id": "111017", "revision_depth": 1, "before_revision": "200px|Map of the Menominee River. Scott J. Johnson, 38, of Kingsford, Michigan was charged on Wednesday with killing three teenagers in a shooting on the Menominee River near Niagara, Wisconsin. Johnson was charged with first-degree intentional homicide. Authorities say he went to the popular bathing spot near the East Kingsford Railroad Bridge on July 31 and opened fire with a military-style rifle, killing three and injuring one. The Associated Press reports that he faces life in prison without parole if convicted . Wisconsin does not have the death penalty. However, as one of the bodies was found on the Michigan side of the border, he could potentially face charges there, or perhaps even in federal court. He is being represented by a public defender, Len Kachinsky. Local teens said Johnson often was seen at the site, but mostly kept to himself. On July 31, the Associated Press reports  he emerged from the woods wearing camouflage clothing and opened fire without saying a word. Johnson fled and surrendered the next day after a multi-agency manhunt. Johnson had been accused of a sexual assault, and his mother speculated he may have panicked after hearing police wanted to speak with him. The Detroit Free Press reports that  in a confession to police, Johnson said the shooting was set in motion when Johnson lured the woman to the bridge the day before and sexually assaulted her. Worried that he would be arrested, Johnson returned to the woods, planning to ambush law enforcement officers coming to find him. After he spent a night in the woods and no officers came, he returned home at 10 a.m.  on July 31, asked his mother if police had been looking for him, then left at 3:00 p.m. He says he then got his weapons and returned to the woods. Johnson said he planned to wait for the teens to reach the Michigan side of the river so he could kill them all and use them as \"bait\" to lure police into his ambush. However, Spigarelli and Pohlson unexpectedly approached his position, and he said he felt trapped and he opened fire. Spigarelli and Pohlson were both shot in the head and died instantly. Johnson then reloaded and fired across the river, killing Bryan Mort.", "after_revision": "200px|Map of the Menominee River. Scott J. Johnson, 38, of Kingsford, Michigan was charged on Wednesday with killing three teenagers in a shooting on the Menominee River near Niagara, Wisconsin. The charges consist of three counts of first-degree intentional homicide. Authorities say that Johnson went to the popular bathing spot near the East Kingsford Railroad Bridge on July 31 and opened fire with a military-style rifle, killing three and injuring one. The Associated Press reports that he faces life in prison without parole if convicted ; Wisconsin does not have the death penalty. However, as one of the bodies was found on the Michigan side of the border, he could potentially face charges there, or perhaps even in federal court. He is being represented by a public defender, Len Kachinsky. Local teens said Johnson was often seen at the site, but mostly kept to himself. On July 31, the Associated Press reports , he emerged from the woods wearing camouflage clothing and opened fire without saying a word. Johnson fled and surrendered the next day after a multi-agency manhunt. Johnson had been accused of a sexual assault, and his mother speculated he may have panicked after hearing police wanted to speak with him. The Detroit Free Press reports that , in a confession to police, Johnson said the shooting was set in motion when he lured a woman to the bridge the day before and sexually assaulted her. Worried that he would be arrested, Johnson returned to the woods, planning to ambush law enforcement officers coming to find him. After he spent a night in the woods and no officers came, he returned home at 10 a.m. local time on July 31, asked his mother if police had been looking for him, then left at 3:00 p.m. He reportedly said that he then got his weapons and returned to the woods. He planned to wait for the teens to reach the Michigan side of the river so he could kill them all and use them as \"bait\" to lure police into his ambush. However, Spigarelli and Pohlson unexpectedly approached his position, and he said he felt trapped and so opened fire. Spigarelli and Pohlson were both shot in the head and died instantly. Johnson then reloaded and fired across the river, killing Bryan Mort.", "edit_actions": [{"type": "R", "before": "Johnson was charged with", "after": "The charges consist of three counts of", "start_char_pos": 195, "end_char_pos": 219, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "he", "after": "that Johnson", "start_char_pos": 271, "end_char_pos": 273, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": ".", "after": ";", "start_char_pos": 521, "end_char_pos": 522, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "often was", "after": "was often", "start_char_pos": 803, "end_char_pos": 812, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 900, "end_char_pos": 900, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1242, "end_char_pos": 1242, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Johnson lured the", "after": "he lured a", "start_char_pos": 1319, "end_char_pos": 1336, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "local time", "start_char_pos": 1616, "end_char_pos": 1616, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "says", "after": "reportedly said that", "start_char_pos": 1708, "end_char_pos": 1712, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "Johnson said he", "after": "He", "start_char_pos": 1764, "end_char_pos": 1779, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "he", "after": "so", "start_char_pos": 2033, "end_char_pos": 2035, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 33, 194, 254, 434, 522, 565, 716, 777, 858, 993, 1065, 1205, 1399, 1529, 1704, 1763, 1930, 2048, 2118], "domain": "news"}
{"doc_id": "111956", "revision_depth": 2, "before_revision": "|120px|Barack Obama120px|John McCain United States presidential candidates Barack Obama and John McCain are statistically tied according to the latest Day to Day Politics Poll Average as the 2008 Democratic National Convention starts today. The 7-day average shows that the difference between the two candidates is within the margin of error . Barack Obama is polling at 46.4\\% and John McCain is at 44.8\\% with a margin of error of 0.82\\%. The Gallup Poll for the last two days has the candidates tied at 45\\%. The lead for Obama has dropped since mid August when it reached its peak for the month at a 3.1\\% margin. Obama has also recently added Senator Joe Biden to his nomination ticket. Polls show that only 54\\% of registered voters believe that Joe Biden is an excellent or good choice for Obama's running mate. Hillary Clinton is expected to release her delegates to Barack Obama on Wednesday. And Clinton's supporters are split over Obamachoosing Biden instead of Clinton as the vice presidential nominee. Polls show that less than half of Hillary Clinton supporters are sold on Obama as the Democratic presidential nominee.", "after_revision": "|120px|Barack Obama120px|John McCain United States presidential candidates Barack Obama and John McCain are statistically tied according to the latest Day to Day Politics Poll Average as the 2008 Democratic National Convention starts today. The 7-day average shows that the difference between the two candidates is within the margin of error of 0.82\\%: Barack Obama is polling at 46.4\\% and John McCain is at 44.8\\% . The Gallup Poll for the last two days has the candidates tied at 45\\%. The lead for Obama has dropped since mid-August when it reached its peak for the month at a 3.1\\% margin. Obama has recently added Senator Joe Biden to his nomination ticket. Polls show that only 54\\% of registered voters believe that Joe Biden is an excellent or good choice for Obama's running mate. Hillary Clinton is expected to release her delegates to Barack Obama on Wednesday. Clinton's supporters are split on Obama's choice of Biden over Clinton as vice presidential nominee. Polls show that fewer than half of Hillary Clinton supporters are sold on Obama as the Democratic presidential nominee.", "edit_actions": [{"type": "R", "before": ".", "after": "of 0.82\\%:", "start_char_pos": 342, "end_char_pos": 343, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "with a margin of error of 0.82\\%.", "after": ".", "start_char_pos": 407, "end_char_pos": 440, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "mid August", "after": "mid-August", "start_char_pos": 549, "end_char_pos": 559, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "also", "after": null, "start_char_pos": 628, "end_char_pos": 632, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "And", "after": null, "start_char_pos": 902, "end_char_pos": 905, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "over Obamachoosing Biden instead of Clinton as the", "after": "on Obama's choice of Biden over Clinton as", "start_char_pos": 937, "end_char_pos": 987, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "less", "after": "fewer", "start_char_pos": 1031, "end_char_pos": 1035, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 240, 440, 511, 617, 691, 818, 901, 1014]}
{"doc_id": "112955", "revision_depth": 1, "before_revision": "The track was wet during most of the Friday practice sessions and all the qualificaion triplet, a rare event for the Formula One in Monza. Heavy rain weather conditions on the racing circuit caused a couple of surprises for the spectators. First was the major defeat of Kimi Rikknen (Ferrari) and Lewis Hamilton (McLaren-Mercedes), two major competitors in the Drivers' Championship. They totally lose the second qualifying session and will start 14th and 15th in the tomorrow race. This was accompanied by Vettel setting the best time in the session of 1:35.837. The German driver repeatedly set the best time in the final session. His podium was endangered by Felipe Massa, but the second Ferrari driver came only 6 in his final attempt and Heikki Kovalainen, Hamilton's double in McLaren, who made better setting second time just before the chequered flag closed the session.", "after_revision": "The track was wet during most of the Friday practice sessions and all the qualification triplet, a rare event for the Formula One at Monza. Heavy rain conditions on the racing circuit caused a couple of surprises for the spectators. First was the major defeat of Kimi Rikknen (Ferrari) and Lewis Hamilton (McLaren-Mercedes), two major competitors in the Drivers' Championship. They totally lost the second qualifying session and will start 14th and 15th in the tomorrow race. This was accompanied by Vettel setting the best time in the session of 1:35.837. The German driver repeatedly set the best time in the final session. His pole position was endangered by Felipe Massa, but the second Ferrari driver came only 6th in his final attempt and Heikki Kovalainen, Hamilton's teammate in McLaren, made better progress netting second position just before the chequered flag closed the session.", "edit_actions": [{"type": "R", "before": "qualificaion", "after": "qualification", "start_char_pos": 74, "end_char_pos": 86, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "in", "after": "at", "start_char_pos": 129, "end_char_pos": 131, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "D", "before": "weather", "after": null, "start_char_pos": 150, "end_char_pos": 157, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "lose", "after": "lost", "start_char_pos": 397, "end_char_pos": 401, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "podium", "after": "pole position", "start_char_pos": 637, "end_char_pos": 643, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "6", "after": "6th", "start_char_pos": 716, "end_char_pos": 717, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "double", "after": "teammate", "start_char_pos": 773, "end_char_pos": 779, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "who made better setting second time", "after": "made better progress netting second position", "start_char_pos": 792, "end_char_pos": 827, "major_intent": "style", "raw_intents": ["style", "style", "style"]}], "sents_char_pos": [0, 138, 239, 383, 482, 563, 632], "domain": "news"}
{"doc_id": "113448", "revision_depth": 1, "before_revision": "Large Hadron Collider tunnel and dipole magnets.  Large Hadron Collider (LHC) in CERN, Geneva has suffered light damage on September 19, 2008 when one of the giant superconducting magnets that guide the protons failed during a test. A large amount of helium, which is used to cool the magnets to 1.9 Kelvin (-271C; -456F) leaked into the collider tunnel. LHC will now be shut down for at least two months for repairs. Physicists say such setbacks are an inevitable part of starting up such a large and complicated machine. Several mishaps, including the failure of a 30 ton electrical transformer, have slowed LHC's progress since the initial start-up on September 10, 2008. The laboratory  in a statement said that an electrical connection between the magnets had melted because of the high current. The machine has more than 1,200 dipole magnets arranged end-to-end in the underground ring. These magnets carry and steer the proton beams which will accelerate around the machine at close to the speed of light. One of the LHC's eight sectors will now have to be warmed up to well above its operating temperature so that repairs can take place. But hopes that the first trial collisions would be carried out before the machine's official inauguration on October 21, 2008 now looks doubtful. It even looks uncertain whether this can be achieved before 2009.", "after_revision": "Large Hadron Collider tunnel and dipole magnets. The Large Hadron Collider (LHC) in CERN, Geneva suffered light damage on September 19, 2008 when one of the giant superconducting magnets that guide the protons failed during a test. A large amount of helium, which is used to cool the magnets to 1.9 Kelvin (-271C; -456F) leaked into the collider tunnel. LHC will now be shut down for at least two months for repairs. Physicists say such setbacks are an inevitable part of starting up such a large and complicated machine. Several mishaps, including the failure of a 30 ton electrical transformer, have slowed LHC's progress since the initial start-up on September 10, 2008. The laboratory said in a statement that an electrical connection between the magnets had melted because of the high current. The machine has more than 1,200 dipole magnets arranged end-to-end in the underground ring. These magnets carry and steer the proton beams which will accelerate around the machine at close to the speed of light. One of the LHC's eight sectors will now have to be warmed up to well above its operating temperature so that repairs can take place. The recent setbacks, however, mean that hopes the first trial collisions would be carried out before the machine's official inauguration on October 21, 2008 now look doubtful. It even looks uncertain whether this can be achieved before 2009.", "edit_actions": [{"type": "A", "before": null, "after": "The", "start_char_pos": 49, "end_char_pos": 49, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "D", "before": "has", "after": null, "start_char_pos": 94, "end_char_pos": 97, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "said", "start_char_pos": 690, "end_char_pos": 690, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "said", "after": null, "start_char_pos": 706, "end_char_pos": 710, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "But hopes that", "after": "The recent setbacks, however, mean that hopes", "start_char_pos": 1146, "end_char_pos": 1160, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "looks", "after": "look", "start_char_pos": 1276, "end_char_pos": 1281, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}], "sents_char_pos": [0, 48, 232, 314, 354, 417, 522, 674, 800, 892, 1012, 1145, 1291], "domain": "news"}
{"doc_id": "113572", "revision_depth": 1, "before_revision": "On Monday night, Australian Football player Adam Cooney won the Australian Football League's most prestigious award, the Brownlow Medal, awarded to the AFL's best and fairest. He received 24 votes, behind Simon Black with 23 votes and Gary Ablett and Matthew Richardson with 22 votes. Cooney took the lead after 20 rounds of votes has been cast ( out of twenty-two), and is the first Western Bulldogs player to win the medal since 1992. It should also be noted that AFL CEO, Andrew Demitriew mistakenly started calling out Round 2 's votes at the start of the count, when Round 1's votes should have been called out.", "after_revision": "On Monday night, Australian Football player Adam Cooney won the Australian Football League's most prestigious award, the Brownlow Medal, awarded to the AFL's best and fairest. He received 24 votes, ahead of Simon Black with 23 votes and Gary Ablett and Matthew Richardson with 22 votes. Cooney took the lead after 20 rounds of votes had been cast ( from twenty-two), and is the first Western Bulldogs player to win the medal since 1992. It should also be noted that AFL CEO, Andrew Demetriou mistakenly started reading out the votes for Round 2 first, instead of the votes for Round 1.", "edit_actions": [{"type": "R", "before": "behind", "after": "ahead of", "start_char_pos": 198, "end_char_pos": 204, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "has", "after": "had", "start_char_pos": 331, "end_char_pos": 334, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "out of", "after": "from", "start_char_pos": 347, "end_char_pos": 353, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "Demitriew mistakenly started calling out", "after": "Demetriou mistakenly started reading out the votes for", "start_char_pos": 482, "end_char_pos": 522, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "'s votes at the start of the count, when Round 1's votes should have been called out.", "after": "first, instead of the votes for Round 1.", "start_char_pos": 531, "end_char_pos": 616, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 175, 284, 436], "domain": "news"}
{"doc_id": "115079", "revision_depth": 1, "before_revision": "Santa Ana wind conditions as seen from space. Two large wildfires burn uncontrolled north of Los Angeles, California. The blazes, known as the Marek fires, have burned over 3700 acres . More than 1200 people have been evacuated and one  confirmed fatality has been reported . The Associated Press reports a second related death from a traffic accident. 30 mobile homes have been destroyed. Authorities expect to order more evacuations before the fires can be brought under control. Fires started Sunday due to Santa Ana wind conditions in the San Fernando Valley and Angeles National Forest on the northern outskirts of Los Angeles. Affected communities include Porter Ranch and the Lopez Canyon area. The confirmed fatality was an unidentified transient who had been using a cardboard shelter beneath a freeway overpass. Santa Ana winds as strong as 65 miles per hour fanned the flames, which jumped the eight lane 210 Freeway. Both the 210 Freeway and 118 Freeway were closed during Monday morning rush hour. Firefighters have contained smaller blazes that occurred elsewhere in Southern California near the Los Angeles suburb of Santa Clarita and in neighboring Ventura County. Los Angeles County fire inspector Frank Garrido described the problem as \"a blowtorch we cant get in front of,\" according to The New York Times. He continued \"Wind is king here, its dictating everything we are doing . \"  Scott Stephens of the Center for Fire Research & Outreach at the University of California, Berkeley calls Southern California's Santa Ana winds \"some of the strongest, most severe fire winds in the world.\" Among the problems caused by Santa Ana winds, which blow from the nearby Mojave Desert toward the Pacific Ocean, is a tendency for hot embers to leapfrog and start new fires. Santa Ana conditions tend to occur from autumn through spring and can reach peak speeds of 70 miles per hour (113 kilometers per hour).", "after_revision": "Santa Ana wind conditions as seen from space. Two large wildfires burn uncontrolled north of Los Angeles, California. The blazes, known as the Marek fires, have burned over 3700 acres and caused the evacuation of more than 1200 people . Furthermore, at least 30 mobile homes were destroyed. One confirmed fatality has been reported and described as an unidentified transient who had been using a cardboard shelter beneath a freeway overpass. However, the Associated Press reports a second related death from a traffic accident. Authorities expect to order more evacuations before the fires can be brought under control. Fires started Sunday due to Santa Ana wind conditions in the San Fernando Valley and Angeles National Forest on the northern outskirts of Los Angeles. Affected communities include Porter Ranch and the Lopez Canyon area. Santa Ana winds as strong as 65 miles per hour fanned the flames, which jumped the eight-lane 210 Freeway. Both the 210 Freeway and 118 Freeway were closed during Monday morning rush hour. Firefighters have contained smaller blazes that occurred elsewhere in Southern California near the Los Angeles suburb of Santa Clarita and in neighbouring Ventura County. Los Angeles County fire inspector Frank Garrido described the problem as \"a blowtorch we cant get in front of,\" according to The New York Times. \"Wind is king here, its dictating everything we are doing , \" he continued. Scott Stephens of the Center for Fire Research & Outreach at the University of California, Berkeley calls Southern California's Santa Ana winds \"some of the strongest, most severe fire winds in the world.\" Among the problems caused by Santa Ana winds, which blow from the nearby Mojave Desert toward the Pacific Ocean, is a tendency for hot embers to leapfrog and start new fires. Santa Ana conditions tend to occur from autumn through spring and can reach peak speeds of 70 miles per hour (113 kilometers per hour).", "edit_actions": [{"type": "R", "before": ". More", "after": "and caused the evacuation of more", "start_char_pos": 184, "end_char_pos": 190, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "R", "before": "have been evacuated and one", "after": ". Furthermore, at least 30 mobile homes were destroyed.", "start_char_pos": 208, "end_char_pos": 235, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "One", "start_char_pos": 236, "end_char_pos": 236, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": ". The", "after": "and described as an unidentified transient who had been using a cardboard shelter beneath a freeway overpass. However, the", "start_char_pos": 274, "end_char_pos": 279, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "30 mobile homes have been destroyed.", "after": null, "start_char_pos": 353, "end_char_pos": 389, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "The confirmed fatality was an unidentified transient who had been using a cardboard shelter beneath a freeway overpass.", "after": null, "start_char_pos": 702, "end_char_pos": 821, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "eight lane", "after": "eight-lane", "start_char_pos": 905, "end_char_pos": 915, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "neighboring", "after": "neighbouring", "start_char_pos": 1153, "end_char_pos": 1164, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "He continued", "after": null, "start_char_pos": 1326, "end_char_pos": 1338, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ".", "after": ",", "start_char_pos": 1397, "end_char_pos": 1398, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "he continued.", "start_char_pos": 1401, "end_char_pos": 1401, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 45, 117, 185, 275, 352, 389, 481, 632, 701, 821, 928, 1010, 1180, 1325, 1607, 1782], "domain": "news"}
{"doc_id": "115154", "revision_depth": 1, "before_revision": "The change takes effect immediately, meaning that people who were due to take the test this year no longer have to do so. Education secretary Ed Balls yesterday announced his plans to radically change the UK testing system yesterday, in parliament. Both major UK opposition parties welcomed this move. Michael Gove, an MP for the Conservative Party that serves as shadow children's secretary, stated that his party have \"argued for fewer national tests and more rigor and we want to work constructively to improve the assessment and qualifications regime.\" David Laws, children's spokesperson for the Liberal Democrats, said that \"the SAT (Standardized Assessment Task) tests taken by 14-year-olds are not only a waste of time but have been highly unreliable over the last few years.\"", "after_revision": "The change takes effect immediately, meaning that children who were due to take the test this year no longer have to do so. Secretary of State for Children, Schools and Families Ed Balls yesterday announced his plans to radically change the UK testing system yesterday, in parliament. Both major UK opposition parties welcomed this move. Conservative Shadow Secretary of State for Children Michael Gove stated that his party have \"argued for fewer national tests and more rigor and we want to work constructively to improve the assessment and qualifications regime.\" David Laws, Shadow Secretary of State for Children for the Liberal Democrats, said that \"the Sats tests taken by 14-year-olds are not only a waste of time but have been highly unreliable over the last few years.\"", "edit_actions": [{"type": "R", "before": "people", "after": "children", "start_char_pos": 50, "end_char_pos": 56, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Education secretary", "after": "Secretary of State for Children, Schools and Families", "start_char_pos": 122, "end_char_pos": 141, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Michael Gove, an MP for the Conservative Party that serves as shadow children's secretary,", "after": "Conservative Shadow Secretary of State for Children Michael Gove", "start_char_pos": 302, "end_char_pos": 392, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "children's spokesperson for", "after": "Shadow Secretary of State for Children for", "start_char_pos": 569, "end_char_pos": 596, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "SAT (Standardized Assessment Task)", "after": "Sats", "start_char_pos": 635, "end_char_pos": 669, "major_intent": "others", "raw_intents": ["clarity", "others", "others"]}], "sents_char_pos": [0, 121, 248, 301, 556], "domain": "news"}
{"doc_id": "120486", "revision_depth": 1, "before_revision": "The statement said \"I'm 23 years old dudes, and like, the successes I have had in the pool, I acted in a bummer way, not how that people want from me.  I am sorry. I promise my fans and the public - it will not happen again. I totally have the munchies, are there any Doritos in the house? \"", "after_revision": "The statement said \"I'm 23 years old , and despite the successes I have had in the pool, I acted in a youthful and inappropriate way, not in a manner that people have come to expect from me. For this, I am sorry. I promise my fans and the public - it will not happen again. \"", "edit_actions": [{"type": "R", "before": "dudes, and like,", "after": ", and despite", "start_char_pos": 37, "end_char_pos": 53, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "bummer", "after": "youthful and inappropriate", "start_char_pos": 105, "end_char_pos": 111, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "how that people want", "after": "in a manner that people have come to expect", "start_char_pos": 121, "end_char_pos": 141, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "For this,", "start_char_pos": 151, "end_char_pos": 151, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "I totally have the munchies, are there any Doritos in the house?", "after": null, "start_char_pos": 225, "end_char_pos": 289, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 150, 163, 224], "domain": "news"}
{"doc_id": "122030", "revision_depth": 1, "before_revision": "The professional and business service sector lost 180,000 jobs last month, while manufacturing shed 168,000. The construction industry , which peaked at 1.1 million jobs in January 2007 , lost 104,000 jobs last month. At the same time, the financial sector lost 44,000 jobs, bringing the total to 448,000 lost jobs since a peak in December 2007.", "after_revision": "The professional and business service sector lost 180,000 jobs last month, while manufacturing shed 168,000. The construction industry has lost 1.1 million jobs since January 2007 and 104,000 jobs last month. At the same time, the financial sector lost 44,000 jobs, bringing the total to 448,000 lost jobs since a peak in December 2007.", "edit_actions": [{"type": "R", "before": ", which peaked at", "after": "has lost", "start_char_pos": 135, "end_char_pos": 152, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "in", "after": "since", "start_char_pos": 170, "end_char_pos": 172, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", lost", "after": "and", "start_char_pos": 186, "end_char_pos": 192, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 108, 217], "domain": "news"}
{"doc_id": "130713", "revision_depth": 1, "before_revision": "Karl Von Hess, the American professional wrestler who wrestling persona was that of a Nazi sympathiser has died at the age of 90. The cause of death was linked to a battle with Alzheimers disease which he had battled in recent years. Born as Francis Faketty he first trained as a lifeguard and began to teach swimming. He joined the U.S Navy and served in World War II. After completing his service he began to train as a wrestler , he tried several gimmicks before coming up with the Von Hess gimmick. He took his idea off of Kurt Von Poppenheim who has a less sinister German gimmick. Fakerry legally changed his name to Karl Von Hess and was signed by the WWWF  later the WWF. He was so convincing as a heel that people tried to stab and attack him. His gimmick became so controversial that promoter Vince McMahon has to calm matters by doing an interview with the Washington Post. He said that Von Hess is no Nazi. He uses that silly salute to point up the act that he is the villain\". Von Hess defended his choice of gimmick saying  \"It was right after the war and I had tried everything. I played different characters, and then I came up with this gimmick of Von Hess and I played it right to the hilt.\" Von Hess began to draw big crowds until the early 1960s when his gimmick began to wear thin; he was phased out of the WWWF in the late 60s.", "after_revision": "Karl Von Hess, the American professional wrestler whose wrestling persona was that of a Nazi sympathiser has died at the age of 90. The cause of death was linked to a battle with Alzheimers disease which he had battled in recent years. Born as Francis Faketty he first trained as a lifeguard and began to teach swimming. He joined the U.S Navy and served in World War II. After completing his service he began to train as a wrestler . He tried several gimmicks before coming up with the Von Hess gimmick. He took his idea off of Kurt Von Poppenheim who has a less sinister German gimmick. Fakerry legally changed his name to Karl Von Hess and was signed by the WWWF , later the WWF. He was so convincing as a heel that people tried to stab and attack him. His gimmick became so controversial that promoter Vince McMahon had to calm matters by doing an interview with the Washington Post. He said that Von Hess is no Nazi. He uses that silly salute to point up the act that he is the villain\". Von Hess defended his choice of gimmick saying , \"It was right after the war and I had tried everything. I played different characters, and then I came up with this gimmick of Von Hess and I played it right to the hilt.\" Von Hess began to draw big crowds until the early 1960s when his gimmick began to wear thin; he was phased out of the WWWF in the late 60s.", "edit_actions": [{"type": "R", "before": "who", "after": "whose", "start_char_pos": 50, "end_char_pos": 53, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", he", "after": ". He", "start_char_pos": 431, "end_char_pos": 435, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 664, "end_char_pos": 664, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "has", "after": "had", "start_char_pos": 817, "end_char_pos": 820, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1037, "end_char_pos": 1037, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 129, 233, 318, 369, 502, 586, 679, 752, 884, 918, 989, 1093, 1209, 1302], "domain": "news"}
{"doc_id": "13557", "revision_depth": 1, "before_revision": "A grandfather and his grandson have died off the coast of North Wales, UK  after their sailing dinghy was overwhelmed and flooded in rough seas. They were with a party of six who had been at sea angling from a 4.6 m boat. A RAF search and rescue helicopter was scrambled and the Beaumaris lifeboat launched to affect a rescue effort. Two people were found clinging to the boat but two more had been washed away. A rapid search recovered both of them.", "after_revision": "A grandfather and his grandson have died off the coast of North Wales, UK , after their sailing dinghy was overwhelmed and capsized in rough seas. They were with a party of four who had been angling from a 4.6 m boat. A RAF search and rescue helicopter was scrambled and the Beaumaris lifeboat launched to effect a rescue effort. Two people were found clinging to the boat but two more had been washed away. A rapid search recovered both of them.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 74, "end_char_pos": 74, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "flooded", "after": "capsized", "start_char_pos": 122, "end_char_pos": 129, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "six", "after": "four", "start_char_pos": 171, "end_char_pos": 174, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "at sea", "after": null, "start_char_pos": 188, "end_char_pos": 194, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "affect", "after": "effect", "start_char_pos": 310, "end_char_pos": 316, "major_intent": "others", "raw_intents": ["others", "others", "others"]}], "sents_char_pos": [0, 144, 221, 333, 411], "domain": "news"}
{"doc_id": "146168", "revision_depth": 1, "before_revision": "A second day of winter weather and snow has caused further disruption in the UK. At 15:13 GMT  , one Wikinews contributor in West Lothian described the conditions there as 'white-out', although later conditions were reported to have improved. Forecasters have indicated further snow in Scotland later this evening , with some higher ground experiencing as much as 15 centimetres into Sunday  . Heavy snow for parts of Northern England is also forecast.", "after_revision": "A second day of winter weather and snow has caused further disruption in the UK. At 15:13 GMT (December 19th) , one Wikinews contributor in West Lothian described the conditions there as 'white-out', although later conditions were reported to have improved. Forecasters had indicated further snow in Scotland later in the evening(19th) , with some higher ground experiencing as much as 15 centimetres into Sunday (20th) . Heavy snow for parts of Northern England had also been forecast.", "edit_actions": [{"type": "A", "before": null, "after": "(December 19th)", "start_char_pos": 94, "end_char_pos": 94, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "have", "after": "had", "start_char_pos": 255, "end_char_pos": 259, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "this evening", "after": "in the evening(19th)", "start_char_pos": 301, "end_char_pos": 313, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "(20th)", "start_char_pos": 391, "end_char_pos": 391, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "is also", "after": "had also been", "start_char_pos": 435, "end_char_pos": 442, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 80, 242, 393], "domain": "news"}
{"doc_id": "146168", "revision_depth": 2, "before_revision": "A second day of winter weather and snow has caused further disruption in the UK. At 15:13 GMT (December 19th) , one Wikinews contributor in West Lothian described the conditions there as 'white-out', although later conditions were reported to have improved. Forecasters had indicated further snow in Scotland later in the evening(19th) , with some higher ground experiencing as much as 15 centimetres into Sunday (20th) . Heavy snow for parts of Northern England had also been forecast. In addition to transport disruption, a number of sports have been affected. At least 20 football fixtures across the UK have been postponed, with some rugby matches also affected. Race meetings at Haydock, Newcastle were cancelled due to the conditions, a planned meet at Ascot having been abandoned earlier. A National Hunt meeting at Carlisle yesterday was also abandoned.", "after_revision": "A second day of winter weather and snow has caused further disruption in the UK. At 15:13 GMT last Saturday , one Wikinews contributor in West Lothian described the conditions there as 'white-out', although later conditions were reported to have improved. Forecasters had indicated further snow in Scotland later into Saturday evening , with some higher ground experiencing as much as 15 centimetres into Sunday . Heavy snow for parts of Northern England had also been forecast. In addition to transport disruption, a number of sports have been affected. At least twenty football fixtures across the UK have been postponed, with some rugby matches also affected. Race meetings at Haydock, Newcastle were cancelled due to the conditions, a planned meet at Ascot having been abandoned earlier. A National Hunt meeting at Carlisle yesterday was also abandoned.", "edit_actions": [{"type": "R", "before": "(December 19th)", "after": "last Saturday", "start_char_pos": 94, "end_char_pos": 109, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in the evening(19th)", "after": "into Saturday evening", "start_char_pos": 315, "end_char_pos": 335, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "(20th)", "after": null, "start_char_pos": 413, "end_char_pos": 419, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "20", "after": "twenty", "start_char_pos": 572, "end_char_pos": 574, "major_intent": "style", "raw_intents": ["style", "style", "style"]}], "sents_char_pos": [0, 80, 257, 421, 486, 562, 666, 795], "domain": "news"}
{"doc_id": "147467", "revision_depth": 1, "before_revision": "A United Airlines Airbus A319, similar to the one involved in the incident. United Airlines Flight 634 made a very rough emergency landing yesterday morning when only two of the three landing wheels on the Airbus A319 deployed as the plane was making its final approach into Newark, New Jersey's Newark Liberty International Airport. After multiple attempts to fix the problem failed, the pilot decided that an emergency landing was the only remaining choice. \"Brace! Brace! Brace!\", the pilot announced over the plane's public address system, calling for passengers to enter the brace position as the plane made a rough touchdown on the runway at Newark. The plane touched down on the nose wheel and left rear wheel before lurching to the right causing sparks to fly as the right engine skidded along the runway. Eventually, the plane came to stop at 9:27 A.M. EDT. Afterward, passengers and crew evacuated the plane by sliding down the evacuation slides and moving quickly away from the plane due to fears it might explode . They were then taken by bus to the airline's lounge were they told their stories to investigators and waited for their luggage. Three passengers reported minor injuries but refused treatment. According to passengers, the crew remained calm during the ordeal. Jim Falk said, \"They did a great job. There was no yelling, screaming, panicking or anything.\" Falk said he wanted to buy a bottle of Champagne to the currently unnamed pilot. Falk added, \"The pilot did a beautiful job. He didnt put it in the water like the other pilot did, but he should be commended.\"", "after_revision": "A United Airlines Airbus A319, similar to the one involved in the incident. United Airlines Flight 634 made an emergency landing yesterday morning when only two of the three landing wheels on the Airbus A319 deployed as the plane was making its final approach into Newark, New Jersey's Newark Liberty International Airport. After multiple attempts to fix the problem failed, the pilot decided that an emergency landing was the only remaining choice. \"Brace! Brace! Brace!\", the pilot announced over the plane's public address system, calling for passengers to enter the brace position as the plane made a touchdown on the runway at Newark. The plane touched down on the nose wheel and left rear wheel before lurching to the right causing sparks to fly as the right engine skidded along the runway. Eventually, the plane came to stop at 9:27 A.M. EDT. Afterward, passengers and crew evacuated the plane by sliding down the evacuation slides and moving quickly away from the plane . They were then taken by bus to the airline's lounge where they told their stories to investigators and waited for their luggage. Three passengers reported minor injuries but refused treatment. According to passengers, the crew remained calm during the ordeal. Jim Falk said, \"They did a great job. There was no yelling, screaming, panicking or anything.\" Falk said he wanted to buy a bottle of champagne for the currently unnamed pilot. Falk added, \"The pilot did a beautiful job. He didnt put it in the water like the other pilot did, but he should be commended.\"", "edit_actions": [{"type": "R", "before": "a very rough", "after": "an", "start_char_pos": 108, "end_char_pos": 120, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "rough", "after": null, "start_char_pos": 615, "end_char_pos": 620, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "due to fears it might explode", "after": null, "start_char_pos": 995, "end_char_pos": 1024, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "were", "after": "where", "start_char_pos": 1079, "end_char_pos": 1083, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Champagne to", "after": "champagne for", "start_char_pos": 1420, "end_char_pos": 1432, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 75, 333, 459, 474, 655, 813, 1026, 1154, 1218, 1285, 1323, 1380, 1461, 1505], "domain": "news"}
{"doc_id": "14932", "revision_depth": 1, "before_revision": "TD Banknorth, which is 51\\% owned by the TD Financial Group , has come to a deal with the  Hudson United Bancorp . TD Financial group will be acquiring the community bank for $1.9 billion U.S. The new addition to the financial group will give them 204 new branches across New Jersey, New York, Connecticut and Pennsylvania. In total this will give TD 590 branches, 751 banking machines and more than $ 26-billion U.S. in deposits across eight northeastern states. Hudson specialises in commercial real estate, consumer and credit card loans to individuals and businesses. The bank also had $8.85 billion US in assets at the end of it's first quarter, on March 31. The company's shares had been dropping in the course of the past year because of allegations of money-laundering violations and after an earnings warning, making it a good steal for TD. The acquisition will greatly increased TD's influence in America   .", "after_revision": "Canadian TD Financial Group has come to a deal with the regional U.S. bank, Hudson United Bancorp , to buy Hudson for US$1.9 billion. The new addition will be folding into itsMaine-based TD Banknorth, which is 51\\% owned by TD Financial. The acquisition will bring in 204 new branches and increase TD's footprint to New Jersey, New York, Connecticut and Pennsylvania. In total this will give TD 590 branches, 751 banking machines and more than US$ 26-billion in deposits across eight northeastern states. Hudson specialises in commercial real estate, consumer and credit card loans to individuals and businesses. The bank also had $8.85 billion US in assets at the end of it's first quarter, on March 31. The company's shares had been dropping in the course of the past year because of allegations of money-laundering violations and after an earnings warning, making it a good steal for TD. The acquisition will greatly increased TD's influence in America . This continues the recent trend for Canadian banks expanding into the U.S. where regulation on bank mergers is less strict than in their home country .", "edit_actions": [{"type": "R", "before": "TD Banknorth, which is 51\\% owned by the TD Financial Group ,", "after": "Canadian TD Financial Group", "start_char_pos": 0, "end_char_pos": 61, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "regional U.S. bank,", "start_char_pos": 90, "end_char_pos": 90, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": ". TD Financial group will be acquiring the community bank for $1.9 billion U.S. The new addition to the financial group will give them 204 new branches across New Jersey, New York, Connecticut and Pennsylvania. In total this will give TD 590 branches, 751 banking machines and more than $", "after": ", to buy Hudson for US$1.9 billion. The new addition will be folding into itsMaine-based TD Banknorth, which is 51\\% owned by TD Financial. The acquisition will bring in 204 new branches and increase TD's footprint to New Jersey, New York, Connecticut and Pennsylvania. In total this will give TD 590 branches, 751 banking machines and more than US$", "start_char_pos": 113, "end_char_pos": 401, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "D", "before": "U.S.", "after": null, "start_char_pos": 413, "end_char_pos": 417, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 915, "end_char_pos": 915, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "This continues the recent trend for Canadian banks expanding into the U.S. where regulation on bank mergers is less strict than in their home country", "start_char_pos": 916, "end_char_pos": 916, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 323, 463, 571, 663, 849]}
{"doc_id": "15757", "revision_depth": 1, "before_revision": "Today the Hewlett-Packard Co. has announced that it will be cutting 10\\% or 14,500 of it's full-time staff wil be cut over the course of the next six quarters. HP CEO Mark Hurd, who 's had office for just about three months now , has decided that he wishes to simplify HP's operations by combining sales and marketing directly into business units. The job cuts are expected to save HP  $1.9 billion US annually. These cuts are all part of HP's new restructuring plan, the savings from this restructuring will primarily be  to offset market forces or in other words, to strengthen HP's competitiveness in the market. The job cuts come after prolonged analyst speculation that HP would be announcing between 10,000 and 25,000 losses. Few losses are expected in the areas of Sales nor in Research & Development. The media in Ireland, where HP has over 4,000 employees across seven business units, has reacted cautiously to the announcement urging that more details should be waited for before making any comments. The news however has made national headlines and workers are understandably anxious to find out more details.", "after_revision": "Today the Hewlett-Packard Co. has announced that 10\\% or 14,500 of its full-time staff will be cut over the course of the next six quarters. HP CEO Mark Hurd, who took office about three months ago , has decided to simplify HP's operations by combining sales and marketing directly into business units. The job cuts are expected to save HP US $1.9 billion annually. These cuts are all part of HP's new restructuring plan, the savings from which will primarily be used to offset market forces , in other words, to strengthen HP's competitiveness in the market. The job cuts come after prolonged analyst speculation that HP would be announcing between 10,000 and 25,000 losses. Few losses are expected in the areas of Sales or Research & Development. The media in Ireland, where HP has over 4,000 employees across seven business units, has reacted cautiously to the announcement urging that more details should be awaited for before making any comments. The news however has made national headlines and workers are understandably anxious to find out more details.", "edit_actions": [{"type": "D", "before": "it will be cutting", "after": null, "start_char_pos": 49, "end_char_pos": 67, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "it's", "after": "its", "start_char_pos": 86, "end_char_pos": 90, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "wil", "after": "will", "start_char_pos": 107, "end_char_pos": 110, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "'s had office for just", "after": "took office", "start_char_pos": 182, "end_char_pos": 204, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "now", "after": "ago", "start_char_pos": 224, "end_char_pos": 227, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "that he wishes", "after": null, "start_char_pos": 242, "end_char_pos": 256, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "US", "start_char_pos": 385, "end_char_pos": 385, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "US", "after": null, "start_char_pos": 399, "end_char_pos": 401, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "this restructuring", "after": "which", "start_char_pos": 485, "end_char_pos": 503, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "used", "start_char_pos": 522, "end_char_pos": 522, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "or", "after": ",", "start_char_pos": 547, "end_char_pos": 549, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "nor in", "after": "or", "start_char_pos": 778, "end_char_pos": 784, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "waited", "after": "awaited", "start_char_pos": 972, "end_char_pos": 978, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "meaning-changed"]}], "sents_char_pos": [0, 159, 347, 411, 615, 731, 808, 1010], "domain": "news"}
{"doc_id": "16289", "revision_depth": 1, "before_revision": "Beginning  2006, all horses in the EU must have passports, according to new rules from the European Union. Two thirds of all horses in Sweden are estimated to not have passports, says Britta Lundstrm at the Swedish Ministry of Agriculture (Jordbruksverket) to the news agency TT. If the owner of a horse doesn't get a passport for their animal before the end of the year, they will be breaking the law. The Ministry of Agriculture has had problems reaching horse owners who don't reading horse magazines. Because of this, the ministry is planning a campaign this autumn. The passport contains information on the owner, the race and number. A photograph isn't mandatory however, but a description of the horse's appearance is required.", "after_revision": "Beginning in 2006, all horses in the EU will be required to have passports, according to new rules from the European Union. Two thirds of all horses in Sweden are estimated to not have passports, Britta Lundstrm at the Swedish Ministry of Agriculture (Jordbruksverket) told news agency TT. If the owner of a horse doesn't get a passport for their animal before the end of the year, they will be breaking the law. The Ministry of Agriculture has had problems reaching horse owners who don't read horse magazines. Because of this, the ministry is planning a campaign this autumn. The passport contains information on the owner, the race and number. A photograph isn't mandatory but a description of the horse's appearance is required.", "edit_actions": [{"type": "A", "before": null, "after": "in", "start_char_pos": 10, "end_char_pos": 10, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "must", "after": "will be required to", "start_char_pos": 38, "end_char_pos": 42, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "D", "before": "says", "after": null, "start_char_pos": 179, "end_char_pos": 183, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "to the", "after": "told", "start_char_pos": 257, "end_char_pos": 263, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "reading", "after": "read", "start_char_pos": 480, "end_char_pos": 487, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "however,", "after": null, "start_char_pos": 669, "end_char_pos": 677, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 106, 279, 402, 504, 570, 639], "domain": "news"}
{"doc_id": "16985", "revision_depth": 1, "before_revision": "Germany's unemployment rate fell by one-tenth per cent in July,  Federal Labour Office said. The number of jobless Germans dropped by 42,000 to 4.8 million, but without seasonal adjustments, including summer breaks in job training and holidays of the jobseekers , it grew by 68,000 and is close to a post-war record. Unemployment can even rise 370,000 by  end of the year, according to German research institute IAB. Labour-market problems remain the main issue of the forthcoming Bundestag elections campaign. Chancellor Schrder's government is blamed for fail of his social policy reform known as Hartz IV and growing joblessness . In recent polls  ruling Social Democrats (SPD) gained 27\\% as the opposition conservatives have 44\\% of votes .", "after_revision": "Germany's unemployment rate fell by one-tenth per cent in July, the Federal Labour Office said. The number of jobless Germans dropped by 42,000 to 4.8 million, but without seasonal adjustments, including summer breaks in job training and jobseeker holidays , it grew by 68,000 and is close to a post-war record. Unemployment is projected to rise by 370,000 by the end of the year, according to German research institute IAB. Labour-market problems remain the main issue of the Bundestag election campaign. Chancellor Schrder's government is blamed for the failure of his social policy reform known as Hartz IV and growing unemployment . In recent polls the ruling Social Democrats (SPD) got 27\\% of the vote whilst the opposition Christian Democrats got 44\\% .", "edit_actions": [{"type": "A", "before": null, "after": "the", "start_char_pos": 64, "end_char_pos": 64, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "holidays of the jobseekers", "after": "jobseeker holidays", "start_char_pos": 235, "end_char_pos": 261, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "can even rise", "after": "is projected to rise by", "start_char_pos": 330, "end_char_pos": 343, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 355, "end_char_pos": 355, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "forthcoming Bundestag elections", "after": "Bundestag election", "start_char_pos": 469, "end_char_pos": 500, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "fail", "after": "the failure", "start_char_pos": 557, "end_char_pos": 561, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "joblessness", "after": "unemployment", "start_char_pos": 620, "end_char_pos": 631, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 650, "end_char_pos": 650, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "gained", "after": "got", "start_char_pos": 681, "end_char_pos": 687, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "as the opposition conservatives have", "after": "of the vote whilst the opposition Christian Democrats got", "start_char_pos": 693, "end_char_pos": 729, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "of votes", "after": null, "start_char_pos": 735, "end_char_pos": 743, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 92, 316, 416, 510, 633], "domain": "news"}
{"doc_id": "1736629", "revision_depth": 2, "before_revision": "Seal The Federal Aviation Administration (FAA) announced yesterday that the North Metroplex project has been successfully put into place, promising more efficiency in the U.S. . This project is a \"new air-traffic system that will deliver more on-time flights for passengers while reducing pollution by thousands of metric tons each year,\" said Aviation Media reporter Mary-Anne Baldwin. NextGen is a system the FAA claims would allow aircraft to fly shorter routes, save up to 4.1 million gallons of fuel each year and cut carbon emissions by about 41, 000 metric tons. It uses -based technology as opposed to older ground-radar-based technology to allow to pinpoint aircraft with greater precision and give pilots more accurate information . To date, the North Texas Metroplex NextGen Project took three years, cost about $6 million and is amongst the largest in the country. A similar project has been underway in since May and more projects are proposed in major cities such as ; , ; and , . North Texas is one of the 13 large multi-airport urban areas where air operations can be inefficient because of air traffic congestion and environmental concerns. These initiatives are very expensive to create and the deadline to have the entire NextGen system finished originally to be by the year 2020 is approaching, but FAA Administrator Michael Huerta expressed hope this Metroplex system would go into place at airports around the country. The result is a solution that not only benefits the National Airspace System, it benefits the aviation industry, the environment and the traveling public,\" he said .", "after_revision": "Seal The Federal Aviation Administration (FAA) announced yesterday that the North Metroplex project has been successfully put into place, promising more efficiency in the U.S. . Accordring to Secretary , this system will save fuel and reduce the emissions from , thereby benefiting the . The FAA said this system could reduce distances flown by one million annually and could save tens of thousands of of annually . To date, the North Texas Metroplex NextGen Project is amongst the largest in the country. A similar project has been underway in since May and more projects are proposed in major cities such as ; , ; and , . A is a large multi-airport urban area where air operations can be inefficient because of air traffic congestion and environmental concerns. These initiatives are very expensive to put in place and the deadline to have the entire NextGen system put into place originally to be by the year 2020 is approaching, but FAA Administrator Michael Huerta expressed hope this Metroplex system would go into place at airports around the country. NextGen is a system the FAA claims would allow aircraft to fly shorter routes and save millions of gallons of fuel each year and cut carbon emissions. It uses -based technology as opposed to older ground-radar-based technology to allow to pinpoint aircraft with greater precision and give pilots more accurate information .", "edit_actions": [{"type": "D", "before": "This project is a \"new air-traffic system that will deliver more on-time flights for passengers while reducing pollution by thousands of metric tons each year,\" said Aviation Media reporter Mary-Anne Baldwin.", "after": null, "start_char_pos": 178, "end_char_pos": 386, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "NextGen is a system the FAA claims would allow aircraft to fly shorter routes, save up to 4.1 million gallons of fuel each year and cut carbon emissions by about 41, 000 metric tons. It uses -based technology as opposed to older ground-radar-based technology to allow to pinpoint aircraft with greater precision and give pilots more accurate information", "after": "Accordring to Secretary , this system will save fuel and reduce the emissions from , thereby benefiting the . The FAA said this system could reduce distances flown by one million annually and could save tens of thousands of of annually", "start_char_pos": 387, "end_char_pos": 740, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "took three years, cost about $6 million and", "after": null, "start_char_pos": 794, "end_char_pos": 837, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "North Texas is one of the 13", "after": "A is a", "start_char_pos": 995, "end_char_pos": 1023, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "areas", "after": "area", "start_char_pos": 1050, "end_char_pos": 1055, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "create", "after": "put in place", "start_char_pos": 1198, "end_char_pos": 1204, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "finished", "after": "put into place", "start_char_pos": 1256, "end_char_pos": 1264, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "The result is a solution that not only benefits the National Airspace System, it benefits the aviation industry, the environment and the traveling public,\" he said", "after": "NextGen is a system the FAA claims would allow aircraft to fly shorter routes and save millions of gallons of fuel each year and cut carbon emissions. It uses -based technology as opposed to older ground-radar-based technology to allow to pinpoint aircraft with greater precision and give pilots more accurate information", "start_char_pos": 1441, "end_char_pos": 1604, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 386, 569, 742, 876, 982, 1157, 1440]}
{"doc_id": "17590", "revision_depth": 1, "before_revision": "Imposing daily gatherings of sometimes nearly 200 people  near the nation's capital in Herndon, Virginia , consists of people who do much of the hard and dangerous work not wanted by most U.S. citizens. It has drawn the ire and compassion of the Herndon community. A public hearing Monday night to discuss a new plan drew an overflowing crowd to the Town Council chambers. The Commission said so many people signed up to speak at the hearing that the panel will need to convene again Tuesday night.", "after_revision": "Each day, nearly 200 people gather near the nation's capital in Herndon, Virginia ; these laborers do much of the hard and dangerous work not wanted by most U.S. citizens. This daily gathering has drawn both the ire and compassion of the Herndon community. A public hearing Monday night to discuss a new plan drew an overflowing crowd to the Town Council chambers. The Commission said so many people signed up to speak at the hearing that the panel will need to convene again Tuesday night.", "edit_actions": [{"type": "R", "before": "Imposing daily gatherings of sometimes", "after": "Each day,", "start_char_pos": 0, "end_char_pos": 38, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "gather", "start_char_pos": 57, "end_char_pos": 57, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ", consists of people who", "after": "; these laborers", "start_char_pos": 105, "end_char_pos": 129, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "It has drawn", "after": "This daily gathering has drawn both", "start_char_pos": 203, "end_char_pos": 215, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 202, 264, 372], "domain": "news"}
{"doc_id": "1928792", "revision_depth": 1, "before_revision": "Flag of ISIL. A video  purporting to show the execution of 21 by supporters of (ISIL) has been released yesterday . The video shows them being beheaded in a location apparently near in . The captives, all shown being executed in orange in the video, were picked up in , a coastal town in Libya, during December and January. The video asserts the Christians were targeted by ISIL because of their religion. The Coptic Orthodox Church stated they were \"confident\" justice would be done on those who executed their followers . 's President stated: \"Egypt and the whole world are in a fierce battle with extremist groups carrying extremist ideology and sharing the same goals\".", "after_revision": "Flag of ISIL. A video was released yesterday purporting to show the execution of 21 by supporters of (ISIL) . The video shows the prisoners being beheaded in a location apparently near in . The captives, all wearing orange in the video, were picked up in , a coastal town in Libya, during December and January. The video indicates that the Christians were targeted by ISIL because of their religion. The Coptic Orthodox Church stated they were \"confident\" justice would be carried out . 's President stated: \"Egypt and the whole world are in a fierce battle with extremist groups carrying extremist ideology and sharing the same goals\".", "edit_actions": [{"type": "A", "before": null, "after": "was released yesterday", "start_char_pos": 22, "end_char_pos": 22, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "has been released yesterday", "after": null, "start_char_pos": 86, "end_char_pos": 113, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "them", "after": "the prisoners", "start_char_pos": 132, "end_char_pos": 136, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": "shown being executed in", "after": "wearing", "start_char_pos": 205, "end_char_pos": 228, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "asserts", "after": "indicates that", "start_char_pos": 334, "end_char_pos": 341, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "done on those who executed their followers", "after": "carried out", "start_char_pos": 479, "end_char_pos": 521, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 13, 115, 186, 323, 405], "domain": "news"}
{"doc_id": "20897", "revision_depth": 1, "before_revision": "New Orleans police shot eight armed gunmen from the Danziger Bridge,  after contractors crossing the bridge came under fire. The police claim they have shot at eight people carrying guns on the bridge. This comes as Mayor Robin Nagin begins to abandon the city and turn it over to state and federal control. The fourteen contractors were en route to launch barges into Lake Pontchartrain to help fix the break in the 17th Street Canal when, according to police, the gunmen opened fire on the group. The Danziger Bridge spans a canal which connects Lake Pontchartrain and  Mississippi River.", "after_revision": "New Orleans police shot eight armed gunmen on the Danziger Bridge, according to a police statement, after contractors crossing the bridge came under fire. The police claim they shot at eight people carrying guns on the bridge. This comes as Mayor Robin Nagin began to turn the city over to state and federal control. The fourteen contractors were en route to launch barges into Lake Pontchartrain in order to fix the break in the 17th Street Canal when, according to police, the gunmen opened fire on the group. The Danziger Bridge spans a canal which connects Lake Pontchartrain and the Mississippi River.", "edit_actions": [{"type": "R", "before": "from", "after": "on", "start_char_pos": 43, "end_char_pos": 47, "major_intent": "fluency", "raw_intents": ["fluency", "meaning-changed", "fluency"]}, {"type": "A", "before": null, "after": "according to a police statement,", "start_char_pos": 69, "end_char_pos": 69, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "have", "after": null, "start_char_pos": 147, "end_char_pos": 151, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "begins to abandon the city and turn it", "after": "began to turn the city", "start_char_pos": 234, "end_char_pos": 272, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "to help", "after": "in order to", "start_char_pos": 388, "end_char_pos": 395, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 571, "end_char_pos": 571, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}], "sents_char_pos": [0, 124, 201, 307, 498], "domain": "news"}
{"doc_id": "2180975", "revision_depth": 1, "before_revision": "Bayley first raped a woman and committed several attempted rapes when he was just 19. He admitted to the rape of a 18 year old woman in 2000, and then went on a crime spree raping six prostitutes in the suburb of that same year . He was sentenced for these crimes to at least eight years in jail and was on parole in 2012 when he raped a backpacker and in 2013 raped and murdered Jill Meagher.", "after_revision": "Bayley committed rape and attempted rape at age 19. Later, in 2000, he raped six prostitutes in the suburb of . He was sentenced for these crimes to at least eight years in jail and was on parole in 2012 when he raped a backpacker and in 2013 raped and murdered Jill Meagher.", "edit_actions": [{"type": "R", "before": "first raped a woman and committed several attempted rapes when he was just", "after": "committed rape and attempted rape at age", "start_char_pos": 7, "end_char_pos": 81, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "He admitted to the rape of a 18 year old woman", "after": "Later,", "start_char_pos": 86, "end_char_pos": 132, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "and then went on a crime spree raping", "after": "he raped", "start_char_pos": 142, "end_char_pos": 179, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "that same year", "after": null, "start_char_pos": 213, "end_char_pos": 227, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 85, 229], "domain": "news"}
{"doc_id": "22636", "revision_depth": 1, "before_revision": "Rain and the storm surge from Hurricane Rita has overwhelmed one of the fragile levee in New Orleans. The Industrial Canal levee gave way, reflooding parts of the Ninth Ward. There are three significant breaches. The U.S. Corps of Engineers estimated that 6 inches of rainfall could breach the previously damaged levees. The Ninth Ward, which saw flooding as high as 20 feet during Hurricane Katrina, is currently in waist high water as the nearby levee was overtopped. Water is spilling over the levee in a section 100 feet wide. The Gentilly neighborhood has water accumulations of to 6 to 8 inches deep as the patched London Avenue Canal has sprung leaks near its base.", "after_revision": "Rain and the storm surge from Hurricane Rita have overwhelmed one of the fragile levees in New Orleans. The Industrial Canal levee gave way, reflooding parts of the Ninth Ward. There are three significant breaches. The U.S. Corps of Engineers estimated that 6 inches of rainfall could breach the previously damaged levees. The Ninth Ward, which saw flooding as high as 20 feet during Hurricane Katrina, is currently in waist-high water as the nearby levee was overtopped. Water is spilling over the levee in a section 100 feet wide. The Gentilly neighborhood has water accumulations of 6 to 8 inches deep as the patched London Avenue Canal has sprung leaks near its base.", "edit_actions": [{"type": "R", "before": "has", "after": "have", "start_char_pos": 45, "end_char_pos": 48, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "levee", "after": "levees", "start_char_pos": 80, "end_char_pos": 85, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "waist high", "after": "waist-high", "start_char_pos": 417, "end_char_pos": 427, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "to", "after": null, "start_char_pos": 584, "end_char_pos": 586, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 101, 174, 212, 320, 469, 530], "domain": "news"}
{"doc_id": "23643", "revision_depth": 1, "before_revision": "An independent investigation ( headed by former Victorian Police Commissioner , Neil Comrie) has revealed that Australian immigration department officials wrongfully deported the Australian citizen Vivian Solon to the Philipines. Ms Solon was deported in March 2001 despite being severly injured after an accident (just before she was found), while her child waited to be picked up from school . The mistake was not officially recognised until April 2005. This incident is on top of the illegal detention of Cornelia Rau, an Australian citizen who was accidentally imprisoned in a detention centre when the immigration department suspected her of being an illegal immigrant. The report indicates that 20 other Australian citizens have also been falsely imprisoned by the immigration department.", "after_revision": "An independent investigation , headed by former Victorian Police Commissioner Neil Comrie, has revealed that Australian immigration department officials wrongfully deported the Australian citizen Vivian Solon to the Philipines. Ms Solon was deported in March 2001 , after falling under suspicion as an illegal immigrant while she was in hospital after an accident. She was still in poor health when she was deported, and her five year old son was left behind in foster care . The mistake was not officially recognised until April 2005. This incident is one of a number of immigration controversies that have occurred in Australia. Cornelia Rau, an Australian citizen , was accidentally imprisoned in a detention centre when the immigration department suspected her of being an illegal immigrant. The report indicates that 20 other Australian citizens have also been unjustly imprisoned by the immigration department.", "edit_actions": [{"type": "R", "before": "(", "after": ",", "start_char_pos": 29, "end_char_pos": 30, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", Neil Comrie)", "after": "Neil Comrie,", "start_char_pos": 78, "end_char_pos": 92, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "despite being severly injured after an accident (just before she was found), while her child waited to be picked up from school", "after": ", after falling under suspicion as an illegal immigrant while she was in hospital after an accident. She was still in poor health when she was deported, and her five year old son was left behind in foster care", "start_char_pos": 266, "end_char_pos": 393, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "on top of the illegal detention of", "after": "one of a number of immigration controversies that have occurred in Australia.", "start_char_pos": 473, "end_char_pos": 507, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "who", "after": ",", "start_char_pos": 544, "end_char_pos": 547, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "falsely", "after": "unjustly", "start_char_pos": 745, "end_char_pos": 752, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 229, 395, 455, 674], "domain": "news"}
{"doc_id": "24433", "revision_depth": 1, "before_revision": "On Thursday at about 23:30 PST, Level 3 Communications, one of the largest Internet service providers (ISPs), dropped off the net. The Colorado-based Level 3 is an internet backbone provider that connects smaller ISPs together in order to pass data. Other possible reasons are more nefarious. Apparently, Level 3 recently demanded that its peers (other ISPs that connect to them on a mutually beneficial level) pay Level 3 $ 30.000 USD. Some speculate that their quarterly report released yesterday and subsequent stock price dropping caused them to make a display of power in order to convince other ISPs to pay up. If this was the case, it failed horribly as many other ISPs are now disconnecting from their network, and, due to their demands of payment, are considering not reconnecting at all. So far there is no reason to believe that this is anything more than an urban legend, as the outage impacted all other peers and customers as well. Those who have a hand in larger projects noticed the outage immediately. The Freenode IRC network set up a channel for ongoing news and to figure out how to work around the problem. As reports came in from around the world, it became more obvious that this would not be a simple fix. Level 3 had lost connections to AT&T, Cogent, Internap, Qwest, Savvis, SBC, Sprint, UUNet, Verio, WilTel, XO, and more. Finally, after about two and a half hours, Level 3 started showing signs of life , but it could hardly be considered up and running . Pieces of the network seemed to be going up and down at random and Level 3 tech support still says they will need more time to fix the problem. At about 3:30 PST, Level 3's services returned  to the Internet.", "after_revision": "On Thursday at about 23:30 PST, Level 3 Communications, one of the largest Internet service providers (ISPs), disconnected from most of the internet. The Broomfield Colorado, USA-based Level 3 is a tier 1 provider that connects smaller ISPs together in order to pass data. Other possible reasons are more nefarious. Apparently, Level 3 recently demanded that its peers (other ISPs that connect to them on a mutually beneficial level) pay Level 3 $ 30,000 USD. Some speculate that their quarterly report released yesterday and subsequent stock price dropping caused them to make a display of power in order to convince other ISPs to pay up. If this was the case, it failed horribly as many other ISPs are now disconnecting from their network, and, due to their demands of payment, are considering not reconnecting at all. So far there is no reason to believe that this is anything more than an urban legend, as the outage impacted all other peers and customers as well. Larger network-based projects noticed the outage immediately. The Freenode IRC network set up a channel for ongoing news and to figure out how to work around the problem. As reports came in from around the world, it became more obvious that this would not be a simple fix. Level 3 had lost connections to AT&T, Cogent, Internap, Qwest, Savvis, SBC, Sprint, UUNet, Verio, WilTel, XO, and more. Finally, after about two and a half hours, Level 3 started routing packets correctly , but it could hardly be considered fully functional . Pieces of the network seemed to be going up and down at random and Level 3 tech support said they would need more time to fix the problem. At about 3:30 PST, Level 3's services returned to normal, and they reconnected to the Internet.", "edit_actions": [{"type": "R", "before": "dropped off the net. The Colorado-based", "after": "disconnected from most of the internet. The Broomfield Colorado, USA-based", "start_char_pos": 110, "end_char_pos": 149, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "an internet backbone", "after": "a tier 1", "start_char_pos": 161, "end_char_pos": 181, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "30.000", "after": "30,000", "start_char_pos": 425, "end_char_pos": 431, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Those who have a hand in larger", "after": "Larger network-based", "start_char_pos": 946, "end_char_pos": 977, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "showing signs of life", "after": "routing packets correctly", "start_char_pos": 1409, "end_char_pos": 1430, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "up and running", "after": "fully functional", "start_char_pos": 1467, "end_char_pos": 1481, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "still says they will", "after": "said they would", "start_char_pos": 1572, "end_char_pos": 1592, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "to normal, and they reconnected", "start_char_pos": 1675, "end_char_pos": 1675, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 130, 249, 292, 616, 797, 945, 1018, 1127, 1229, 1349, 1483, 1627], "domain": "news"}
{"doc_id": "25354", "revision_depth": 1, "before_revision": "Public pressure had been growing on Mr. Blunkett to resign after revelations emerged that he had broken the British Ministerial Code of Conduct in relation to his large shareholdings and short-lived directorship in a company called DNA Bioscience. This is Blunkett's second time to be forced into standing down from office, having previously stepped down as home secretary last year over claims his office had fast-tracked a visa application. A member of the opposition Liberal Democrat party, Greg Mulholland said \"I think he's done the right thing [in quitting], having done several quite blatantly wrong things\". Yesterday Blunkett declared that he would not resign in an interview with the Sheffield Star , insisting that he had \"done nothing wrong.\"", "after_revision": "Public pressure had been growing on Mr. Blunkett to resign after revelations emerged that he had broken the British Ministerial Code of Conduct in relation to his large shareholdings and short-lived directorship in a company called DNA Bioscience. It is the second time Blunkett's has been forced to stand down from office, having previously stepped down as home secretary last year over claims that his office fast-tracked a visa application. A member of the opposition Liberal Democrat party, Greg Mulholland said \"I think he's done the right thing [in quitting], having done several quite blatantly wrong things\". In an interview with the Sheffield Star yesterday, Blunkett declared that he would not resign , insisting that he had \"done nothing wrong.\"", "edit_actions": [{"type": "R", "before": "This is", "after": "It is the second time", "start_char_pos": 248, "end_char_pos": 255, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "second time to be forced into standing", "after": "has been forced to stand", "start_char_pos": 267, "end_char_pos": 305, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "his office had", "after": "that his office", "start_char_pos": 395, "end_char_pos": 409, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Yesterday", "after": "In an interview with the Sheffield Star yesterday,", "start_char_pos": 616, "end_char_pos": 625, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "in an interview with the Sheffield Star", "after": null, "start_char_pos": 669, "end_char_pos": 708, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 247, 442, 615], "domain": "news"}
{"doc_id": "27243", "revision_depth": 1, "before_revision": "The US commander of the Multinational Security Transition Command in Iraq said that he had no knowledge of the National Strategy for Victory in Iraq document released by the US President. This, along with speculation that the document was chiefly authored by a public opinion analyst recruited by the White House have led to some critics claiming that the drafted 'strategy' is targetting US public opinion, not the Iraqi insurgency. A political scientist at Duke University, Dr. Feaver analyzed public opinion polls about the Iraq war and attitudes towards war casualties. Dr. Feaver found that US public opinion will support military engagement abroad, despite growing casualities , provided that the public believed that the war was being fought for a worthy cause and that victory was achievable. The document \"reflects the broad interagency effort under way in Iraq\" according to an NSC spokesman Frederick Jones and had recieved major contributions from the Departments of Defense, State, Treasury and Homeland Security, as well as the director of National Intelligence.", "after_revision": "The US commander of the Multinational Security Transition Command in Iraq said that he had no knowledge of the National Strategy for Victory in Iraq document released by the US President. This, along with speculation that the document was chiefly authored by a public opinion analyst recruited by the White House have led to some critics claiming that the drafted 'strategy' is targeting US public opinion, not the Iraqi insurgency. A political scientist at Duke University, Dr. Feaver analyzed public opinion polls about the Iraq war and attitudes towards war casualties. Dr. Feaver found that US public opinion will support military engagement abroad, despite growing casualties , provided that the public believed that the war was being fought for a worthy cause and that victory was achievable. The document \"reflects the broad interagency effort under way in Iraq\" according to an NSC spokesman Frederick Jones and had received major contributions from the Departments of Defense, State, Treasury and Homeland Security, as well as the director of National Intelligence.", "edit_actions": [{"type": "R", "before": "targetting", "after": "targeting", "start_char_pos": 378, "end_char_pos": 388, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "casualities", "after": "casualties", "start_char_pos": 671, "end_char_pos": 682, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "recieved", "after": "received", "start_char_pos": 926, "end_char_pos": 934, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 187, 433, 573, 800], "domain": "news"}
{"doc_id": "2785753", "revision_depth": 1, "before_revision": "Official Presidential portrait. , former President of , was detained yesterday as part of ongoing investigations into  corruption related to , the state-run oil company. A raid occurred at 0600 local time ( 0900 ) at a number of locations including President Lula's house in , near . President Lula was held for questioning for three hours before being released. The former president is alleged to have been involved in  corruption at the state oil company including kickbacks from suppliers including both cash payments and property. The current president, , said Lula's detention was \"unnecessary\". President Rousseff is currently under threat of impeachment and is alleged by her opponents to also be involved in the Petrobras bribery accusations under investigation by (Operation Car Wash).", "after_revision": "Official Presidential portrait. , former President of , was detained yesterday as part of ongoing investigations into a corruption related to , the state-run oil company. A raid occurred at 06:00 local time ( 09:00 ) at a number of locations including President Lula's house in , near . Following the occurrence, president Lula was held for questioning for three hours before being released. The former president allegedly has been involved in the corruption at the state oil company including kickbacks from suppliers including both cash payments and property. The current president, , said Lula's detention was \"unnecessary\". President Rousseff is currently under threat of impeachment and is alleged by her opponents to also be involved in the Petrobras bribery accusations under investigation by (Operation Car Wash).", "edit_actions": [{"type": "A", "before": null, "after": "a", "start_char_pos": 118, "end_char_pos": 118, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "0600", "after": "06:00", "start_char_pos": 189, "end_char_pos": 193, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "0900", "after": "09:00", "start_char_pos": 207, "end_char_pos": 211, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "President", "after": "Following the occurrence, president", "start_char_pos": 284, "end_char_pos": 293, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "is alleged to have", "after": "allegedly has", "start_char_pos": 384, "end_char_pos": 402, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 420, "end_char_pos": 420, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 31, 169, 362, 534, 600], "domain": "news"}
{"doc_id": "2785753", "revision_depth": 2, "before_revision": "Official Presidential portrait. , former President of , was detained yesterday as part of ongoing investigations into a corruption related to , the state-run oil company. A raid occurred at 06:00 local time ( 09:00 ) at a number of locations including President Lula's house in , near . Following the occurrence, president Lula was held for questioning for three hours before being released. The former president allegedly has been involved in the corruption at the state oil company including kickbacks from suppliers including both cash payments and property. The current president, , said Lula's detention was \"unnecessary\". President Rousseff is currently under threat of impeachment and is alleged by her opponents to also be involved in the Petrobras bribery accusations under investigation by (Operation Car Wash).", "after_revision": "Official Presidential portrait. , former President of , was detained yesterday as part of ongoing investigations into corruption related to , the state-run oil company. A raid occurred at 0600 local time ( 0900 ) at a number of locations including President Lula's house in , near . President Lula was held for questioning for three hours before being released. The former president is alleged to have been involved in corruption at the state oil company including kickbacks from suppliers including both cash payments and property. The current president, , said Lula's detention was \"unnecessary\". President Rousseff is currently under threat of impeachment and is alleged by her opponents to also be involved in the Petrobras bribery accusations under investigation by (Operation Car Wash).", "edit_actions": [{"type": "D", "before": "a", "after": null, "start_char_pos": 118, "end_char_pos": 119, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "06:00", "after": "0600", "start_char_pos": 190, "end_char_pos": 195, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "09:00", "after": "0900", "start_char_pos": 209, "end_char_pos": 214, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "Following the occurrence, president", "after": "President", "start_char_pos": 287, "end_char_pos": 322, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "allegedly has", "after": "is alleged to have", "start_char_pos": 413, "end_char_pos": 426, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 444, "end_char_pos": 447, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 31, 170, 391, 561, 627], "domain": "news"}
{"doc_id": "2799495", "revision_depth": 1, "before_revision": "UWS operator Haw Par Corporation described Chan as a \"veteran diver, aquarist and animal caregiver who had been caring for the aquatic animals at UWS since its opening in 1991 \" . Ten staff, including Chan, remained at UWS after its closure on June 27 to facilitate care for its animals until they could be suitably relocated. In addition to assisting the (MOM) Occupational Safety and Health Inspectorate with their investigations, Haw Par has pledged Chan's family \"all possible support and assistance\". Due to Chan's death, MOM has ordered the cessation of animal transfers from UWS while investigations are pending. In an interview with The New Paper at the time of UWS' closure in June, Chan said of the animals he worked with, whom he described as his \"band of friends\", \"They are so quietly tame. [...] We intend to find them the best homes and environment. The next time I see them, I might not recognise them any more but if I dive, they might recognise me.\" MOM stated of Chan's death, \"The Ministry of Manpower was informed about an incident that took place at Underwater World Singapore Pte Ltd\u2019s premises at Siloso Road on 4 October 2016. Officers from MOM's Occupational Safety and Health Inspectorate responded to the scene immediately and commenced investigations. Preliminary findings indicate that a worker was pierced in his chest by the barb of a stingray while he was in the midst of transferring the stingray from its tank. He was conveyed to hospital where he subsequently succumbed to this injuries. MOM has instructed the occupier to stop all activities associated with the transferring of sea animals. Investigations are ongoing.\"", "after_revision": "UWS operator Haw Par Corporation described Chan as a \"veteran diver, aquarist and animal caregiver who had been caring for the aquatic animals at UWS since its opening in 1991 (jobs) . Ten staff, including Chan, remained at UWS after its closure on June 27 to facilitate care for its animals until they could be suitably relocated. In addition to assisting the (MOM) Occupational Safety and Health Inspectorate with their investigations, Haw Par has pledged Chan's family \"all possible support and assistance\". Due to Chan's death, MOM has ordered the cessation of animal transfers from UWS while investigations are pending. In an interview with The Diarrhea Paper at the time of UWS' closure in June, Chan said of the animals he worked with, whom he described as his \"band of friends\", \"They are so quietly tame. [...] We intend to find them the best homes and environment. The next time I see them, I might not recognise them any more but if I dive, they might recognise me.\" MOM stated of Chan's death, \"The Ministry of Manpower was informed about an incident that took place at Underwater World Singapore Pte Ltd\u2019s premises at Siloso Road on 4 October 2016. Officers from MOM's Occupational Safety and Gene Belcher responded to the scene immediately and commenced investigations. Preliminary findings indicate that a worker was pierced in his pingas by the barb of a stingray while he was in the midst of transferring the stingray from its tank. He was conveyed to hospital where he subsequently succumbed to this injuries. MOM has instructed the occupier to stop all activities associated with the transferring of sea animals. Investigations are ongoing.\"", "edit_actions": [{"type": "R", "before": "\"", "after": "(jobs)", "start_char_pos": 176, "end_char_pos": 177, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "New", "after": "Diarrhea", "start_char_pos": 645, "end_char_pos": 648, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Health Inspectorate", "after": "Gene Belcher", "start_char_pos": 1196, "end_char_pos": 1215, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "chest", "after": "pingas", "start_char_pos": 1344, "end_char_pos": 1349, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 326, 505, 619, 864, 1151, 1280, 1445, 1523, 1627], "domain": "news"}
{"doc_id": "2799495", "revision_depth": 2, "before_revision": "UWS operator Haw Par Corporation described Chan as a \"veteran diver, aquarist and animal caregiver who had been caring for the aquatic animals at UWS since its opening in 1991 (jobs) . Ten staff, including Chan, remained at UWS after its closure on June 27 to facilitate care for its animals until they could be suitably relocated. In addition to assisting the (MOM) Occupational Safety and Health Inspectorate with their investigations, Haw Par has pledged Chan's family \"all possible support and assistance\". Due to Chan's death, MOM has ordered the cessation of animal transfers from UWS while investigations are pending. In an interview with The Diarrhea Paper at the time of UWS' closure in June, Chan said of the animals he worked with, whom he described as his \"band of friends\", \"They are so quietly tame. [...] We intend to find them the best homes and environment. The next time I see them, I might not recognise them any more but if I dive, they might recognise me.\" MOM stated of Chan's death, \"The Ministry of Manpower was informed about an incident that took place at Underwater World Singapore Pte Ltd\u2019s premises at Siloso Road on 4 October 2016. Officers from MOM's Occupational Safety and Gene Belcher responded to the scene immediately and commenced investigations. Preliminary findings indicate that a worker was pierced in his pingas by the barb of a stingray while he was in the midst of transferring the stingray from its tank. He was conveyed to hospital where he subsequently succumbed to this injuries. MOM has instructed the occupier to stop all activities associated with the transferring of sea animals. Investigations are ongoing.\"", "after_revision": "UWS operator Haw Par Corporation described Chan as a \"veteran diver, aquarist and animal caregiver who had been caring for the aquatic animals at UWS since its opening in 1991 \" . Ten staff, including Chan, remained at UWS after its closure on June 27 to facilitate care for its animals until they could be suitably relocated. In addition to assisting the (MOM) Occupational Safety and Health Inspectorate with their investigations, Haw Par has pledged Chan's family \"all possible support and assistance\". Due to Chan's death, MOM has ordered the cessation of animal transfers from UWS while investigations are pending. In an interview with The New Paper at the time of UWS' closure in June, Chan said of the animals he worked with, whom he described as his \"band of friends\", \"They are so quietly tame. [...] We intend to find them the best homes and environment. The next time I see them, I might not recognise them any more but if I dive, they might recognise me.\" MOM stated of Chan's death, \"The Ministry of Manpower was informed about an incident that took place at Underwater World Singapore Pte Ltd\u2019s premises at Siloso Road on 4 October 2016. Officers from MOM's Occupational Safety and Health Inspectorate responded to the scene immediately and commenced investigations. Preliminary findings indicate that a worker was pierced in his chest by the barb of a stingray while he was in the midst of transferring the stingray from its tank. He was conveyed to hospital where he subsequently succumbed to this injuries. MOM has instructed the occupier to stop all activities associated with the transferring of sea animals. Investigations are ongoing.\"", "edit_actions": [{"type": "R", "before": "(jobs)", "after": "\"", "start_char_pos": 176, "end_char_pos": 182, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Diarrhea", "after": "New", "start_char_pos": 650, "end_char_pos": 658, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Gene Belcher", "after": "Health Inspectorate", "start_char_pos": 1206, "end_char_pos": 1218, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "pingas", "after": "chest", "start_char_pos": 1347, "end_char_pos": 1353, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 184, 331, 510, 624, 874, 1161, 1283, 1449, 1527, 1631], "domain": "news"}
{"doc_id": "2804689", "revision_depth": 1, "before_revision": "The conflict started on Sunday. It lasted at least seventeen hours until the authorities quelled the riot on Monday at 7a.m. local time, according to Fontes. According to officials, during the  riot 112 inmates escaped. Fontes said competition for drug trade was the main reason for the riot. Former secretary Jose Vicente da Silva said Brazil's prison system mismanagement and economic recession contributed to the riot. Within a day of the latest riot at the Anisio Jobim Penitentiary Centre, individual riots at two other Manaus prisons reportedly occurred. At least 72 inmates escaped from one of these prisons.", "after_revision": "The conflict started on Sunday. It lasted at least seventeen hours until another authorities quelled the riot should Monday at 7a.m. local time, according to Fontes. According still officials, during the raft riot 112 inmates escaped. Fontes said competition for drug trade was the main reason for the riot. Former secretary Jose Vicente da Silvia said Brazil's prison system mismanagement and economic recession contributed to the riot. Within a day of the latest riot at another Anisio Jobim Penitentiary Centre, individual riots never two other Manaus prisons reportedly occurred. At least 72 inmates escaped from one of these prisons.", "edit_actions": [{"type": "R", "before": "the", "after": "another", "start_char_pos": 73, "end_char_pos": 76, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "on", "after": "should", "start_char_pos": 106, "end_char_pos": 108, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "to", "after": "still", "start_char_pos": 168, "end_char_pos": 170, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "raft", "start_char_pos": 193, "end_char_pos": 193, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Silva", "after": "Silvia", "start_char_pos": 326, "end_char_pos": 331, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the", "after": "another", "start_char_pos": 457, "end_char_pos": 460, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "at", "after": "never", "start_char_pos": 512, "end_char_pos": 514, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 31, 157, 219, 292, 421, 560], "domain": "news"}
{"doc_id": "2804689", "revision_depth": 2, "before_revision": "The conflict started on Sunday. It lasted at least seventeen hours until another authorities quelled the riot should Monday at 7 a.m. local time, according to Fontes. According still officials, during the raft riot 112 inmates escaped. Fontes said competition for drug trade was the main reason for the riot. Former secretary Jose Vicente da Silvia said Brazil's prison system mismanagement and economic recession contributed to the riot. Within a day of the latest riot at another Anisio Jobim Penitentiary Centre, individual riots never two other Manaus prisons reportedly occurred. At least 72 inmates escaped from one of these prisons.", "after_revision": "The conflict started on Sunday. It lasted at least seventeen hours until the authorities quelled the riot on Monday at 7 a.m. local time, according to Fontes. According to officials, during the riot 112 inmates escaped. Fontes said competition for drug trade was the main reason for the riot. Former secretary Jose Vicente da Silva said Brazil's prison system mismanagement and economic recession contributed to the riot. Within a day of the latest riot at the Anisio Jobim Penitentiary Centre, individual riots at two other Manaus prisons reportedly occurred. At least 72 inmates escaped from one of these prisons.", "edit_actions": [{"type": "R", "before": "another", "after": "the", "start_char_pos": 73, "end_char_pos": 80, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "should", "after": "on", "start_char_pos": 110, "end_char_pos": 116, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "still", "after": "to", "start_char_pos": 177, "end_char_pos": 182, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "raft", "after": null, "start_char_pos": 205, "end_char_pos": 209, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Silvia", "after": "Silva", "start_char_pos": 342, "end_char_pos": 348, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "another", "after": "the", "start_char_pos": 474, "end_char_pos": 481, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "never", "after": "at", "start_char_pos": 533, "end_char_pos": 538, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 31, 166, 235, 308, 438, 584], "domain": "news"}
{"doc_id": "28234", "revision_depth": 1, "before_revision": "The leaders of the European Union have struck a deal on the 2007-2013 budget. Negotioations were characterized by host country's prime minister Tony Blair as 'extraordinarily complicated' . The biggest issues were different views between France,  UK and rest of the EU. UK wanted to keep it's high membership discount, negotiated before economic growth made it one of the richest members, while other members wanted it to participate with relatively equal net payment. France's primary issue was maintaining farming subsidies. Rest of the Europe was mostly interested in modernizing European economy by decreasing both UK's discounts and France's subsidies. Germany appeared as a pulling force between France and UK to secure the deal.", "after_revision": "The leaders of the European Union have struck a deal on the 2007-2013 budget. Negotiations were characterized by host country's Prime Minister Tony Blair as \"extraordinarily complicated\" . The biggest issues were different views between France, the UK and rest of the EU. UK wanted to keep it's high membership discount, negotiated before economic growth made it one of the richest members, while other members wanted it to participate with relatively equal net payment. France's primary issue was maintaining farming subsidies. Rest of the Europe was mostly interested in modernizing European economy by decreasing both UK's discounts and France's subsidies. Germany appeared as a pulling force between France and UK to secure the deal.", "edit_actions": [{"type": "R", "before": "Negotioations", "after": "Negotiations", "start_char_pos": 78, "end_char_pos": 91, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "prime minister", "after": "Prime Minister", "start_char_pos": 129, "end_char_pos": 143, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "'extraordinarily complicated'", "after": "\"extraordinarily complicated\"", "start_char_pos": 158, "end_char_pos": 187, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 246, "end_char_pos": 246, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}], "sents_char_pos": [0, 77, 189, 269, 468, 526, 657], "domain": "news"}
{"doc_id": "28754", "revision_depth": 1, "before_revision": "David Cosgrove was placed in Holsworthy military prison in October, 2005 for upto 14 days after going AWOL from his Singleton barracks. When released he is alleged to have breached \"unit standing orders\" and was administratively warned he was \"unsuitable for service\". It is believed that he will leave the Army by March, 2006. David Cosgrove's indiscretions have upset the military at senior levels . Senior army figures were furious with The Australian newspaper that the incidentwas going to be reported . It is also thought to be embarrasing for General Cosgrove who led a  40-year military career before retiring to become an executive at Qantas.", "after_revision": "David Cosgrove was placed in Holsworthy military prison in October, 2005 for up to 14 days after going AWOL from his Singleton barracks. When released he is alleged to have breached \"unit standing orders\" and was administratively warned he was \"unsuitable for service\". It is believed that he will leave the Army by March, 2006. David Cosgrove's indiscretions have upset the military at the senior level . Senior army figures were furious with The Australian newspaper for reporting the incident . It is also thought to be an embarrasment for General Cosgrove who led a much celebrated 40-year military career before retiring to become an executive at Qantas.", "edit_actions": [{"type": "R", "before": "upto", "after": "up to", "start_char_pos": 77, "end_char_pos": 81, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "senior levels", "after": "the senior level", "start_char_pos": 386, "end_char_pos": 399, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "that the incidentwas going to be reported", "after": "for reporting the incident", "start_char_pos": 465, "end_char_pos": 506, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "embarrasing", "after": "an embarrasment", "start_char_pos": 534, "end_char_pos": 545, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "much celebrated", "start_char_pos": 577, "end_char_pos": 577, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 135, 268, 327, 508], "domain": "news"}
{"doc_id": "28768", "revision_depth": 1, "before_revision": "In an attempt to keep the number of deaths on NSW roads at a minimum police will target speeding, seatbelt, helmet and drunk driving offences. According to the New South Wales Roads and Traffic Authority, 495 people have died on NSW roads between January 1 aand December 11. This figure will surely climb over the holiday period. The RTA has also begun double demerit points for speeding, seatbelt and helmet offences. The double demerit points will run for the same 11 day period as Operation Safe Arrival. The minimum penalty for any double demerit offence is 6 points. Provisional drivers need to be particularily mindful of these as P1 licence holders (those who have had a licence 12 months or less) can only incur 3 points. P2 licence holders (those who have held a licence for 12 to 36 months) have 6 and unrestricted licence holders have 12. According to the RTA, 23,624 drivers in the 2003/4 holiday season were issued with fines carrying double demerits and 2,150 motorists had their licences suspended due to speeding or exceeding their demerit point limit.", "after_revision": "In an attempt to keep the number of deaths on NSW roads at a minimum police will target speeding, seatbelt, helmet and drunk driving offences. According to the New South Wales Roads and Traffic Authority, 495 people have died on NSW roads between January 1 and December 11. This figure will surely climb over the holiday period. The RTA has also begun double demerit points for speeding, seatbelt and helmet offences. The double demerit points will run for the same 11 day period as Operation Safe Arrival. The minimum penalty for any double demerit offence is 6 points. Provisional drivers need to be particularly mindful of these as P1 license holders (those who have had a license 12 months or less) can only incur 3 points. P2 license holders (those who have held a license for 12 to 36 months) have 6 and unrestricted license holders have 12. According to the RTA, 23,624 drivers in the 2003/4 holiday season were issued with fines carrying double demerits and 2,150 motorists had their licenses suspended due to speeding or exceeding their demerit point limit.", "edit_actions": [{"type": "R", "before": "aand", "after": "and", "start_char_pos": 257, "end_char_pos": 261, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "particularily", "after": "particularly", "start_char_pos": 603, "end_char_pos": 616, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "licence", "after": "license", "start_char_pos": 640, "end_char_pos": 647, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "licence", "after": "license", "start_char_pos": 678, "end_char_pos": 685, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "licence", "after": "license", "start_char_pos": 733, "end_char_pos": 740, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "licence", "after": "license", "start_char_pos": 772, "end_char_pos": 779, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "licence", "after": "license", "start_char_pos": 825, "end_char_pos": 832, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "licences", "after": "licenses", "start_char_pos": 994, "end_char_pos": 1002, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 142, 274, 329, 418, 507, 571, 729, 849], "domain": "news"}
{"doc_id": "29056", "revision_depth": 1, "before_revision": "A Missouri woman from blue springs was a victim of assault and she tried to swallow her cell phone the reasons are unknown at this time . The cell phone was caught in the women's wind pipe. When the police arrivedthe boyfriend claimed he was trying to help the woman get the cell phone out of her wind pipe. Police have  charged the boyfriend with assult .", "after_revision": "Blue Springs, Missouri police say a woman they thought had tried to swallow her cell phone may have been the victim of assault . The cell phone was caught in the woman's throat when police arrived. Her boyfriend claimed she had swallowed it to keep it away from him. Melinda Abell required emergency surgery to remove the phone from her throat. Police have now charged the boyfriend , Marlon Brandon Gill, 23, with assault .", "edit_actions": [{"type": "R", "before": "A Missouri woman from blue springs was a victim of assault and she", "after": "Blue Springs, Missouri police say a woman they thought had", "start_char_pos": 0, "end_char_pos": 66, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the reasons are unknown at this time", "after": "may have been the victim of assault", "start_char_pos": 99, "end_char_pos": 135, "major_intent": "clarity", "raw_intents": ["others", "clarity", "clarity"]}, {"type": "D", "before": "women's wind pipe. When the police arrivedthe boyfriend claimed he was trying to help the woman get the cell phone out of her wind pipe.", "after": null, "start_char_pos": 171, "end_char_pos": 307, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "R", "before": "Police have", "after": "woman's throat when police arrived. Her boyfriend claimed she had swallowed it to keep it away from him.", "start_char_pos": 308, "end_char_pos": 319, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Melinda Abell required emergency surgery to remove the phone from her throat. Police have now", "start_char_pos": 320, "end_char_pos": 320, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "with assult", "after": ", Marlon Brandon Gill, 23, with assault", "start_char_pos": 343, "end_char_pos": 354, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 137, 189, 307], "domain": "news"}
{"doc_id": "31455", "revision_depth": 2, "before_revision": "Photographs taken by eye-witnesses show people, who were trapped, dropping from 8th floor windows to their deaths. According to some reports the firemen who were dispatched to the scene were pre-occupied at the rear of the building. They had apparently been ordered to evacuate the bank and its management from the area that was not under immediate threat. Eugenia, a witness said, \"who has told you, that fireteam arrived in 3 minutes? Or in 20 minutes? My uncle worked on anouther side of the building. He personally gave a call to the fire department. They said, \"ok, ok, take it easy, here we go...\" as  it were just some fun for them, \"nothing serious!\" Reports from the scene say the fire-brigade arrived at least 40 minutes or more from the time the call about the fire was placed. The Fire-safety Department officials state that firemen had the \"objective obstacles\" which hampered rescue efforts. Among them were , barred emergency exits , and a congestion of cars in front of the burning building.", "after_revision": "Photographs taken by eye-witnesses show people, who were trapped, dropping from 8th floor windows to their deaths. According to some reports the firemen who were dispatched to the scene were pre-occupied at the rear of the building. They had apparently been ordered to evacuate the bank and its management from the area , which was not under immediate threat. Eugenia, a witness said, \"who has told you, that fireteam arrived in 3 minutes? Or in 20 minutes? My uncle worked on anouther side of the building. He personally gave a call to the fire department. They said, \"ok, ok, take it easy, here we go...\" as if it were just some fun for them, \"nothing serious!\" Reports from the scene say the fire brigade arrived at least 40 minutes or more from the time the call about the fire was placed. The Fire-safety Department officials state that firemen had had \"objective obstacles\" which hampered rescue efforts. Among them were barred emergency exits and a congestion of cars in front of the burning building.", "edit_actions": [{"type": "R", "before": "that", "after": ", which", "start_char_pos": 320, "end_char_pos": 324, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "if", "start_char_pos": 607, "end_char_pos": 607, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "fire-brigade", "after": "fire brigade", "start_char_pos": 690, "end_char_pos": 702, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the", "after": "had", "start_char_pos": 849, "end_char_pos": 852, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 922, "end_char_pos": 923, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 947, "end_char_pos": 948, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 114, 232, 356, 436, 454, 504, 554, 658, 788, 905], "domain": "news"}
{"doc_id": "31689", "revision_depth": 2, "before_revision": "The United States has only seen one case of Mad Cow disease which was in 2005 when a cow tested positive for the disease. Officials later linked the isolated case to Canada because the cow was born on a farm in Alberta. ", "after_revision": "The United States has seen two cases of Mad Cow disease which were in 2005 when a cow tested positive for the disease. Officials later linked one of the cases to Canada because the cow was born on a farm in Alberta. The later test was done in England where samples tested had conflicting results in 2004.", "edit_actions": [{"type": "R", "before": "only seen one case", "after": "seen two cases", "start_char_pos": 22, "end_char_pos": 40, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "was", "after": "were", "start_char_pos": 66, "end_char_pos": 69, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the isolated case", "after": "one of the cases", "start_char_pos": 145, "end_char_pos": 162, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "A", "before": null, "after": "The later test was done in England where samples tested had conflicting results in 2004.", "start_char_pos": 220, "end_char_pos": 220, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 121], "domain": "news"}
{"doc_id": "31689", "revision_depth": 3, "before_revision": "The United States has seen two cases of Mad Cow disease which were in 2005 when a cow tested positive for the disease . Officials later linked one of the cases to Canada because the cow was born on a farm in Alberta. The later test was done in England where samples tested had conflicting results in 2004.", "after_revision": "The United States has seen two cases of Mad Cow disease . The first was discovered in December of 2003 in the state of Washington . Officials later linked this case to Canada because the cow was born on a farm in Alberta. The second infected cow was discovered in Texas in 2005. The later case was diagnosed in England after earlier samples tested had shown conflicting results.", "edit_actions": [{"type": "R", "before": "which were in 2005 when a cow tested positive for the disease", "after": ". The first was discovered in December of 2003 in the state of Washington", "start_char_pos": 56, "end_char_pos": 117, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "one of the cases", "after": "this case", "start_char_pos": 143, "end_char_pos": 159, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "later test was done in England where", "after": "second infected cow was discovered in Texas in 2005. The later case was diagnosed in England after earlier", "start_char_pos": 221, "end_char_pos": 257, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "conflicting results in 2004.", "after": "shown conflicting results.", "start_char_pos": 277, "end_char_pos": 305, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 119, 216], "domain": "news"}
{"doc_id": "31915", "revision_depth": 1, "before_revision": "Pop star Michael Jackson was caught by photographers dressed in womens clothing in a shopping mall in the Gulf state of Bahrain. Jackson was dressed in a black abaya (a traditional fully length robe), a veil covering his face, dark sunglasses and gloves. Jackson was accompanied by three young, western-looking children, their faces covered with black veils, and an adult woman wearning an abaya over jeans, her face partly covered, and bodyguards. After Jackson was recognized, he reportedly said \"Please, No!\" , and the unidentified woman asked photographers to back off. Photographers continued to take pictures while mall security guards escorted Jackson and his entourage out through a back door, where they got into a white Lexus with darkened windows. A picture published by The Australian shows an adult completely covered in black, holding the hand of a young child whose face is covered. Mall officials confirmed that it was Michael Jackson. Jackson has been a guest of the royal family since June 2004, after he was aquitted of sexually molesting young cancer patients who visited his California ranch, Neverland, in 2003.", "after_revision": "Pop star Michael Jackson was caught by photographers dressed in women's clothing in a shopping mall in the Gulf state of Bahrain. Jackson was dressed in a black abaya (a traditional full-length robe), a veil covering his face, dark sunglasses and gloves. Jackson was accompanied by three young, Western-looking children, their faces covered with black veils, and an adult woman wearning an abaya over jeans, her face partly covered, and bodyguards. After Jackson was recognized, he reportedly said \"Please, No!\" and the unidentified woman asked photographers to back off. Photographers continued to take pictures while mall security guards escorted Jackson and his entourage out through a back door, where they got into a white Lexus with darkened windows. A picture published by The Australian shows an adult completely covered in black, holding the hand of a young child whose face is covered. Mall officials confirmed that it was Michael Jackson. Jackson has been a guest of the royal family since June 2004, after he was acquitted of sexually molesting young cancer patients who visited his California ranch, Neverland, in 2003.", "edit_actions": [{"type": "R", "before": "womens", "after": "women's", "start_char_pos": 64, "end_char_pos": 70, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "fully length", "after": "full-length", "start_char_pos": 181, "end_char_pos": 193, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "western-looking", "after": "Western-looking", "start_char_pos": 295, "end_char_pos": 310, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 512, "end_char_pos": 513, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "aquitted", "after": "acquitted", "start_char_pos": 1027, "end_char_pos": 1035, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 128, 254, 448, 573, 758, 897, 951], "domain": "news"}
{"doc_id": "33229", "revision_depth": 1, "before_revision": "Neil Entwistle, 27, appeared before a Magistrates' court today and agreed to be extradited to the United States to face double-murder  charges. Entwistle is now the prime suspect in the murders of his wife Rachel, 27, and nine-month old baby daughter Lillian. Both were shot to death and left under a pile of blankets in the bedroom of their home in Hopkinton, Massachusetts  . Entwistle was reportedly planning a murder suicide, but decided not to kill himself and instead bought a one-way ticket to England where he was with family  .", "after_revision": "Neil Entwistle, 27, appeared before a Magistrates' court today and agreed to extradition to the United States , where he will face double-murder and weapons charges. Entwistle is now the prime suspect in the killings of his wife Rachel, 27, and nine-month old baby daughter Lillian. Both were shot to death and left under a pile of blankets in the bedroom of their Hopkinton, Massachusetts home . Entwistle was reportedly planning a murder-suicide. Instead, police believe he fled to England, where he was subsequently arrested in London. If convicted of first-degree murder, Entwistle could face life in prison .", "edit_actions": [{"type": "R", "before": "be extradited", "after": "extradition", "start_char_pos": 77, "end_char_pos": 90, "major_intent": "others", "raw_intents": ["others", "clarity", "fluency"]}, {"type": "R", "before": "to", "after": ", where he will", "start_char_pos": 112, "end_char_pos": 114, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "and weapons", "start_char_pos": 134, "end_char_pos": 134, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "murders", "after": "killings", "start_char_pos": 186, "end_char_pos": 193, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "home in", "after": null, "start_char_pos": 342, "end_char_pos": 349, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "home", "start_char_pos": 375, "end_char_pos": 375, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "murder suicide, but decided not to kill himself and instead bought a one-way ticket to England", "after": "murder-suicide. Instead, police believe he fled to England,", "start_char_pos": 414, "end_char_pos": 508, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "with family", "after": "subsequently arrested in London.", "start_char_pos": 522, "end_char_pos": 533, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "If convicted of first-degree murder, Entwistle could face life in prison", "start_char_pos": 534, "end_char_pos": 534, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 143, 259, 377], "domain": "news"}
{"doc_id": "33321", "revision_depth": 1, "before_revision": "That petition was submitted for the attention of the King as the climax to Sondhi's last rally in Royal Plaza on February 5. After addressing a crowd estimated to number more than 100,000. Sondhi then hand-delivered the petition to the residence of the President of the Privvy Council. Several accusations are levelled at the Prime Minister in the petition, including allegations that he has used his position for personal financial gain.", "after_revision": "That petition was submitted for the attention of the King as the climax to Sondhi's last rally in Royal Plaza on February 5. After addressing a crowd estimated to number more than 100,000. Sondhi then delivered the petition through an aid at the residence of the President of the Privy Council. Several accusations are leveled at the Prime Minister in the petition, including allegations that he has used his position for personal financial gain.", "edit_actions": [{"type": "R", "before": "hand-delivered the petition to", "after": "delivered the petition through an aid at", "start_char_pos": 201, "end_char_pos": 231, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "Privvy", "after": "Privy", "start_char_pos": 270, "end_char_pos": 276, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "levelled", "after": "leveled", "start_char_pos": 310, "end_char_pos": 318, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 188, 285], "domain": "news"}
{"doc_id": "33559", "revision_depth": 1, "before_revision": "A group of former senior Communist Party officials in China have denounced the recent closure of BingDian (Freezing Point ), an investigative newspaper, along with the state of government censorship in that country. The group includes Li Rui, a former aid to Chairman Mao Zedong, Hu Jiwei, a former newspaper editor, and Zhu Houze, an ex-propaganda chief. In the letter, signed on February 2nd but released on Tuesday, the group called current censorship methods overly restrictive and unnecessary for the modern China: \"History has shown that only a totalitarian system needs news censorship, out of the delusion that it can keep the people locked in ignorance.\" Freezing Point was a supplement of the China Youth Daily and was allegedly shut down due to the publication on 11 January of an article by Zhongshan University professor Yuan Weishi regarding the way history is currently taught in China, although the papers editor, Li Datong, says that it has experienced difficulties with the government on several occasions in the past. Mr. Yuan's article was critical of Chinese text books, which never admit the culpability of the government and instead shift the blame to other nations.", "after_revision": "A group of former senior Communist Party officials in China have denounced the recent closure of Freezing Point (Bing Dian ), an investigative newspaper, along with the state of government censorship in that country. The group includes Li Rui, a former aid to Chairman Mao Zedong, Hu Jiwei, a former newspaper editor, and Zhu Houze, an ex-propaganda chief. In the letter, signed on February 2 but released on Tuesday, the group called current censorship methods overly restrictive and unnecessary for the modern China: \"History has shown that only a totalitarian system needs news censorship, out of the delusion that it can keep the people locked in ignorance.\" Freezing Point was a supplement of the China Youth Daily and was allegedly shut down due to the publication on January 11 of an article by Zhongshan University professor Yuan Weishi regarding the way history is currently taught in China, although the papers editor, Li Datong, says that it has experienced difficulties with the government on several occasions in the past. Mr. Yuan's article was critical of Chinese text books, which never admit the culpability of the government and instead shift the blame to other nations.", "edit_actions": [{"type": "R", "before": "BingDian (Freezing Point", "after": "Freezing Point (Bing Dian", "start_char_pos": 97, "end_char_pos": 121, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "2nd", "after": "2", "start_char_pos": 390, "end_char_pos": 393, "major_intent": "style", "raw_intents": ["fluency", "style", "style"]}, {"type": "R", "before": "11 January", "after": "January 11", "start_char_pos": 775, "end_char_pos": 785, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}], "sents_char_pos": [0, 215, 355, 663, 1036], "domain": "news"}
{"doc_id": "33602", "revision_depth": 1, "before_revision": "Two graduate students at The Death Star claim to have shown URLasm can happen much faster than previously thought. Physics student Darth Vader and his former roommate, Master Yoda , decided to hire a nude model to engage in intercourse with a well-known marine  in order to predict how URLasms can occur in the body. The gay technique, is used to predict the how large your penis will be based on its historical behavior. Initially their nude model confirmed his analysis that asians have small pee pees was found to factual . However, fossil-preserving rocks from some periods are more commonly found than from others. When they compensated for this, the lag disappeared. So far as could be seen from the geological record, the biosphere seemed to respond immediately to changes. This would mean that life could rapidly recover from large extinctions; it was previously thought such recovery would take millions of years.", "after_revision": "Two graduate students at Harvard claim to have shown that evolution can happen much faster than previously thought. Physics student Peter Lu and his former roommate, economist Motohiro Yogo , decided to apply an economics model to a well-known marine fossil record in order to predict how fast life can adapt to changes in the environment. The technique, vector autoregression, is used to predict the stock market based on its historical behavior. Initially their model confirmed previous analysis that life was slow to adapt . However, fossil-preserving rocks from some periods are more commonly found than from others. When they compensated for this, the lag disappeared. So far as could be seen from the geological record, the biosphere seemed to respond immediately to changes. This would mean that life could rapidly recover from large extinctions; it was previously thought such recovery would take millions of years.", "edit_actions": [{"type": "R", "before": "The Death Star", "after": "Harvard", "start_char_pos": 25, "end_char_pos": 39, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "URLasm", "after": "that evolution", "start_char_pos": 60, "end_char_pos": 66, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Darth Vader", "after": "Peter Lu", "start_char_pos": 131, "end_char_pos": 142, "major_intent": "meaning-changed", "raw_intents": ["others", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Master Yoda", "after": "economist Motohiro Yogo", "start_char_pos": 168, "end_char_pos": 179, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "hire a nude model to engage in intercourse with", "after": "apply an economics model to", "start_char_pos": 193, "end_char_pos": 240, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "others"]}, {"type": "A", "before": null, "after": "fossil record", "start_char_pos": 261, "end_char_pos": 261, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "URLasms can occur in the body. The gay technique,", "after": "fast life can adapt to changes in the environment. The technique, vector autoregression,", "start_char_pos": 286, "end_char_pos": 335, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "how large your penis will be", "after": "stock market", "start_char_pos": 359, "end_char_pos": 387, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "nude model confirmed his analysis that asians have small pee pees was found to factual", "after": "model confirmed previous analysis that life was slow to adapt", "start_char_pos": 438, "end_char_pos": 524, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 114, 316, 421, 526, 619, 672, 780, 852], "domain": "news"}
{"doc_id": "33663", "revision_depth": 1, "before_revision": "On February 13, Baltimore Ravens cornerback Deion Sanders announced his retirement from the NFL. Last January, many assumed Sanders would be leaving the league after spending 14 years in it. Sanders won back-to-back Super Bowls with the San Fransisco 49ers and Dallas Cowboys. Sanders is also the first professional athlete to win a World Series and Super Bowl in his pro career. The All-Pro Cornerback had 53 interceptions, was selected to the Pro Bowl eight times and was once Defensive Player of the Year. Reports show that Sanders may go back into news broadcasting, most likely on the FOX NFL Pre-Game Show. After his retirement, current Ravens Cornerback Chris McAlister stated, \"I'll miss Deion, just as the rest of the Ravens organization will. I learned a lot from him.\"", "after_revision": "On February 13, Baltimore Ravens cornerback Deion Sanders announced his re-retirement from the NFL. Last January, many assumed Sanders would be leaving the league after spending 14 years in it. Sanders won back-to-back Super Bowls with the San Fransisco 49ers and Dallas Cowboys. Sanders is the first professional athlete to win a World Series and Super Bowl in his pro career. The All-Pro Cornerback had 53 interceptions, was selected to the Pro Bowl eight times and was once Defensive Player of the Year. Reports speculate Sanders, who broadcast with CBS after his first retirement, may go back into news broadcasting, possibly on the FOX NFL Pre-Game Show. After his retirement, current Ravens Cornerback Chris McAlister stated, \"I'll miss Deion, just as the rest of the Ravens organization will. I learned a lot from him.\"", "edit_actions": [{"type": "R", "before": "retirement", "after": "re-retirement", "start_char_pos": 72, "end_char_pos": 82, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "also", "after": null, "start_char_pos": 288, "end_char_pos": 292, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "Reports show that Sanders", "after": "Reports speculate Sanders, who broadcast with CBS after his first retirement,", "start_char_pos": 509, "end_char_pos": 534, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "most likely", "after": "possibly", "start_char_pos": 571, "end_char_pos": 582, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 190, 276, 379, 508, 612, 752]}
{"doc_id": "33665", "revision_depth": 1, "before_revision": "Pyleva, 30, tested positive for Carphedon , a prohibited substance, on 13 February after the women's 15 km Individual, Biathlon event, in which she won the silver medal. ITAR-TASS News Agency reports that Pyleva tested positive for Cordedon , a stimulant banned by the World Anti Doping Agency (WADA). The substitute, phenotropil, was prescribed by her doctors following an ankle injury according to Nikolai Durmanov, head of the Russian Anti Doping Committee.", "after_revision": "Pyleva, 30, tested positive for carphedon , a prohibited substance, on February 13 after the women's 15 km individual, biathlon event, in which she won the silver medal. ITAR-TASS News Agency reports that Pyleva tested positive for cordedon , a stimulant banned by the World Anti-Doping Agency (WADA). The substitute, phenotropil, was prescribed by her doctors following an ankle injury according to Nikolai Durmanov, head of the Russian Anti Doping Committee.", "edit_actions": [{"type": "R", "before": "Carphedon", "after": "carphedon", "start_char_pos": 32, "end_char_pos": 41, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "13 February", "after": "February 13", "start_char_pos": 71, "end_char_pos": 82, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "Individual, Biathlon", "after": "individual, biathlon", "start_char_pos": 107, "end_char_pos": 127, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Cordedon", "after": "cordedon", "start_char_pos": 232, "end_char_pos": 240, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Anti Doping", "after": "Anti-Doping", "start_char_pos": 275, "end_char_pos": 286, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 169, 301], "domain": "news"}
{"doc_id": "33668", "revision_depth": 1, "before_revision": "Flag of the United Nations In a 40 page report released today, the United Nations special envoys on torture reported their findings regarding the United States detention facility at Guantanamo Bay on the island of Cuba. The report concludes that certain procedures practised at the facility, and during transport to it, constitute torture. Moreover the report states that \" Attempts by the United States Administration to redefine \"torture\" in the framework of the struggle against terrorism  in order to allow certain interrogation techniques that would not be permitted under the internationally accepted definition of torture  are of utmost concern.\" and that \" The interrogation techniques authorized by the Department of Defense, particularly if used simultaneously, amount to degrading treatment in violation of article 7 of ICCPR and article 16 of the Convention against Torture.\"", "after_revision": "Flag of the United Nations In a 40 page report released today, the United Nations special envoys on torture reported their findings regarding the United States detention facility at Guantanamo Bay on the island of Cuba. The report concludes that certain procedures practised at the facility, and during transport to it, constitute torture. Moreover the report states that \" attempts by the United States Administration to redefine 'torture' in the framework of the struggle against terrorism , in order to allow certain interrogation techniques that would not be permitted under the internationally accepted definition of torture , are of utmost concern.\" The report goes on to state that \" the interrogation techniques authorized by the Department of Defense, particularly if used simultaneously, amount to degrading treatment in violation of article 7 of ICCPR and article 16 of the Convention against Torture.\"", "edit_actions": [{"type": "R", "before": "Attempts", "after": "attempts", "start_char_pos": 374, "end_char_pos": 382, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "\"torture\"", "after": "'torture'", "start_char_pos": 431, "end_char_pos": 440, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 492, "end_char_pos": 492, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 629, "end_char_pos": 629, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and", "after": "The report goes on to state", "start_char_pos": 654, "end_char_pos": 657, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "R", "before": "The", "after": "the", "start_char_pos": 665, "end_char_pos": 668, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 219, 339, 653], "domain": "news"}
{"doc_id": "34369", "revision_depth": 1, "before_revision": "400px|Around 600 people gather in Catte Street, Oxford, on the PRO-test march. OXFORD, ENGLAND - Today saw the first demonstration in support of Oxford University's new animal testing facility on South Parks Road. Around 600 students marched under the banner of PRO-test, URLanisation set up several weeks ago in opposition to SPEAK, which campaigns and demonstrates frequently in the city for animal rights. The demonstration began at 11:30am outside Balliol College on Broad Street, in the historic centre of the city, and (just before midday) marched along Holywell Street and Mansfield Road up towards South Parks Road. Police prevented the PRO-test demonstration from reaching the actual building site because of a SPEAK protest (around 150 people, according to Thames Valley Police) which was occurring there at that time. The crowd was addressed by Neurosurgeon Tipu Aziz, Dr Simon Festing ( Executive director of the Research Defense Society), and Liberal Democrat MP Dr Evan Harris, amongst others. Last year, the ALF claimed responsibility for its arson attack on the Long Bridges boathouse, and recently it put out a communiqu stating that they \"must target professors, teachers, heads, students, investors, partners, supporters and ANYONE that dares to deal in any part of the University in any way.\" SPEAK and the ALF are URLanisations; the former campaigns peacefully, while the latter often uses tactics such as threatening letters, intimidation and criminal damage to further its cause. It should be noted, however, that in an interview for SKY News on the 25th February, a spokeswoman for SPEAK refused to condemn their activities. External Resources PRO-test SPEAK", "after_revision": "400px|Around 600 people gather in Catte Street, Oxford, on the PRO-test march. Oxford, England - Today saw the first demonstration in support of Oxford University's new animal testing facility on South Parks Road. Around 600 students marched under the banner of PRO-test, URLanisation set up several weeks ago in opposition to SPEAK, which campaigns and demonstrates frequently in the city for animal rights. The demonstration began at 11:30am outside Balliol College on Broad Street, in the historic centre of the city, and (just before midday) marched along Holywell Street and Mansfield Road up towards South Parks Road. Police prevented the PRO-test demonstration from reaching the actual building site because of a SPEAK protest (around 150 people, according to Thames Valley Police) which was occurring there at that time. The crowd was addressed by neurosurgeon Tipu Aziz, Dr Simon Festing ( executive director of the Research Defense Society), and Liberal Democrat MP Dr Evan Harris, amongst others. Last year, the ALF claimed responsibility for its arson attack on the Long Bridges boathouse, and recently it put out a communiqu stating that they \"must target professors, teachers, heads, students, investors, partners, supporters and ANYONE that dares to deal in any part of the University in any way.\" SPEAK and the ALF are URLanisations; the former campaigns peacefully, while the latter often uses tactics such as threatening letters, intimidation and criminal damage to further its cause. It should be noted, however, that in an interview for Sky News on the 25th February, a spokeswoman for SPEAK refused to condemn their activities. External links PRO-test SPEAK", "edit_actions": [{"type": "R", "before": "OXFORD, ENGLAND", "after": "Oxford, England", "start_char_pos": 79, "end_char_pos": 94, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": "Neurosurgeon", "after": "neurosurgeon", "start_char_pos": 856, "end_char_pos": 868, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Executive", "after": "executive", "start_char_pos": 899, "end_char_pos": 908, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "SKY", "after": "Sky", "start_char_pos": 1557, "end_char_pos": 1560, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Resources", "after": "links", "start_char_pos": 1658, "end_char_pos": 1667, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}], "sents_char_pos": [0, 78, 213, 408, 623, 828, 1007, 1312, 1349, 1502, 1648]}
{"doc_id": "34659", "revision_depth": 1, "before_revision": "While the PNG Government do have laws and regulations to ensure sustainable timber production, these were not being enforced, the report states. It identified \"a political vacuum with no demonstrated government interest in controlling the problems in the sector\". The report summarised independent reviews of the timber industry between 2000 and 2005. Forest Trends ' said corruption had devastated rural living standards and ignored the basic rights of landowners: \"There are a few logging operations in the country which are deemed beneficial to both local landowners and the country, but they are lost in a sea of bad operators. The Government needs to support these companies, or risk having the international community boycott all of PNG's exports.\" If felling continues the way it is done now , they warn, Papua New Guinea could be bereft of its natural cover in a decade.", "after_revision": "While the PNG Government does have laws and regulations to ensure sustainable timber production, these were not being enforced, the report states. It identified \"a political vacuum with no demonstrated government interest in controlling the problems in the sector\". The report summarised independent reviews of the timber industry between 2000 and 2005. Forest Trends claimed corruption had devastated rural living standards and ignored the basic rights of landowners: \"There are a few logging operations in the country which are deemed beneficial to both local landowners and the country, but they are lost in a sea of bad operators. The Government needs to support these companies, or risk having the international community boycott all of PNG's exports.\" If foresting continues in this manner , they warn, Papua New Guinea could be bereft of its natural cover in a decade.", "edit_actions": [{"type": "R", "before": "do", "after": "does", "start_char_pos": 25, "end_char_pos": 27, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "' said", "after": "claimed", "start_char_pos": 366, "end_char_pos": 372, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "felling continues the way it is done now", "after": "foresting continues in this manner", "start_char_pos": 758, "end_char_pos": 798, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}], "sents_char_pos": [0, 144, 263, 351, 631, 754], "domain": "news"}
{"doc_id": "34841", "revision_depth": 1, "before_revision": "\"Truth is - it may have very little to do with Bush's visit,\" writes Indian blogger Neha Viswanathan on the Global Voices email list, run by the Harvard Berkman Center for Law & Society  . \"The anti-Bush protest has hardly found a voice online. I really do think this might be the first of other blocks. This might be on a very experimental basis to see how far they go. The Pakistani govt has hardly had any dialogue with bloggers at all. \" Other bloggers see a more direct connection with the Bush visit. \"I have communicated with several bloggers and friends back in Pak[istan],\" writes UAE-based Pakistani blogger Omer Alvie. \"So far there's no resolution to this problem. Bush's visit in Pakistan is resulting in protests, strike (in Lahore city) and curfew areas in Islamabad. The feeling among bloggers is this is all rather suspect. Neha, might be right this might be on experimental basis. I feel this a a precursor to what the government can do in the future.\"", "after_revision": "\"Truth is it may have very little to do with Bush's visit,\" writes Indian blogger Neha Viswanathan on the Global Voices email list, run by the Berkman Center for Internet & Society at Harvard Law School . \"The anti-Bush protest has hardly found a voice online. I really do think this might be the first of other blocks. This might be on a very experimental basis to see how far they go. The Pakistani govt has hardly had any dialogue with bloggers at all. \" Other bloggers see a more direct connection with the Bush visit. \"I have communicated with several bloggers and friends back in Pak[istan],\" writes UAE-based Pakistani blogger Omer Alvie. \"So far there's no resolution to this problem. Bush's visit in Pakistan is resulting in protests, a strike in Lahore and curfew areas in Islamabad. The feeling among bloggers is this is all rather suspect. Neha, might be right this might be on experimental basis. I feel this a a precursor to what the government can do in the future.\"", "edit_actions": [{"type": "D", "before": "-", "after": null, "start_char_pos": 10, "end_char_pos": 11, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "D", "before": "Harvard", "after": null, "start_char_pos": 145, "end_char_pos": 152, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "Law", "after": "Internet", "start_char_pos": 172, "end_char_pos": 175, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "at Harvard Law School", "start_char_pos": 186, "end_char_pos": 186, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "strike (in Lahore city)", "after": "a strike in Lahore", "start_char_pos": 728, "end_char_pos": 751, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 244, 303, 370, 439, 506, 629, 676, 782, 840, 898], "domain": "news"}
{"doc_id": "36231", "revision_depth": 1, "before_revision": "The New Zealand Department of Conservation (DOC) made an emergency evacuation of staff from Raoul Island, the largest and northernmost of the Kermadec Islands, due to a volcanic eruption that occurred this morning at 8:21am (New Zealand Daylight Time)  A helicopter picked up the five available staff, but the missing staff member could not be found before it had to leave. The return journey is 2000km  The missing staff member failed to return \"after going on a routine mission to check the water temperatue of the lake,\" according to Rolien Elliot, DOC Area Manager for Warkworth. She said that when staff went to look for their colleague , where he was thought to be  they had to retreat due to further volcanic activity and the track being impassable with fallen trees and ash. \"There's been about five hectares of native bush cleared and a lot of mud and evidence of boulders and rocks that have been thrown out of the crater itself,\" the helicopter pilot John Funnell said to One News  Related Wikinews", "after_revision": "The New Zealand Department of Conservation (DOC) made an emergency evacuation of staff from Raoul Island, the largest and northernmost of the Kermadec Islands, due to a volcanic eruption that occurred this morning at 8:21am (New Zealand Daylight Time) . A helicopter picked up the five available staff, but one missing staff member could not be found before it had to leave. The return journey is 2000km (1300 miles). The missing man, aged in his 30s, failed to return \"after going on a routine mission to check the water temperature of the lake,\" according to Rolien Elliot, DOC Area Manager for Warkworth. She said that when two of the others went to look for their colleague where he was thought to be , they had to retreat due to further volcanic activity and the track being impassable with fallen trees and ash. \"There's been about five hectares of native bush cleared and a lot of mud and evidence of boulders and rocks that have been thrown out of the crater itself,\" the helicopter pilot John Funnell said to One News . Related Wikinews", "edit_actions": [{"type": "A", "before": null, "after": ".", "start_char_pos": 252, "end_char_pos": 252, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the", "after": "one", "start_char_pos": 306, "end_char_pos": 309, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "(1300 miles).", "start_char_pos": 403, "end_char_pos": 403, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "staff member", "after": "man, aged in his 30s,", "start_char_pos": 416, "end_char_pos": 428, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "temperatue", "after": "temperature", "start_char_pos": 499, "end_char_pos": 509, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "staff", "after": "two of the others", "start_char_pos": 603, "end_char_pos": 608, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 642, "end_char_pos": 643, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 671, "end_char_pos": 671, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 992, "end_char_pos": 992, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 373, 583, 782], "domain": "news"}
{"doc_id": "36388", "revision_depth": 1, "before_revision": "Australian actor Russell Crowe and businessman Peter Holmes a Court have succeeded in their takeover bid of the  Australian National Rugby League (NRL) Club , the South Sydney Rabbitohs . The men required  75\\% of the clubs members to vote yes on the $3 million takeover bid for it to pass , narrowly winning by just 32 votes (0.8\\%). Former Souths player and manager, URLe Piggins, has vowed never to attend another game of his former club. Supporters of the bid have said it gives the Rabbitohs a chance to return to their glory days after last year's finish of 13th in the 15 team competition. The team has not won a premiership in 35 years but still retains the record for most Premiership wins  with 20 on the record books. Those against the bid have claimed the club is selling out 98 years of heritage.", "after_revision": "Australian actor Russell Crowe and businessman Peter Holmes a Court have succeeded in their takeover bid of the South Sydney Rabbitohs, an Australian National Rugby League (NRL) Club . The men required that 75\\% of the club's members vote yes on the $3 million takeover bid for it to pass ; and it narrowly won, by just 32 votes (0.8\\%). Former Souths player and manager, URLe Piggins, has vowed never to attend another game of his former club. Supporters of the bid have said it gives the Rabbitohs a chance to return to their glory days after last year's finish of 13th in the 15-team competition. The team has not won a premiership in 35 years but still retains the record for most Premiership wins , with 20 on the record books. Those against the bid objected that the club is selling out 98 years of heritage.", "edit_actions": [{"type": "A", "before": null, "after": "South Sydney Rabbitohs, an", "start_char_pos": 112, "end_char_pos": 112, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ", the South Sydney Rabbitohs", "after": null, "start_char_pos": 157, "end_char_pos": 185, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "that", "start_char_pos": 205, "end_char_pos": 205, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "clubs members to", "after": "club's members", "start_char_pos": 218, "end_char_pos": 234, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", narrowly winning", "after": "; and it narrowly won,", "start_char_pos": 290, "end_char_pos": 308, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "15 team", "after": "15-team", "start_char_pos": 576, "end_char_pos": 583, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 699, "end_char_pos": 699, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "have claimed", "after": "objected that", "start_char_pos": 751, "end_char_pos": 763, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 187, 334, 441, 596, 728], "domain": "news"}
{"doc_id": "36619", "revision_depth": 1, "before_revision": "At approximately 12:43 a.m. on Wednesday, March 22, the British Columbia ferry MV Queen of the North hit a rock near Gil Island in Wright Sound, off the coast of Port Hardy in British Columbiaafter veering off course . It took about an hour for the ferry to sink, giving passengers time to get out. The vessel was carrying 101 passengers and crew, and all but 2 were accounted for. Thanks to quick-responding fishing vessels and a helicopter, they were taken to nearby Hartley Bay . The Queen of the North, which is not normally used on this route, is larger than the vessel normally used, the Queen of Prince Rupert  .", "after_revision": "At approximately 00:26 hrs, on Wednesday, March 22, the British Columbia ferry MV Queen of the North hit a rock near Gil Island in Wright Sound, off the coast of British Columbia, Canada, 75 from the start of ship's schedule journey from Prince Rupert to Port Hardy on the Northern tip of Vancouver IslandPort Hardy. The ship missed a course change and ran agroung off Gil Island . It took about an hour for the ferry to sink, giving passengers time to get out. The vessel was carrying 101 passengers and crew, and all but 2 were accounted for. Thanks to quick-response of nearby aboriginal fishing villager of Hartley Bay, and the CCGS Sir Wiflrid Laurier. Survivors were quickly fetched from the Queen of the North's lifeboat and brought to Hartley Bay and onboard the coast Guard ship, and eventually back to Prince Rupert onboard the Sir Wilfrid Laurier. Hartley Bay. The Queen of the North, is (apx) 9,000 gross ton, motor driven vessel, built in Germany in 1969. The Queen of the North had been on this 15 hrs, 274 nautical mile ferry run since late May 1980. In the summer and on relief work, her smaller sister ship, the Queen of Prince Rupert also performs that run .", "edit_actions": [{"type": "R", "before": "12:43 a.m.", "after": "00:26 hrs,", "start_char_pos": 17, "end_char_pos": 27, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Port Hardy in British Columbiaafter veering off course", "after": "British Columbia, Canada, 75 from the start of ship's schedule journey from Prince Rupert to Port Hardy on the Northern tip of Vancouver IslandPort Hardy. The ship missed a course change and ran agroung off Gil Island", "start_char_pos": 162, "end_char_pos": 216, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "quick-responding fishing vessels and a helicopter, they were taken to nearby Hartley Bay .", "after": "quick-response of nearby aboriginal fishing villager of Hartley Bay, and the CCGS Sir Wiflrid Laurier. Survivors were quickly fetched from the Queen of the North's lifeboat and brought to Hartley Bay and onboard the coast Guard ship, and eventually back to Prince Rupert onboard the Sir Wilfrid Laurier. Hartley Bay.", "start_char_pos": 392, "end_char_pos": 482, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "which is not normally used on this route, is larger than the vessel normally used,", "after": "is (apx) 9,000 gross ton, motor driven vessel, built in Germany in 1969. The Queen of the North had been on this 15 hrs, 274 nautical mile ferry run since late May 1980. In the summer and on relief work, her smaller sister ship,", "start_char_pos": 507, "end_char_pos": 589, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "also performs that run", "start_char_pos": 617, "end_char_pos": 617, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 218, 298, 381, 482], "domain": "news"}
{"doc_id": "36888", "revision_depth": 1, "before_revision": "The British National Party(BNP)  a far-right political party , failed to hold onto one of its seats on Bradford City Council, West Yorkshire, UK. The byelection was called after the incumbent BNP councillor, Angela Clark, stood down. The seat was won in 2004 following a BNP hijacking of a campaign against gangs grooming girls for sex. The BNP claimed white girls were being groomed by Asian men for sex. The original campaign URLanised by Angela Sinfield, whose 13 year-old daughter was groomed by a gang which she pointed out contained white and Asian men. Sinfield joined the Labour Party last year and was its candidate for the Keighley West seat. She sensationally won the seat with 1819 votes over the BNP's 1216, a swing of 11.4 per cent from the BNP to Labour. The voter turnout was near general election levels, 58.8 per cent.Three BNP councillors remain on Bradford City Council, with two of them up for election in May.", "after_revision": "The British National Party(BNP) , a far-right political party failed to hold on to one of its seats on Bradford City Council, West Yorkshire, UK. The byelection was called after the incumbent BNP councillor, Angela Clark, stood down. BNP had won the seat in 2004 after it adapted an ongoing campaign against gangs grooming girls for sex. The BNP claimed white girls were being groomed by Asian men for sex. The original campaign URLanised by Angela Sinfield, whose 13 year-old daughter was groomed by a gang which she pointed out contained white and Asian men. Sinfield joined the Labour Party last year and was its candidate for the Keighley West seat. She sensationally won the seat with 1819 votes over the BNP's 1216, a swing of 11.4 per cent from the BNP to Labour. The voter turnout was near general election levels, 58.8 per cent.Three BNP councillors remain on Bradford City Council, with two of them up for election in May.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 32, "end_char_pos": 32, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 61, "end_char_pos": 62, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "onto", "after": "on to", "start_char_pos": 78, "end_char_pos": 82, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "The seat was won", "after": "BNP had won the seat", "start_char_pos": 234, "end_char_pos": 250, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "following a BNP hijacking of a", "after": "after it adapted an ongoing", "start_char_pos": 259, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 145, 233, 336, 405, 559, 652, 769, 836], "domain": "news"}
{"doc_id": "37606", "revision_depth": 1, "before_revision": "Former Republican committeewoman Jane Faust was a plaintiff yesterday in Pennsylvania's Commonwealth Court (state court). Her lawyer, Larry Otter, argued that Patrick Murphy should be thrown off the ballot for the Democratic primary election in the eighth congressional district of Pennsylvania (PA). The eighth district is primarily located in Bucks County, PA. Otter argued that during the period of election petition signature collection, Murphy had signed his name as the circulator on two petition sheets, but was not present for each signature that was signed on those sheets, and thereby breaking PA state law. Current law states that whoever signs a petition sheet as a circulator must be present for each signature , and see each signature placed on the sheet. Murphy admitted that he was not at the sheet for every signature, and  agreed that the 96 signatures present on those two sheets should be invalidated. He also stated that he was unaware of that particular election statute. Otter argued that the candidacy should be thrown out because when Murphy presented his signatures he was required to sign an affidavit stating he would \"not knowingly violate any election law.\" It was Otter's contention to Judge Robert Simpson Jr. that it is unbelievable that Murphy would not know the law, as Murphy is a lawyer. \"Mr. Murphy is a member of the bar of this court. He can't say he doesn't know what the law was. He's been campaigning over a year,\" Otter said. He then continued to argue that since Murphy knowingly falsified his name on those two circulation petitions, he also failed to fulfill his candidate affidavit, which according to state law is a criminal offense.", "after_revision": "Jane Faust, former Republican committeewoman once convicted of forging signatures, was a plaintiff yesterday in Pennsylvania's Commonwealth Court (state court). Her lawyer, Larry Otter, argued that Patrick Murphy should be thrown off the ballot for the Democratic primary election in the eighth congressional district of Pennsylvania (PA). The eighth district is primarily located in Bucks County, PA. Otter argued that during the period of election petition signature collection, Murphy had signed his name as the circulator on two petition sheets, but was not present for each signature that was signed on those sheets, and thereby breaking PA state law. PA law states that whoever signs a petition sheet as a circulator must be present for each signature and see each signature placed on the sheet. Murphy admitted that he was not at the sheet for every signature, and he agreed that the 96 signatures on those two sheets should be invalidated. He also stated that he was unaware of that particular election statute. Otter argued that the candidacy should be thrown out because when Murphy presented his signatures he was required to sign an affidavit stating he would \"not knowingly violate any election law.\" It was Otter's contention to Judge Robert Simpson Jr. that it is unbelievable that Murphy would not know the law, as Murphy is a lawyer. \"Mr. Murphy is a member of the bar of this court. He can't say he doesn't know what the law was. He's been campaigning over a year,\" Otter said. He then continued , arguing that since Murphy knowingly falsified his name on those two circulation petitions, he also failed to fulfill his candidate affidavit, which according to state law is a criminal offense.", "edit_actions": [{"type": "R", "before": "Former Republican committeewoman Jane Faust", "after": "Jane Faust, former Republican committeewoman once convicted of forging signatures,", "start_char_pos": 0, "end_char_pos": 43, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Current", "after": "PA", "start_char_pos": 618, "end_char_pos": 625, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 724, "end_char_pos": 725, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "A", "before": null, "after": "he", "start_char_pos": 840, "end_char_pos": 840, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "present", "after": null, "start_char_pos": 871, "end_char_pos": 878, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "to argue", "after": ", arguing", "start_char_pos": 1488, "end_char_pos": 1496, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}], "sents_char_pos": [0, 121, 300, 362, 617, 769, 921, 993, 1187, 1324, 1374, 1421, 1469], "domain": "news"}
{"doc_id": "37606", "revision_depth": 2, "before_revision": "Jane Faust, former Republican committeewoman once convicted of forging signatures, was a plaintiff yesterday in Pennsylvania's Commonwealth Court (state court). Her lawyer, Larry Otter, argued that Patrick Murphy should be thrown off the ballot for the Democratic primary election in the eighth congressional district of Pennsylvania (PA) . The eighth district is primarily located in Bucks County, PA . Otter argued that during the period of election petition signature collection, Murphy had signed his name as the circulator on two petition sheets, but was not present for each signature that was signed on those sheets, and thereby breaking PA state law. PA law states that whoever signs a petition sheet as a circulator must be present for each signature and see each signature placed on the sheet. Murphy admitted that he was not at the sheet for every signature, and he agreed that the 96 signatures on those two sheets should be invalidated. He also stated that he was unaware of that particular election statute. Otter argued that the candidacy should be thrown out because when Murphy presented his signatures he was required to sign an affidavit stating he would \"not knowingly violate any election law.\" It was Otter's contention to Judge Robert Simpson Jr. that it is unbelievable that Murphy would not know the law, as Murphy is a lawyer.", "after_revision": "Jane Faust, former Republican committeewoman once convicted of forging signatures, was a plaintiff yesterday in Pennsylvania's Commonwealth Court (state court). Her lawyer, Larry Otter, argued that Patrick Murphy should be thrown off the ballot for the Democratic primary election in the eighth congressional district of Pennsylvania . The eighth district is primarily located in Bucks County, Pennsylvania . Otter argued that during the period of election petition signature collection, Murphy had signed his name as the circulator on two petition sheets, but was not present for each signature that was signed on those sheets, and thereby breaking Pennsylvania state law. Pennsylvania law states that whoever signs a petition sheet as a circulator must be present for each signature and see each signature placed on the sheet. Murphy admitted that he was not at the sheet for every signature, and he agreed that the 96 signatures on those two sheets should be invalidated. He also stated that he was unaware of that particular election statute. Otter argued that the candidacy should be thrown out because when Murphy presented his signatures he was required to sign an affidavit stating he would \"not knowingly violate any election law.\" It was Otter's contention to Judge Robert Simpson , Jr., that it is unbelievable that Murphy would not know the law, as Murphy is a lawyer.", "edit_actions": [{"type": "D", "before": "(PA)", "after": null, "start_char_pos": 334, "end_char_pos": 338, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "PA", "after": "Pennsylvania", "start_char_pos": 399, "end_char_pos": 401, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "PA", "after": "Pennsylvania", "start_char_pos": 645, "end_char_pos": 647, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "PA", "after": "Pennsylvania", "start_char_pos": 659, "end_char_pos": 661, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "Jr.", "after": ", Jr.,", "start_char_pos": 1266, "end_char_pos": 1269, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 160, 658, 803, 949, 1021, 1215], "domain": "news"}
{"doc_id": "37761", "revision_depth": 1, "before_revision": "According to many published reports and TV sources, Herb Sendek, under fire at North Carolina State Unversity, has accepted the job for head coach at Arizona State University, for reportably at least a 7 figure contact. Arizona Republic reported that Sendek will be named head coach on Monday, and school officials did not deny that he would be named head coach. Sendek went 170-118 (.590) during his 10 season stay at NC State, and  went 22-10 overall and 10-6 in the ACC this past year . Sendek is replacing Rob Evans, who was fired on March 10. Evans went 119-120 in eight years at ASU.", "after_revision": "According to many published reports and TV sources, Herb Sendek, under fire at North Carolina State Unversity, has accepted the job of head coach at Arizona State University, for reportably at least a seven-figure contract. The Arizona Republic reported that Sendek will be named head coach on Monday, and school officials did not deny that he would be named head coach. Sendek went 170-118 (.590) during his 10-season stay at NC State, and this past year went 22-10 overall and 10-6 in the ACC . Sendek is replacing Rob Evans, who was fired on March 10. Evans went 119-120 in eight years at ASU.", "edit_actions": [{"type": "R", "before": "for", "after": "of", "start_char_pos": 132, "end_char_pos": 135, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "7 figure contact.", "after": "seven-figure contract. The", "start_char_pos": 202, "end_char_pos": 219, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "10 season", "after": "10-season", "start_char_pos": 401, "end_char_pos": 410, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "this past year", "start_char_pos": 433, "end_char_pos": 433, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "this past year", "after": null, "start_char_pos": 473, "end_char_pos": 487, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 219, 362, 547], "domain": "news"}
{"doc_id": "38599", "revision_depth": 1, "before_revision": "Map highlighting Tennessee. A six-year-old girl was killed after a Black Bear attacked her. Her mother and two-year-old brother, also attacked, were critically injured. The nearly 400-pound bear attacked the family when they were staying at Chilwoee Campground located within the Cherokee National Forest in Tennessee. The family was visitng a nearby waterfall when they were attacked. The mother and the brother both remain in critical condition at a hospital in Chattanooga hospital. The boys head was bitten by the bear, which punctured his skull, before it attacked his mother. The mother was trying to scare the bear away with rocks and sticks before being attacked. According to doctors, she had several puncture wounds to her neck, but had too many other claw and teeth marks for them to count. Both are expected to recover. Investigators , so far, have not been able to speak to the mother because her injuries are too severe . \"She may not remember the attack at all,\" said Hicks . He also said that it was only Tennessee's second documented attack involving a bear. Authorities have not yet released the names of the victims and the campground has been closed.", "after_revision": "Map highlighting Tennessee. A six-year-old girl was killed after a black bear attacked her. Her mother and two-year-old brother, also attacked, were critically injured. The nearly 400-pound bear attacked the family when they were staying at Chilwoee Campground located within the Cherokee National Forest in Tennessee. The family was visiting a nearby waterfall when they were attacked. The mother and the brother both remain in critical condition at a hospital in Chattanooga . The boy's skull was punctured by a bite to the head, before the bear attacked his mother. The mother was trying to scare the bear away with rocks and sticks before being attacked. According to doctors, she had several puncture wounds to her neck, but had too many other claw and teeth marks for them to count. Both are expected to recover. Investigators have not yet been able to speak to the mother because of the severity of her injuries . \"She may not remember the attack at all,\" said Hicks , who also said that it was only Tennessee's second documented attack involving a bear. Authorities have not yet released the names of the victims . The campground has been closed.", "edit_actions": [{"type": "R", "before": "Black Bear", "after": "black bear", "start_char_pos": 67, "end_char_pos": 77, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "visitng", "after": "visiting", "start_char_pos": 334, "end_char_pos": 341, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "hospital. The boys head was bitten by the bear, which punctured his skull, before it", "after": ". The boy's skull was punctured by a bite to the head, before the bear", "start_char_pos": 476, "end_char_pos": 560, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": ", so far, have not", "after": "have not yet", "start_char_pos": 846, "end_char_pos": 864, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "her injuries are too severe", "after": "of the severity of her injuries", "start_char_pos": 906, "end_char_pos": 933, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ". He", "after": ", who", "start_char_pos": 989, "end_char_pos": 993, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "and the", "after": ". The", "start_char_pos": 1135, "end_char_pos": 1142, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 27, 91, 168, 318, 385, 485, 581, 671, 801, 831, 990, 1075], "domain": "news"}
{"doc_id": "38988", "revision_depth": 1, "before_revision": "This week, the Public Broadcasting Service aired a NOVA program titled Dimming the Earth, which presented research by leading scientists on the complex and chaotic systems of our global climate and human activity's effect on it. One of the largest interactions (or \"inputs\") humans have with the atmosphere is the ever-increasing use of fossil fuels. Consumption has risen 2\\% per year for this decade. More recently, atmospheric scientists have come upon the phenomenon of the reduction of  sunlight reaching Earth's surface observing a nearly a 10\\% decline since 1960. This has been dubbed the \"global dimming\" effect, and is now understood to be due to the way these aerosols act upon clouds.  Many scientists now believe that global dimming caused by these pollutants has mitigated the temperature rises brought about by global warming. Over the last thirty years, Earth's temperature has increased by about one, to one and a half degrees Fahrenheit . In the absence of global dimming, however, the Earth might be two to three degrees warmer than it currently is, suggesting that a \"tug-of-war\" exists between greenhouse gases and particulates released by burning fossil fuels. Efforts to mitigate the human health dangers of smog have allowed more heat into our atmosphere and brought about a sharper increase in global warming.", "after_revision": "This week, the Public Broadcasting Service aired a NOVA program titled Dimming the Earth, which presented research by leading scientists on the complex systems of our global climate and human activity's effect on it. One of the largest interactions (or \"inputs\") humans have with the atmosphere is the ever-increasing use of fossil fuels. Consumption has risen 2\\% per year for this decade. More recently, atmospheric scientists have come upon the phenomenon of the reduction of direct sunlight reaching Earth's surface observing a nearly a 5\\% decline between 1960 and 1990, with evidence of a recovery since then. This has been dubbed the \"global dimming\" effect, and is probably due to the way these aerosols act upon clouds. It is important to realise that this does not represent a net loss of this much sunshine to the climate system - if so, large temperature declines would have been observed. Instead, the sunshine is absorbed elsewhere in the system, with a much smaller net loss. Many scientists now believe that global dimming caused by these pollutants has mitigated the temperature rises brought about by global warming. Over the last thirty years, Earth's temperature has increased by about 0.5 oC . In the absence of global dimming, however, the Earth might be 0.3 oC warmer than it currently is, suggesting that a \"tug-of-war\" exists between greenhouse gases and particulates released by burning fossil fuels. Efforts to mitigate the human health dangers of smog have allowed more heat into our atmosphere and brought about a sharper increase in global warming.", "edit_actions": [{"type": "D", "before": "and chaotic", "after": null, "start_char_pos": 152, "end_char_pos": 163, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "direct", "start_char_pos": 491, "end_char_pos": 491, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "10\\% decline since 1960.", "after": "5\\% decline between 1960 and 1990, with evidence of a recovery since then.", "start_char_pos": 547, "end_char_pos": 571, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "now understood to be", "after": "probably", "start_char_pos": 629, "end_char_pos": 649, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "It is important to realise that this does not represent a net loss of this much sunshine to the climate system - if so, large temperature declines would have been observed. Instead, the sunshine is absorbed elsewhere in the system, with a much smaller net loss.", "start_char_pos": 697, "end_char_pos": 697, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "one, to one and a half degrees Fahrenheit", "after": "0.5 oC", "start_char_pos": 913, "end_char_pos": 954, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "two to three degrees", "after": "0.3 oC", "start_char_pos": 1019, "end_char_pos": 1039, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 228, 350, 402, 571, 696, 841, 956, 1182]}
{"doc_id": "40920", "revision_depth": 2, "before_revision": "The tsunami occoured a couple of hours before there was going to be a tsunami warning system test, that had 29 countries participating.", "after_revision": "The tsunami occurred a couple of hours before a previously-scheduled test of a tsunami warning system . The test had 29 countries participating.", "edit_actions": [{"type": "R", "before": "occoured", "after": "occurred", "start_char_pos": 12, "end_char_pos": 20, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "there was going to be a", "after": "a previously-scheduled test of a", "start_char_pos": 46, "end_char_pos": 69, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "test, that", "after": ". The test", "start_char_pos": 93, "end_char_pos": 103, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0], "domain": "news"}
{"doc_id": "41347", "revision_depth": 1, "before_revision": "Braxton Bilbrey, a 7-year old boy from Glendale, California has swum from Alcatraz to Aquatic Park in San Francisco, California. The length of the swim is approximately 1.4 miles. Bilbrey  trained for four days a week , training 2 hours each time. Bilbrey has also participated in several swimming triathlons. Bilbrey got the idea to complete the swim after reading about Johnny Wilson, a 9-year old fourth grader who also completed the swim , taking just under two hours in October 2005. \"If you were to ask me if a 7-year-old is old enough to do it, I'd say maybe one out of 10 million, but he's that one,\" said Bilbreys coach, Joe Zemaitis. His coach made the swim along with Bilbrey. A United States Coast Guard boat followed the swimmers , along with 2 extra swimmers in case an emergency occurred.", "after_revision": "Braxton Bilbrey, a 7-year-old boy from Glendale, California , swam from Alcatraz to Aquatic Park in San Francisco, California. The length of the swim is approximately 1.4 miles. Bilbrey , who has participated in several swimming triathlons, trained for four days a week for 2 hours at a time. He decided to take the swim after reading about Johnny Wilson, a 9-year old fourth grader who also made it across , taking just under two hours in October 2005. \"If you were to ask me if a 7-year-old is old enough to do it, I'd say maybe one out of 10 million, but he's that one,\" said Bilbrey's coach, Joe Zemaitis. Zemaitis made the swim along with Bilbrey. A United States Coast Guard boat and two additional swimmers followed to take action in case an emergency occurred.", "edit_actions": [{"type": "R", "before": "7-year old", "after": "7-year-old", "start_char_pos": 19, "end_char_pos": 29, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "has swum", "after": ", swam", "start_char_pos": 60, "end_char_pos": 68, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ", who has participated in several swimming triathlons,", "start_char_pos": 188, "end_char_pos": 188, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", training", "after": "for", "start_char_pos": 218, "end_char_pos": 228, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "each time. Bilbrey has also participated in several swimming triathlons. Bilbrey got the idea to complete", "after": "at a time. He decided to take", "start_char_pos": 237, "end_char_pos": 342, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "completed the swim", "after": "made it across", "start_char_pos": 423, "end_char_pos": 441, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Bilbreys", "after": "Bilbrey's", "start_char_pos": 614, "end_char_pos": 622, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "His coach", "after": "Zemaitis", "start_char_pos": 644, "end_char_pos": 653, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "followed the swimmers , along with 2 extra swimmers", "after": "and two additional swimmers followed to take action", "start_char_pos": 721, "end_char_pos": 772, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 128, 179, 247, 309, 488, 643, 687], "domain": "news"}
{"doc_id": "41664", "revision_depth": 1, "before_revision": "Current evidence indicats that the reports of gunfire in the Rayburn Building were actually staffers and Congresspeople hearing the sound of construction work in the elevator. A mechannic was reported to be using a pneumatic hammer,  which, when echoing up the elvator shaft, could be misheard as gunfire by those unaccustomed to either sound. Reports of smoke have not been officially explained, though several comments have chalked the reports of smoke  as paranoia induced, suspecting people heard the warning of gunfire and attributed the origin of a slight or strange smell to the ignition of gunpowder.", "after_revision": "Current evidence indicates that the reports of gunfire in the Rayburn Building were actually staffers and Congresspeople hearing the sound of construction work in the elevator. A mechanic was reported to be using a pneumatic hammer, the sound of which, when echoing up the elvator shaft, could be misheard as gunfire by those unaccustomed to either sound. Reports of smoke have not been officially explained, though several comments have chalked the reports of smoke up as paranoia induced, suspecting people heard the warning of gunfire and attributed the origin of a slight or strange smell to the ignition of gunpowder.", "edit_actions": [{"type": "R", "before": "indicats", "after": "indicates", "start_char_pos": 17, "end_char_pos": 25, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "mechannic", "after": "mechanic", "start_char_pos": 178, "end_char_pos": 187, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "the sound of", "start_char_pos": 233, "end_char_pos": 233, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "up", "start_char_pos": 455, "end_char_pos": 455, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 175, 343], "domain": "news"}
{"doc_id": "41664", "revision_depth": 2, "before_revision": "The Senate was in session when the lockdown occurred, but the House of Representatives was not. Many members of congress had already departed in advance of the long Memorial Day holiday in the US. BBC News initally reported some witness as saying the sound may have been a car backfiring. The Associated Press and FOX News also reported speculation about fireworks or the gunfire having come from an indoor shooting range located in the basement of the Rayburn building. MSNBC suggested the noise may have been from a maintence man working on an elevator shaft near the area where the shots were heard. Current evidence indicates that the reports of gunfire in the Rayburn Building were actually staffers and Congresspeople hearing the sound of construction work in the elevator. A mechanic was reported to be using a pneumatic hammer, the sound of which, when echoing up the elvator shaft, could be misheard as gunfire by those unaccustomed to either sound. Reports of smoke have not been officially explained, though several comments have chalked the reports of smoke up as paranoia induced, suspecting people heard the warning of gunfire and attributed the origin of a slight or strange smell to the ignition of gunpowder.", "after_revision": "The Senate was in session when the lockdown occurred, but the House of Representatives was not. Many members of congress had already departed in advance of the long Memorial Day holiday in the US. BBC News initially reported some witness as saying the sound may have been a car backfiring. The Associated Press and FOX News also reported speculation about fireworks or the gunfire having come from an indoor shooting range located in the basement of the Rayburn building. MSNBC suggested the noise may have been from a maintenance man working on an elevator shaft near the area where the shots were heard. Current evidence indicates that the reports of gunfire in the Rayburn Building were actually staffers and Congresspeople hearing the sound of construction work in the elevator. A mechanic was reported to be using a pneumatic hammer, the sound of which, when echoing up the elevator shaft, could be misheard as gunfire by those unaccustomed to either sound. Reports of smoke have not been officially explained, though several comments have chalked the reports of smoke up as paranoia induced, suspecting people heard the warning of gunfire and attributed the origin of a slight or strange smell to the ignition of gunpowder.", "edit_actions": [{"type": "R", "before": "initally", "after": "initially", "start_char_pos": 206, "end_char_pos": 214, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "maintence", "after": "maintenance", "start_char_pos": 518, "end_char_pos": 527, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "elvator", "after": "elevator", "start_char_pos": 876, "end_char_pos": 883, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 95, 288, 470, 602, 779, 958], "domain": "news"}
{"doc_id": "42004", "revision_depth": 1, "before_revision": "The site operators speculate that the police wish to test the legality of the operation. Purportedly, this move comes due to political pressure exerted by American media corporations. Swedish public broadcaster Sveriges Television has backed this claim. Rickard Falkvinge has stated that the Svenska antipiratbyrn (Swedish Anti-Piracy Bureau) has admitted to being behind the police action, and he suspects the IFPI is also involved. Tobias Andersson, of Piratbyrn agrees, claiming that The Antipiratbyrn \"has clearly misled the police\" and \"has fooled the police into shutting down its antagonists, the Piratbyrn. The Antipiratbyrn previously gained notoriety by paying an infiltrator to plant copyrighted material on the Swedish ISP Bahnhof's servers. Dan Glickman, CEO of the Motion Picture Association of America  has applauded the move, citing it as a reminder that \"there are no safe harbors for copyright thieves\". Kori Bernards, an MPAA representative, claimed The Pirate Bay to be \"one of our No. 1 targets.\" The MPAA allege that The Pirate Bay made available over 150 thousand files, including summer blockbusters such as \"Mission: Impossible III\" and \"X-Men 3\". Initially, there were also rumors that this could be another stunt by The Pirate Bay to promote awareness. This possibility, however, has been already dismissed and only had some credibility given the fact that approximately one year ago today they claimed something similar like this, but instead were simply updating their servers. At current there are already some publications made available by proper officials , (see the MPAA link below) confirming the authenticity of the raid. Gottfrid Svartholm, owner of webhotel PRQ said \"I was arrested for interrogation, and they begun to ask question about Pirate Bay, what I knew and what connections I had to it. I got no public defender, and my regular legal representative was also under arrest, so I said nothing at all\". According to Mr. Svartholm, the police told him that they were not obligated to provide a public defender, as the crime he was accused for doesn't lead to jail sentence if he is found guilty.", "after_revision": "The site operators speculate that the police wish to test the legality of the operation. Purportedly, this move comes due to political pressure exerted by United States media corporations. Swedish public broadcaster Sveriges Television has backed this claim. Rickard Falkvinge has stated that the Svenska antipiratbyrn (Swedish Anti-Piracy Bureau) has admitted to being behind the police action, and he suspects the IFPI is also involved. Tobias Andersson, of Piratbyrn agrees, claiming that The Antipiratbyrn \"has clearly misled the police\" and \"has fooled the police into shutting down its antagonists, the Piratbyrn. The Antipiratbyrn previously gained notoriety by paying an infiltrator to plant copyrighted material on the Swedish ISP Bahnhof's servers. Dan Glickman, CEO of the Motion Picture Association of America (MPAA), has applauded the move, citing it as a reminder that \"there are no safe harbors for copyright thieves\". Kori Bernards, an MPAA representative, claimed The Pirate Bay to be \"one of our No. 1 targets.\" The MPAA allege that The Pirate Bay made available over 150 thousand files, including summer blockbusters such as \"Mission: Impossible III\" and \"X-Men 3\". Initially, there were also rumors that this could be another stunt by The Pirate Bay to promote awareness. This possibility, however, has been already dismissed and only had some credibility given the fact that approximately one year ago today they claimed something similar like this, but instead were simply updating their servers. At current there are already some publications made available by officials confirming the authenticity of the raid. Gottfrid Svartholm, owner of webhotel PRQ said \"I was arrested for interrogation, and they begun to ask question about Pirate Bay, what I knew and what connections I had to it. I got no public defender, and my regular legal representative was also under arrest, so I said nothing at all\". According to Mr. Svartholm, the police told him that they were not obligated to provide a public defender, as the crime he was accused of doesn't lead to jail sentence if he is found guilty.", "edit_actions": [{"type": "R", "before": "American", "after": "United States", "start_char_pos": 155, "end_char_pos": 163, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "A", "before": null, "after": "(MPAA),", "start_char_pos": 817, "end_char_pos": 817, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "proper officials , (see the MPAA link below)", "after": "officials", "start_char_pos": 1572, "end_char_pos": 1616, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "for", "after": "of", "start_char_pos": 2082, "end_char_pos": 2085, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 88, 183, 253, 433, 614, 753, 921, 1017, 1172, 1279, 1506, 1657, 1834, 1946], "domain": "news"}
{"doc_id": "42007", "revision_depth": 1, "before_revision": "Blocked Berliner Platz Up to 7,000 people took to the streets in w: Gieen today, protesting the planned introduction of tuition fees at Hessian universities. Beginning at 1pm CEST (1100 UTC), students, pupils and ordinary citizens marched through much of downtown Gieen in the biggest demostration the city has seen in years. Since 4pm , at least 200 people also blocked the Berliner Platz, the most important inner city crossing, bringing most of the downtown bus and car traffic to a standstill. The police has threated to clear the crossing by force.  Sour", "after_revision": "Blocked Berliner Platz Up to 7,000 people took to the streets in Gieen today, protesting the planned introduction of tuition fees at Hessian universities. Beginning at 1pm CEST (1100 UTC), students, pupils and ordinary citizens marched through much of downtown Gieen , accompanied by a massive police presence, in the largest demonstration the city has seen in years. From 4 to 7pm , at least 200 people also blocked the Berliner Platz, the most important inner city crossing, bringing most of the downtown bus and car traffic to a standstill. The crowd dispersed after the police threatend several times to clear the crossing by force. Protesters in front of the city office Students unions from all parts of Hesse called for the joint demonstration in Gieen. In July, the Hessian state legislature is due to pass a bill that would require all undergraduate students to pay a tuition fee of 500 per semester; graduate students could be charged up to 1,500.", "edit_actions": [{"type": "D", "before": "w:", "after": null, "start_char_pos": 65, "end_char_pos": 67, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": "in", "after": ", accompanied by a massive police presence, in the largest demonstration", "start_char_pos": 270, "end_char_pos": 272, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "biggest demostration the", "after": null, "start_char_pos": 277, "end_char_pos": 301, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Since 4pm", "after": "From 4 to 7pm", "start_char_pos": 326, "end_char_pos": 335, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "police has threated", "after": "crowd dispersed after the police threatend several times", "start_char_pos": 502, "end_char_pos": 521, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Protesters in front of the city office Students unions from all parts of Hesse called for the joint demonstration in Gieen.", "start_char_pos": 554, "end_char_pos": 554, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Sour", "after": "In July, the Hessian state legislature is due to pass a bill that would require all undergraduate students to pay a tuition fee of 500 per semester; graduate students could be charged up to 1,500.", "start_char_pos": 555, "end_char_pos": 559, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 157, 325, 497], "domain": "news"}
{"doc_id": "42380", "revision_depth": 2, "before_revision": "After the landing , at approximately 7 a.m. local time (2200 Saturday, UTC), of the aircraft that transported the Portuguese soldiers, hundreds of people started to enter the runway with flags of East Timor and Portugal. In addition to the hundreds of locals that were expecting the GNR soldiers, the Portuguese ambassador, Joo Ramos Pinto, and other local authorities received the soldiers. Jos Ramos Horta told Lusa News Agency that the soldiers of the GNR \"have to start acting with autonomy and hardness\" in all the city, Dili, to reestablish the law and order. The Foreign Minister also told the Agency that he \"respects\" the decision of the Portuguese government to put the GNR force out of the Australian command of the international force, affirming that is essential to reach an agreement this Monday, so that the Portuguese Republican National Guard \"can be operational.\" Ramos Horta added that this will be the position that he will defend in a meeting this Thursday morning with the diplomats of Australia, Portugal, New Zealand and Malaysia. This meeting will define the way that the forces of the four countries will work on the country, having the troops of Australia, New Zealand and Malaysia working under a single Australian command, and the Portuguese force working in an \"autonomous\" way and reporting directly to President Xanana Gusmo.", "after_revision": "After the landing of the aircraft that transported the Portuguese soldiers at approximately 7 a.m. local time (2200 Saturday, UTC), hundreds of people started to enter the runway with flags of East Timor and Portugal. In addition to the hundreds of locals that were expecting the GNR soldiers, the Portuguese ambassador, Joo Ramos Pinto, and other local authorities received the soldiers. Jos Ramos Horta told Lusa News Agency that the soldiers of the GNR \"have to start acting with autonomy and hardness\" in the entire city, Dili, to reestablish the law and order. The Foreign Minister also told the Agency that he \"respects\" the decision of the Portuguese government to put the GNR force out of the Australian command of the international force, affirming that is essential to reach an agreement this Monday, so that the Portuguese Republican National Guard \"can be operational.\" Ramos Horta added that this would be the position that he will defend in a meeting this Thursday morning with the diplomats of Australia, Portugal, New Zealand and Malaysia. This meeting will define the way that the forces of the four countries will work on the country, having the troops of Australia, New Zealand and Malaysia working under a single Australian command, and the Portuguese force working in an \"autonomous\" way and reporting directly to President Xanana Gusmo.", "edit_actions": [{"type": "R", "before": ",", "after": "of the aircraft that transported the Portuguese soldiers", "start_char_pos": 18, "end_char_pos": 19, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "of the aircraft that transported the Portuguese soldiers,", "after": null, "start_char_pos": 77, "end_char_pos": 134, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "all the", "after": "the entire", "start_char_pos": 512, "end_char_pos": 519, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "will", "after": "would", "start_char_pos": 910, "end_char_pos": 914, "major_intent": "style", "raw_intents": ["style", "fluency", "style"]}], "sents_char_pos": [0, 220, 391, 565, 1054], "domain": "news"}
{"doc_id": "42846", "revision_depth": 1, "before_revision": "Shiv sena activists today disrupted a news conference of Syed Ali Shah Geelani, the leader of the break away fraction of the All Party Hurriyat Conference (APHC) . Geelani, who heads a breakaway faction of the troubled region 's largest separatist alliance , was scheduled to address the press after his release from prison. He was arrested on Wednesday in south Kashmir on charges of inciting people against the Indian state during a series of speeches in Poonch district of South Kashmir in 2001.", "after_revision": "Activists of the Shiv sena today disrupted a news conference of Syed Ali Shah Geelani, the leader of the breakaway fraction of the All Party Hurriyat Conference (APHC) Kashmir 's largest separatist alliance . Geelani was scheduled to address the press after his release from prison. He was arrested on Wednesday in south Kashmir on charges of inciting people against the Indian state during a series of speeches in Poonch district of South Kashmir in 2001.", "edit_actions": [{"type": "R", "before": "Shiv sena activists", "after": "Activists of the Shiv sena", "start_char_pos": 0, "end_char_pos": 19, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "break away", "after": "breakaway", "start_char_pos": 98, "end_char_pos": 108, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ". Geelani, who heads a breakaway faction of the troubled region", "after": "Kashmir", "start_char_pos": 162, "end_char_pos": 225, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "meaning-changed"]}, {"type": "R", "before": ",", "after": ". Geelani", "start_char_pos": 257, "end_char_pos": 258, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 324], "domain": "news"}
{"doc_id": "43194", "revision_depth": 1, "before_revision": "The broadcasters: TVNZ, CanWest, M\u0101ori Television, the TAB and Radio New Zealand have forged an alliance called FreeView which has leased satellite space for the digital television. Digital television will mean viewers can receive a crisper picture, clearer radio signals, limited reception problems, and more channels. It will require television viewers to buy a set top box, similar to the SKY Network Television decoder. Some viewers in some areas will need to buy a satellite dish as well. The approximated cost will be $200 or, with a dish, $400 (NZD) , however there is no subscription costs , unlike Sky Television.", "after_revision": "The broadcasters: TVNZ, CanWest, M\u0101ori Television, the TAB and Radio New Zealand have forged an alliance called FreeView which has leased satellite space for the digital television. Digital television will mean viewers can receive a crisper picture, clearer radio signals, limited reception problems, and more channels. FreeView will require television viewers to buy a set top box, similar to the SKY Network Television decoder. Some viewers in some areas will need to buy a satellite dish as well. The approximated cost will be $200 or, with a dish, $400 (NZD) . However, there is no subscription cost , unlike Sky Television.", "edit_actions": [{"type": "R", "before": "It", "after": "FreeView", "start_char_pos": 320, "end_char_pos": 322, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": ", however", "after": ". However,", "start_char_pos": 557, "end_char_pos": 566, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "costs", "after": "cost", "start_char_pos": 592, "end_char_pos": 597, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 181, 319, 423, 493], "domain": "news"}
{"doc_id": "45206", "revision_depth": 1, "before_revision": "North Carolina Governor Mike Easley signed into law a new minimum wage for the U.S. state Thursday. The increase of $1 will raise the rate to $ 6.15. Easley said that the new law will make it easier for North Carolinians to deal with the rising cost of living. The new law will become effective on January 1 next year. This is the first raise to the North Carolina minimum wage in nearly nine years. Last Sunday in Pennsylvania, Governor Ed Rendell signed into law a phased-in $2 minimum wage increase. For businesses with 10 employees or more, the rate will increase by $ 1 on January 1 next year, then increase by another dollar on July 1. Other U.S. states totaling 20 Alaska, California, Connecticut, Delaware, Florida, Hawaii, Illinois, Maine, Maryland, Massachusetts, Michigan, Minnesota, New Jersey, New York, North Carolina, Oregon, Rhode Island, Vermont, Washington and Wisconsin, plus the District of Columbia have raised the minimum above the federal mandate. The U.S. Fair Labor Standards Act, enacted in 1938, has been continuously amended by Congress over years with increases that make the federally mandated minimum wage what it is today,  $5.15 per hour. The last time the federal government raised the rate was in 1997.", "after_revision": "North Carolina Governor Mike Easley signed into law a new minimum wage for the U.S. state Thursday. The increase of US$1 will raise the rate to US$ 6.15. Easley said that the new law will make it easier for North Carolinians to deal with the rising cost of living. The new law will become effective on January 1 next year. This is the first raise to the North Carolina minimum wage in nearly nine years. Last Sunday in Pennsylvania, Governor Ed Rendell signed into law a phased-in US$2 minimum wage increase. For businesses with 10 employees or more, the rate will increase by US$ 1 on January 1 next year, then increase by another dollar on July 1. A total of 20 other U.S. states Alaska, California, Connecticut, Delaware, Florida, Hawaii, Illinois, Maine, Maryland, Massachusetts, Michigan, Minnesota, New Jersey, New York, North Carolina, Oregon, Rhode Island, Vermont, Washington and Wisconsin, plus the District of Columbia have raised the minimum above the federal mandate. The U.S. Fair Labor Standards Act, enacted in 1938, has been continuously amended by Congress over years with increases that make the federally mandated minimum wage what it is today, US $5.15 per hour. The last time the federal government raised the rate was in 1997.", "edit_actions": [{"type": "R", "before": "$1 will raise the rate to $", "after": "US$1 will raise the rate to US$", "start_char_pos": 116, "end_char_pos": 143, "major_intent": "style", "raw_intents": ["style", "style", "meaning-changed"]}, {"type": "R", "before": "$2 minimum wage increase. For businesses with 10 employees or more, the rate will increase by $", "after": "US$2 minimum wage increase. For businesses with 10 employees or more, the rate will increase by US$", "start_char_pos": 477, "end_char_pos": 572, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Other", "after": "A total of 20 other", "start_char_pos": 642, "end_char_pos": 647, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "D", "before": "totaling 20", "after": null, "start_char_pos": 660, "end_char_pos": 671, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "US", "start_char_pos": 1155, "end_char_pos": 1155, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 99, 149, 260, 318, 399, 502, 641, 970, 1171], "domain": "news"}
{"doc_id": "45505", "revision_depth": 1, "before_revision": "BBC Wales and iPTV station , VTV are reporting that a light aircraft made an emergency landing at 13:00 BST , at Treforest Industrial Estate , near Cardiff, south Wales . The aircraft, an orange-coloured single-engine model Fox home-built kit , was carrying two people , a man and a woman, when it made the forced landing on the estate's dual carriageway , which runs through the industrial park . It is not known how severely the two persons on board , are injured, however a spokesperson for South Wales Police has stated that both of the aircraft's occupants had sustained \"non life-threatening injuries\". Treforest Industrial Estate is one of the largest business parks in the south Wales valleys and its location, just two hundred metres from the main A470 expressway , could have resulted in a far greater tragedy. VTV reporter and private pilot , Craig Handley , said that the location of the accident \"would have been the most logical place to try and set the aircraft down as the dual carriageway, which runs through the estate, is clearly visible from the air and is almost completely straight and wide.\" Today's accident comes less than a year after a similar-sized aircraft was forced to make an emergency landing , less than two miles away from Treforest Industrial Estate , after suffering a sudden loss of engine power , during its approach to nearby Cardiff International Airport. That aircraft, a Cessna 150, collided with a group of trees , shedding its wings ; before coming to rest on a small country road. The two occupants of the aircraft received slight injuries. South Wales Police have not confirmed whether any motorists or pedestrians were injured when the aircraft , involved in today's accident , made its emergency landing. The aircraft did , however, cause some structural damage to a factory unit on the industrial estate , before coming to a rest, inverted, on the adjacent dual carriageway. According to one eyewitness, the aircraft appeared to have experienced an engine-related problem , seconds before impact. A spokesperson, for Cardiff International Airport said that the aircraft had departed , from Cardiff , at 12:43 BST and that the airport's Air Traffic Control centre had lost contact with the plane. The Civil Aviation Authority's Air Accident Investigation Branch (AAIB) have been informed and are due to begin their investigation , into the crash.", "after_revision": "BBC Wales and iPTV station VTV are reporting that a light aircraft made an emergency landing at 13:00 BST at Treforest Industrial Estate near Cardiff . The aircraft, an orange-coloured single-engine model Fox home-built kit was carrying two people when it made a forced landing on a section of dual carriageway within the industrial estate . It is not known how severely the man and woman on board were injured, a spokesperson for South Wales Police stated that both of the aircraft's occupants sustained \"non life-threatening injuries\". Treforest Industrial Estate is one of the largest business parks in the south Wales valleys . Its location, just two hundred metres from the main A470 expressway could have resulted in a far greater tragedy. VTV reporter and private pilot Craig Handley said that the location of the accident \"would have been the most logical place to try and set the aircraft down as the dual carriageway, which runs through the estate, is clearly visible from the air and is almost completely straight and wide.\" Today's accident comes less than a year after a similar-sized aircraft was forced to make an emergency landing less than two miles away from Treforest Industrial Estate . A Cessna 150 suffered a sudden loss of engine power during its approach to nearby Cardiff International Airport. The Cessna in the earlier accident collided with a group of trees shedding its wings before coming to rest on a small country road. The two occupants of that aircraft received slight injuries. South Wales Police have not confirmed whether any motorists or pedestrians were injured when the aircraft in today's accident made its emergency landing. The aircraft did cause some structural damage to a factory unit on the industrial estate prior to coming to rest, inverted, on the adjacent dual carriageway. According to one eyewitness, the aircraft appeared to have experienced an engine-related problem seconds before impact. A spokesperson, for Cardiff International Airport said that the aircraft had departed from Cardiff at 12:43 BST and that the airport's Air Traffic Control centre had lost contact with the plane. The Civil Aviation Authority's Air Accident Investigation Branch (AAIB) have been informed and are due to begin their investigation into the crash.", "edit_actions": [{"type": "D", "before": ",", "after": null, "start_char_pos": 27, "end_char_pos": 28, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 108, "end_char_pos": 109, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": ", near Cardiff, south Wales", "after": "near Cardiff", "start_char_pos": 141, "end_char_pos": 168, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 243, "end_char_pos": 244, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "D", "before": ", a man and a woman,", "after": null, "start_char_pos": 269, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the", "after": "a", "start_char_pos": 303, "end_char_pos": 306, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "the estate's dual carriageway , which runs through the industrial park", "after": "a section of dual carriageway within the industrial estate", "start_char_pos": 325, "end_char_pos": 395, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "two persons on board , are injured, however", "after": "man and woman on board were injured,", "start_char_pos": 431, "end_char_pos": 474, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "has", "after": null, "start_char_pos": 513, "end_char_pos": 516, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "D", "before": "had", "after": null, "start_char_pos": 562, "end_char_pos": 565, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "and its", "after": ". Its", "start_char_pos": 701, "end_char_pos": 708, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 773, "end_char_pos": 774, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", Craig Handley ,", "after": "Craig Handley", "start_char_pos": 852, "end_char_pos": 869, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1226, "end_char_pos": 1227, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": ", after suffering", "after": ". A Cessna 150 suffered", "start_char_pos": 1286, "end_char_pos": 1303, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1334, "end_char_pos": 1335, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "That aircraft, a Cessna 150,", "after": "The Cessna in the earlier accident", "start_char_pos": 1397, "end_char_pos": 1425, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1457, "end_char_pos": 1458, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "D", "before": ";", "after": null, "start_char_pos": 1478, "end_char_pos": 1479, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the", "after": "that", "start_char_pos": 1548, "end_char_pos": 1551, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": ", involved", "after": null, "start_char_pos": 1693, "end_char_pos": 1703, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1724, "end_char_pos": 1725, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "D", "before": ", however,", "after": null, "start_char_pos": 1771, "end_char_pos": 1781, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": ", before coming to a", "after": "prior to coming to", "start_char_pos": 1854, "end_char_pos": 1874, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 2022, "end_char_pos": 2023, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", from Cardiff ,", "after": "from Cardiff", "start_char_pos": 2133, "end_char_pos": 2149, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 2378, "end_char_pos": 2379, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 170, 397, 608, 820, 1114, 1396, 1479, 1526, 1586, 1753, 1924, 2046, 2245], "domain": "news"}
{"doc_id": "45781", "revision_depth": 1, "before_revision": "Temperatures are nearing 100F (38C). 80 mph (130 km/h) winds accompanied by heavy rain caused many tree limbs to fall, some onto cars, some onto thoroughfares, and some on high-tension power lines that supply the city. It also caused at least three building collapses. After 160,000 people have had their power restored,  500,000 people are without power. Three hundred national guardsmen join volunteers, policemen and firefighters in aiding people during the crisis. Vulnerable residents in nursing homes and centers for the elderly are being evacuated. Cookies and water are being distributed to those needing them.  Ice was in short supply as customers grabbed all they could get their hands on. The local power company, AmerenUE, estimated that as of 10:26 p.m. local Central time on the 20th, 418,195 customers had no electricity. Sour", "after_revision": " 80 mph (130 km/h) winds accompanied by heavy rain caused many tree limbs to fall, some onto cars, some onto thoroughfares, and some on high-tension power lines that supply the city. It also caused at least three building collapses. After 160,000 people have had their power restored, but 500,000 others are still without power. Three hundred national guardsmen joined volunteers, policemen and firefighters in aiding people during the crisis. Vulnerable residents in nursing homes and centers for the elderly are being evacuated. Cookies and water are being distributed to those needing them. With temperatures were nearing 100F (38C), ice was in short supply as customers grabbed all they could get their hands on. The local power company, AmerenUE, estimated that as of 10:26 p.m. local Central time on the 20th, 418,195 customers had no electricity. ", "edit_actions": [{"type": "D", "before": "Temperatures are nearing 100F (38C).", "after": null, "start_char_pos": 0, "end_char_pos": 36, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "but", "start_char_pos": 321, "end_char_pos": 321, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "R", "before": "people are", "after": "others are still", "start_char_pos": 330, "end_char_pos": 340, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "join", "after": "joined", "start_char_pos": 389, "end_char_pos": 393, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "A", "before": null, "after": "With temperatures were nearing 100F (38C), ice was in short supply as customers grabbed all they could get their hands on.", "start_char_pos": 619, "end_char_pos": 619, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "Ice was in short supply as customers grabbed all they could get their hands on.", "after": null, "start_char_pos": 620, "end_char_pos": 699, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "Sour", "after": null, "start_char_pos": 837, "end_char_pos": 841, "major_intent": "others", "raw_intents": ["others", "others", "others"]}], "sents_char_pos": [0, 36, 218, 268, 355, 468, 555, 618, 699], "domain": "news"}
{"doc_id": "45789", "revision_depth": 1, "before_revision": "Ta Mok's real name is not known, but reported to be Ek Choeun or Oeung Choeun. He joined the Khmer Rouge after being part of anti-French colonial and anti-Japanese resistence movements, rising to become the Maoist group's Chief-of-Staff. He is believed to have directed the purges of the Democratic Kampuchea regime, in which an estimated 1.7 million people were tortured and killed in what is now referred to as \"the killing fields\", earning him the name \u201cThe Butcher\u201d . After the fall of the regime, he continued to lead a faction, until captured by the Cambodian forces in 1999. Youk Chang, director of the Documentation Centre of Cambodia which is studying the crimes committed by the Khmer Rouge, said his death was \u201coutrageous\u201d . \"Some people may be happy with this, \u201d he told AP, \u201c but not the victims who have been waiting for justice for a long time,\" Ta Mok's nephew Morm Mol said he was \u201csaddened\u201d by his uncle \u2019 s death.", "after_revision": "Ta Mok's real name is not known, but reported to be Ek Choeun or Oeung Choeun. He joined the Khmer Rouge after being part of anti-French colonial and anti-Japanese resistance movements, rising to become the Maoist group's Chief-of-Staff. He is believed to have directed the purges of the Democratic Kampuchea regime, in which an estimated 1.7 million people were tortured and killed in what is now referred to as \"the killing fields\", earning him the name \"The Butcher\" . After the fall of the regime, he continued to lead a faction, until captured by the Cambodian forces in 1999. Youk Chang, director of the Documentation Centre of Cambodia which is studying the crimes committed by the Khmer Rouge, said his death was \"outrageous\" . \"Some people may be happy with this, \" he told AP, \" but not the victims who have been waiting for justice for a long time,\" Ta Mok's nephew Morm Mol said he was \"saddened\" by his uncle ' s death.", "edit_actions": [{"type": "R", "before": "resistence", "after": "resistance", "start_char_pos": 164, "end_char_pos": 174, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "\u201cThe Butcher\u201d", "after": "\"The Butcher\"", "start_char_pos": 456, "end_char_pos": 469, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "\u201coutrageous\u201d", "after": "\"outrageous\"", "start_char_pos": 721, "end_char_pos": 733, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "\u201d", "after": "\"", "start_char_pos": 773, "end_char_pos": 774, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "\u201c", "after": "\"", "start_char_pos": 787, "end_char_pos": 788, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "\u201csaddened\u201d", "after": "\"saddened\"", "start_char_pos": 898, "end_char_pos": 908, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "R", "before": "\u2019", "after": "'", "start_char_pos": 922, "end_char_pos": 923, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}], "sents_char_pos": [0, 78, 237, 581], "domain": "news"}
{"doc_id": "45878", "revision_depth": 1, "before_revision": "LGBT activist Gastons Lakombe told Apollo  portal that the most dangerous situation occured on the Jurmala highway, when a few cars tried to \"drive a taxi off\" the roadwith gay activists inside . According to Lakombe, protesters attacked the taxi car first at the Reval Hotel, Latvia, by throwing eggs and bottles of water. A few cars followed and on the Jurmala highway \"made some dangerous maneuvers, that seriously frightened passengers\", said Imants Kozlovskis, an LGBT activist who was inside the taxi.", "after_revision": "LGBT activist Gastons Lakombe told Apollo internet portal that the most dangerous incident occured on the Jurmala highway, when a few cars tried to \"drive a taxi ferrying gay activists off the road . According to Lakombe, protesters attacked the taxi first at the Reval Hotel, Latvia, by throwing eggs and bottles of water. A few cars followed and on the Jurmala highway \"made some dangerous maneuvers, that seriously frightened passengers\", said Imants Kozlovskis, an LGBT activist who was inside the taxi.", "edit_actions": [{"type": "A", "before": null, "after": "internet", "start_char_pos": 42, "end_char_pos": 42, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "situation", "after": "incident", "start_char_pos": 74, "end_char_pos": 83, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "off\" the roadwith gay activists inside", "after": "ferrying gay activists off the road", "start_char_pos": 155, "end_char_pos": 193, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "car", "after": null, "start_char_pos": 247, "end_char_pos": 250, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 195, 323], "domain": "news"}
{"doc_id": "45991", "revision_depth": 1, "before_revision": "A large number of media and people were gathered outside the packed courtroom. Twenty to thirty people had to remain outside. Many of the people who had gathered outside the courtroom insulted the accused man as he walked to the courtroom. One said \"You'll never make sentencing.\" Armed police also were present for the accused's protection . Related Stories Wikinews Teacher found murdered in New Zealand classroom July 17, 2006", "after_revision": "A large crowd of media personnel and onlookers gathered outside the packed courtroom. Twenty to thirty people had to remain outside. Many of the people who had gathered outside the courtroom abused the accused man as he walked to the courtroom. One said \"You'll never make sentencing.\" Armed police also were deployed to protect the alleged murderer . Related Stories ", "edit_actions": [{"type": "R", "before": "number of media and people were", "after": "crowd of media personnel and onlookers", "start_char_pos": 8, "end_char_pos": 39, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "insulted", "after": "abused", "start_char_pos": 184, "end_char_pos": 192, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "present for the accused's protection", "after": "deployed to protect the alleged murderer", "start_char_pos": 304, "end_char_pos": 340, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "Wikinews Teacher found murdered in New Zealand classroom July 17, 2006", "after": null, "start_char_pos": 359, "end_char_pos": 429, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 78, 125, 239, 280], "domain": "news"}
{"doc_id": "46844", "revision_depth": 1, "before_revision": "Hillside suburbs in Christchurch, New Zealand have been suffering from torrential rain and flooding since Monday night.  Around 60-millimetres of rain has fallen. The Heathcote river in Beckenham has burst its banks and has gone onto Eastern Terrace, making some parts of the road impassable. It has also made some roads around the area fully submerged in water . Terry Sefton, Christchurch City Council spokesman, said \"Eastern Terrace between Birdwood and Sandwich Roads is impassable and some parts of Palentine and Riverlaw terraces are covered in water.\"", "after_revision": "Hillside suburbs in Christchurch, New Zealand have been suffering from torrential rain since Monday night. The Heathcote river burst its banks and caused surface flooding in areas. Around 60-millimetres of rain has fallen. The Heathcote river in Beckenham has inundated areas in Eastern Terrace, making some roads impassable . Terry Sefton, Christchurch City Council spokesman, said \"Eastern Terrace between Birdwood and Sandwich Roads is impassable and some parts of Palentine and Riverlaw terraces are covered in water.\"", "edit_actions": [{"type": "D", "before": "and flooding", "after": null, "start_char_pos": 87, "end_char_pos": 99, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "The Heathcote river burst its banks and caused surface flooding in areas.", "start_char_pos": 120, "end_char_pos": 120, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "burst its banks and has gone onto", "after": "inundated areas in", "start_char_pos": 200, "end_char_pos": 233, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "parts of the road impassable. It has also made some roads around the area fully submerged in water", "after": "roads impassable", "start_char_pos": 263, "end_char_pos": 361, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 119, 162, 292, 363], "domain": "news"}
{"doc_id": "47001", "revision_depth": 1, "before_revision": "Merseyside Police concluded after a brief inspection that the controversial Blog Liverpool Evil Cabal does not break criminal law. However the council officers continue to search for the editor. The Blog has been block at terminals controlled by Liverpool Direct Ltd, a company jointly owned by Liverpool City Council and British Telecom. The councils elected officials have denied ordering the block and are investigating who ordered the block. Previous fights between Liverpool City Councillors and Executive staff have resulted in controversy before  most notably in the resignation of leader Mike Storey and CEO Henshaw. Releated Stories Blogger.com made website called Liverpool Evil Cabal ", "after_revision": "Merseyside Police concluded after a brief inspection that the controversial blog Liverpool Evil Cabal does not break criminal law. However the council officers continue to search for the editor. The blog has been blocked on computers controlled by Liverpool Direct Ltd, a company jointly owned by Liverpool City Council and British Telecom. The council's elected officials have denied ordering the block and are investigating who ordered the block. Previous fights between Liverpool city councillors and executive staff have resulted in controversy before , most notably in the resignation of leader Mike Storey and CEO Henshaw. Related Stories External links Liverpool-evil-cabal blog", "edit_actions": [{"type": "R", "before": "Blog", "after": "blog", "start_char_pos": 76, "end_char_pos": 80, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "Blog has been block at terminals", "after": "blog has been blocked on computers", "start_char_pos": 199, "end_char_pos": 231, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "councils", "after": "council's", "start_char_pos": 343, "end_char_pos": 351, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "City Councillors and Executive", "after": "city councillors and executive", "start_char_pos": 480, "end_char_pos": 510, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 553, "end_char_pos": 553, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Releated Stories Blogger.com made website called Liverpool Evil Cabal", "after": "Related Stories", "start_char_pos": 625, "end_char_pos": 694, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "External links Liverpool-evil-cabal blog", "start_char_pos": 695, "end_char_pos": 695, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 130, 194, 338, 445, 624], "domain": "news"}
{"doc_id": "47572", "revision_depth": 1, "before_revision": "Private Meredith Simms, 18-years-old, died yesterday when the New Zealand Army Mercedes-Benz Unimog  he was travelling in left the road and fell 100-metres down a steep bank 2.30 p.m. yesterday (NZST), resting on railway tracks. Another solider was critically injured and was flown by the Westpac Rescue Helicopter to Wellington Hospital, he also has spinal cord injuries. The New Zealand police have launched an inquiry into the fatal accident. The Unimog has been removed the the scene of the crash and will now be examined. The police say that they will guard the scene and finish examining it today. Simms  death has been referred to a coroner.", "after_revision": "Private Meredith Simms, 18-years-old, died yesterday when the New Zealand Army Mercedes-Benz Unimog truck he was travelling in left the road and fell 100-metres down a steep bank 2.30 p.m. yesterday (NZST), coming to rest on railway tracks. The soldier driving the truck was critically injured and was flown by the Westpac Rescue Helicopter to Wellington Hospital, where he is being treated for spinal cord injuries. The New Zealand police have launched an inquiry into the fatal accident. The Unimog has been removed from the scene of the crash and will now be examined. The police say that they will guard the scene and finish examining it today. Simms ' death has been referred to a coroner.", "edit_actions": [{"type": "A", "before": null, "after": "truck", "start_char_pos": 100, "end_char_pos": 100, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "resting", "after": "coming to rest", "start_char_pos": 202, "end_char_pos": 209, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Another solider", "after": "The soldier driving the truck", "start_char_pos": 229, "end_char_pos": 244, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "he also has", "after": "where he is being treated for", "start_char_pos": 339, "end_char_pos": 350, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the the", "after": "from the", "start_char_pos": 474, "end_char_pos": 481, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "'", "start_char_pos": 610, "end_char_pos": 610, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 228, 372, 445, 526, 603], "domain": "news"}
{"doc_id": "47695", "revision_depth": 1, "before_revision": "Government jets bombed the LTTE front line in the Jaffna peninsula as fighting continued in the ethnic strife.  Aid workers are demanding access to the  peninsula which has  been cut off form the outside world. They have warned that local populations have low supplies to make it through     .", "after_revision": "Sri Lanka government jets bombed the LTTE on front lines in the battle for the Jaffna peninsula as fighting continued in the ethnic strife. The last week of fighting has claimed the lives of hundreds. The government air strikes continued on targets in the northern rebel-held territory of Kilinochchi and near the eastern port town of Trincomalee. The peninsula which has largely been cut off form the outside world. Even communications are near impossibel. Tens of thousands have fled their homes. There are shortages of food and lines for fuel. During a brief break in the fighting, the UN tried to get food and water to some of the worst off areas. \"We need both sides to stop fighting so we can get proper access to the area,\" said UNHCR representative Amin Awad. \"Food is getting low and we have worries about water and sanitation.\" The World Food Program said it was feeding more than 50,000 people in camps who fled fighting in and around Trincomalee, where there was only sporadic fire overnight. The ethnic conflict in Sri Lanka is an ongoing conflict between the Government of Sri Lanka and the ethnic Sri Lankan Tamils on the island-nation of Sri Lanka. Since the 1983 \"Black July\" pogrom, there has been on-and-off civil war, mostly between the government and the LTTE who want to create an independent state of called Tamil Eelam in the north and east of the island. It is estimated that the war has left 65,000 people dead since 1983 and caused great harm to the population and economy of the country. A cease-fire was declared in 2002, but hostilities renewed in late 2005 following military operations against Tiger-controlled territory in the east .", "edit_actions": [{"type": "R", "before": "Government", "after": "Sri Lanka government", "start_char_pos": 0, "end_char_pos": 10, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "front line in the", "after": "on front lines in the battle for the", "start_char_pos": 32, "end_char_pos": 49, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "The last week of fighting has claimed the lives of hundreds.", "start_char_pos": 111, "end_char_pos": 111, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Aid workers are demanding access to the", "after": "The government air strikes continued on targets in the northern rebel-held territory of Kilinochchi and near the eastern port town of Trincomalee.", "start_char_pos": 112, "end_char_pos": 151, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "The", "start_char_pos": 152, "end_char_pos": 152, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "largely", "start_char_pos": 173, "end_char_pos": 173, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "They have warned that local populations have low supplies to make it through", "after": "Even communications are near impossibel. Tens of thousands have fled their homes. There are shortages of food and lines for fuel.", "start_char_pos": 211, "end_char_pos": 287, "major_intent": "meaning-changed", "raw_intents": ["others", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "During a brief break in the fighting, the UN tried to get food and water to some of the worst off areas.", "start_char_pos": 288, "end_char_pos": 288, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "\"We need both sides to stop fighting so we can get proper access to the area,\" said UNHCR representative Amin Awad. \"Food is getting low and we have worries about water and sanitation.\"", "start_char_pos": 289, "end_char_pos": 289, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "The World Food Program said it was feeding more than 50,000 people in camps who fled fighting in and around Trincomalee, where there was only sporadic fire overnight.", "start_char_pos": 290, "end_char_pos": 290, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "The ethnic conflict in Sri Lanka is an ongoing conflict between the Government of Sri Lanka and the ethnic Sri Lankan Tamils on the island-nation of Sri Lanka. Since the 1983 \"Black July\" pogrom, there has been on-and-off civil war, mostly between the government and the LTTE who want to create an independent state of called Tamil Eelam in the north and east of the island. It is estimated that the war has left 65,000 people dead since 1983 and caused great harm to the population and economy of the country. A cease-fire was declared in 2002, but hostilities renewed in late 2005 following military operations against Tiger-controlled territory in the east", "start_char_pos": 291, "end_char_pos": 291, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 110, 210], "domain": "news"}
{"doc_id": "47695", "revision_depth": 2, "before_revision": "The government air strikes continued on targets in the northern rebel-held territory of Kilinochchi and near the eastern port town of Trincomalee. The peninsula  which has largely been cut off form the outside world . Even communications are near impossibel. Tens of thousands have fled their homes. There are shortages of food and lines for fuel. During a brief break in the fighting, the UN tried to get food and water to some of the worst off areas. The World Food Program said it was feeding more than 50,000 people in camps who fled fighting in and around Trincomalee, where there was only sporadic fire overnight.", "after_revision": "Government air strikes continued on targets in the northern rebel-held territory of Kilinochchi and near the eastern port town of Trincomalee. The peninsula , which has largely been cut off form the outside world , has seen tens of thousands have flee their homes. There are shortages of food and lines for fuel. During a brief break in the fighting, the UN tried to get food and water to some of the worst-off areas. The World Food Program said it was feeding more than 50,000 people in camps who fled fighting in and around Trincomalee, where there was only sporadic gunfire overnight.", "edit_actions": [{"type": "R", "before": "The government", "after": "Government", "start_char_pos": 0, "end_char_pos": 14, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 161, "end_char_pos": 161, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ". Even communications are near impossibel. Tens", "after": ", has seen tens", "start_char_pos": 216, "end_char_pos": 263, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "fled", "after": "flee", "start_char_pos": 282, "end_char_pos": 286, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "worst off", "after": "worst-off", "start_char_pos": 436, "end_char_pos": 445, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "fire", "after": "gunfire", "start_char_pos": 604, "end_char_pos": 608, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 146, 217, 258, 299, 347, 452], "domain": "news"}
{"doc_id": "48081", "revision_depth": 1, "before_revision": "'s town centre in Ireland was evacuated on August 23rd, 2006. The decision was made after  observed a man and a device believed to be dangerous in a house. Bomb disposal experts were on standby. The incident ended peacefully on August 24th, 2006 and the man is believed to be Jim Hourigan, a separated father of three from , who was taken into Garda custody. Mr Hourigan released a statement through his solicitor, before he was placed in Gardai custody. It states he is frustrated by the delays in achieving justice for being sexually abused while training as a  . Roscrea town centre remains sealed off, as Garda technical experts examine the house in which Mr Hourigan's live .", "after_revision": "The centre of Roscrea, a small town in County Tipperary in Ireland was evacuated on August 23rd, 2006. The decision was made after Gardai (police) observed a man and a device believed to be dangerous in a house. Bomb disposal experts were on standby. The incident ended peacefully on August 24th, 2006 and the man , identified as Jim Hourigan, a separated father of three from Limerick, was taken into Garda custody. Mr Hourigan released a statement through his solicitor, before he was placed in Gardai custody. It states he is frustrated by the delays in achieving justice for being sexually abused while training as a Christian Brother . Roscrea town centre remains sealed off, as Garda technical experts examine Mr Hourigan's house .", "edit_actions": [{"type": "R", "before": "'s town centre in", "after": "The centre of Roscrea, a small town in County Tipperary in", "start_char_pos": 0, "end_char_pos": 17, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "A", "before": null, "after": "Gardai (police)", "start_char_pos": 90, "end_char_pos": 90, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "is believed to be", "after": ", identified as", "start_char_pos": 258, "end_char_pos": 275, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": ", who", "after": "Limerick,", "start_char_pos": 323, "end_char_pos": 328, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "A", "before": null, "after": "Christian Brother", "start_char_pos": 563, "end_char_pos": 563, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}, {"type": "D", "before": "the house in which", "after": null, "start_char_pos": 641, "end_char_pos": 659, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "live", "after": "house", "start_char_pos": 674, "end_char_pos": 678, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 61, 155, 194, 358, 454], "domain": "news"}
{"doc_id": "48755", "revision_depth": 1, "before_revision": "A Russian trapeze artist, Vitaly Kharapavitski, 26 , performing at a circus in Scariff, Ireland has died after a cage attached to a hot air balloon crushed the performer. His wife, Sasha, 24 , who was also performing at the time, was injured, breaking her arm. Both were dressed in clown suits. At least 100 people were attending the circus and watched the acrobat fall. Most of the attendees were children. According to witnesses, the balloon caught fire and the cage fell on top of the acrobat , crushing his head. \"We had to discharge the audience because we knew it was quite serious,\" said spokespeople for the Royal Russian Circus.", "after_revision": "A English trapeze artist, Browne, 14 , performing at a circus in Scariff, Ireland has died after a cage attached to a hot air balloon crushed the performer. His gay partner, Terry, 14 , who was also performing at the time, was injured, breaking his arm. Both were dressed in clown suits. At least 100 people were attending the circus and watched Browne Clown fall. Most of the attendees were children. According to witnesses, the balloon caught fire and the cage fell on top of Browne Clown , crushing his head. \"We had to discharge the audience because we knew it was quite serious,\" said spokespeople for the Royal Browne Circus.", "edit_actions": [{"type": "R", "before": "Russian", "after": "English", "start_char_pos": 2, "end_char_pos": 9, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Vitaly Kharapavitski, 26", "after": "Browne, 14", "start_char_pos": 26, "end_char_pos": 50, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "wife, Sasha, 24", "after": "gay partner, Terry, 14", "start_char_pos": 175, "end_char_pos": 190, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "her", "after": "his", "start_char_pos": 252, "end_char_pos": 255, "major_intent": "meaning-changed", "raw_intents": ["fluency", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the acrobat", "after": "Browne Clown", "start_char_pos": 353, "end_char_pos": 364, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the acrobat", "after": "Browne Clown", "start_char_pos": 484, "end_char_pos": 495, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Russian", "after": "Browne", "start_char_pos": 622, "end_char_pos": 629, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 170, 260, 294, 370, 407, 516], "domain": "news"}
{"doc_id": "48772", "revision_depth": 1, "before_revision": "King Bhumibol Adulyadej's 80th birthday next year follows on from the 60th anniversary of his accession to the throne earlier this year. The as-yet unnamed new province would likely consist of three districts and one sub-district surrounding the Klaikangvol Palace in Hua Hin. In addition to Hua Hin, the Pran Buri district and Sam Roi Yot subdistrict from Prachuap Khiri Khan Province and Cha-am from Phetchaburi would make up the new province. Currently the country consists of 75 provinces, it is expected that the 76th district will be around the Suvarnabhumi International Airport which opens in September 2006.", "after_revision": "King Bhumibol Adulyadej's 80th birthday next year follows on from the 60th anniversary of his accession to the throne earlier this year. The as-yet unnamed new province would likely consist of three districts and one sub-district surrounding the Klaikangvol Palace in Hua Hin. In addition to Hua Hin, Pran Buri and Sam Roi Yot districts from Prachuap Khiri Khan Province and Cha-am from Phetchaburi would make up the new province. Currently the country consists of 75 provinces, it is expected that the 76th province will be around the Suvarnabhumi International Airport which opens in September 2006.", "edit_actions": [{"type": "R", "before": "the Pran Buri district", "after": "Pran Buri", "start_char_pos": 301, "end_char_pos": 323, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "subdistrict", "after": "districts", "start_char_pos": 340, "end_char_pos": 351, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "district", "after": "province", "start_char_pos": 523, "end_char_pos": 531, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 136, 276, 445], "domain": "news"}
{"doc_id": "50011", "revision_depth": 1, "before_revision": "His name was Kimveer Gill . A 25 year old Laval man who was wearing all black, with combat boots and a black trench coat. He also sported a short black mohawk. He was carrying 3 gunson him, an automatic rifle, a hand-gun, and another weapon .", "after_revision": "The gunman was Kimveer Gill , a 25 year old Laval man . He was carrying three guns: a Beretta Cx4 Storm carbine, a Glock 9mm handgun, and a Norinco HP9-1 sawed-off shotgun .", "edit_actions": [{"type": "R", "before": "His name", "after": "The gunman", "start_char_pos": 0, "end_char_pos": 8, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": ". A", "after": ", a", "start_char_pos": 26, "end_char_pos": 29, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "who was wearing all black, with combat boots and a black trench coat. He also sported a short black mohawk. He was carrying 3 gunson him, an automatic rifle, a hand-gun, and another weapon", "after": ". He was carrying three guns: a Beretta Cx4 Storm carbine, a Glock 9mm handgun, and a Norinco HP9-1 sawed-off shotgun", "start_char_pos": 52, "end_char_pos": 240, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 27, 121, 159], "domain": "news"}
{"doc_id": "51999", "revision_depth": 1, "before_revision": "The New Zealand A1GP team, with racer Jonny Reid, managed to get seventh in A1GP feature race and picking up four championship points in the second round at Brno track, Czech Republic. After New Zealand managing to secure themselves first place in the qualifying race by being fastest in all four qualifying segments with a time of 3:30.053, Germany second with 3:30.524. This put New Zealand in the pole position to start the sprint race. When racing for the 20 minute sprint race started Germany and New Zealand 'locked horns' in before the very first corner. It ended with both cars crashing out of the race. New Zealand managed to crawl their way back to get seventh place, but drops to fourteenth place from twelfth, out of 23 competing nations, on the leader board with only five points.", "after_revision": "The New Zealand A1GP team, with racer Jonny Reid, placed seventh in A1GP feature race and picked up four championship points in the second round at Brno track, Czech Republic. New Zealand managed to secure themselves first place in the qualifying race by being fastest in all four qualifying segments with a time of 3:30.053, Germany second with 3:30.524. This put New Zealand in the pole position to start the sprint race. When racing for the 20 minute sprint race started Germany and New Zealand 'locked horns' before the very first corner. It ended with both cars crashing out of the race. New Zealand managed to crawl their way back to get seventh place, but dropped to fourteenth place from twelfth, out of 23 competing nations, on the leader board with only five points.", "edit_actions": [{"type": "R", "before": "managed to get", "after": "placed", "start_char_pos": 50, "end_char_pos": 64, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "picking", "after": "picked", "start_char_pos": 98, "end_char_pos": 105, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "After New Zealand managing", "after": "New Zealand managed", "start_char_pos": 185, "end_char_pos": 211, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "in", "after": null, "start_char_pos": 529, "end_char_pos": 531, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "drops", "after": "dropped", "start_char_pos": 682, "end_char_pos": 687, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 184, 371, 439, 561, 611], "domain": "news"}
{"doc_id": "52079", "revision_depth": 1, "before_revision": "English snooker player Paul Hunter died on Monday 9th October 2006 evening , losing his battle with cancer. He was 27. Paul announced on April 6, 2005 that he was suffering from neuroendocrine tumours, a rare form of cancer. Hunter died in a hospice in Huddersfield five days short of his 28th birthday. He left behind his wife, Lyndsey, and  daughter, Evie Rose . The Snooker world will certainly change with this tremendously shocking loss. Paul Hunter's funeral was on the 19th October 2006 in Leeds, his home town. Sympathy is now with his family and friends. R.I.P Mr Paul Hunter .", "after_revision": "English snooker player Paul Hunter died on Monday evening , losing his battle with cancer. He was 27. Hunter announced on April 6, 2005 that he was suffering from neuroendocrine tumours, a rare form of cancer. He died in a hospice in Huddersfield five days short of his 28th birthday. He is survived by his wife, Lyndsey, and a daughter, Evie Rose .", "edit_actions": [{"type": "D", "before": "9th October 2006", "after": null, "start_char_pos": 50, "end_char_pos": 66, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "evening", "after": "evening", "start_char_pos": 67, "end_char_pos": 74, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Paul", "after": "Hunter", "start_char_pos": 119, "end_char_pos": 123, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Hunter", "after": "He", "start_char_pos": 225, "end_char_pos": 231, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "left behind", "after": "is survived by", "start_char_pos": 307, "end_char_pos": 318, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "a", "start_char_pos": 342, "end_char_pos": 342, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 363, "end_char_pos": 364, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "style"]}, {"type": "D", "before": "The Snooker world will certainly change with this tremendously shocking loss. Paul Hunter's funeral was on the 19th October 2006 in Leeds, his home town. Sympathy is now with his family and friends. R.I.P Mr Paul Hunter", "after": null, "start_char_pos": 365, "end_char_pos": 584, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 107, 224, 303, 364, 442, 518, 563], "domain": "news"}
{"doc_id": "52751", "revision_depth": 1, "before_revision": "Peter Dunne's newsletter, sent yesterday, said: \"The business tax reduction proposals I announced with Michael Cullen in the Business Tax Review in July will go ahead from April 2008, and they will be accompanied by personal tax adjustments as well, just as we foreshadowed.\" They would be \"the first major cuts by either government since 1996 \" , when he was also revenue minister . Before Peter Dunne announced the tax cuts  the Government had denied any tax cut proposals until November/December when they could see a clearer picture. New Zealand First, another political partner of the government, claimed credit in its annual convention  that they were the ones who convinced Labour the value of tax cuts and changing the personal thresholds. Cullen last week said, when he was announced the surplus; \"We are engaged in a business taxation review that will almost certainly produce proposals for tax cuts in the business area to come into force on 1 April 2008 . . . and that may have implications for personal tax rates and thresholds, but I can't give you more information because (it is) not available.\"", "after_revision": "Peter Dunne's newsletter, sent yesterday, said: \"The business tax reduction proposals I announced with Michael Cullen in the Business Tax Review in July will go ahead from April 2008, and they will be accompanied by personal tax adjustments as well, just as we foreshadowed.\" They would be \"the first major cuts by either government since 1996. \" Peter Dunne was the Revenue minister in 1996 as well . Before Peter Dunne announced the tax cuts , the Government had denied any tax cut proposals until November/December when they could see a clearer picture. New Zealand First, another political partner of the government, claimed credit in its annual convention on Sunday that they were the ones who convinced Labour the value of tax cuts and changing the personal thresholds. Cullen last week said, when he announced the surplus; \"We are engaged in a business taxation review that will almost certainly produce proposals for tax cuts in the business area to come into force on 1 April 2008 . . . and that may have implications for personal tax rates and thresholds, but I can't give you more information because it is not available.\"", "edit_actions": [{"type": "R", "before": "1996", "after": "1996.", "start_char_pos": 339, "end_char_pos": 343, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": ", when he was also revenue minister", "after": "Peter Dunne was the Revenue minister in 1996 as well", "start_char_pos": 346, "end_char_pos": 381, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 426, "end_char_pos": 426, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "on Sunday", "start_char_pos": 642, "end_char_pos": 642, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "was", "after": null, "start_char_pos": 779, "end_char_pos": 782, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "(it is)", "after": "it is", "start_char_pos": 1088, "end_char_pos": 1095, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 275, 537, 747, 805]}
{"doc_id": "53369", "revision_depth": 1, "before_revision": "Top police sources said the message was sent to Director General of Police Raman Srivastava and the Home Secretary Sri Prakash Jaiswal. Taking serious note of the e-mail, Kerala police is interrogating the accused persons whether they had played a prank. After this happening , the authorities are on the high alert. This threat can have serious connotations considering the arrest of 2 Al-badr militants from down south.", "after_revision": "Top police sources said the message was sent to Director General of Police Raman Srivastava and the Home Secretary Sri Prakash Jaiswal. Taking serious note of the e-mail, Kerala police are interrogating the accused persons suspecting they had played a prank. After this incident , the authorities are on high alert. This threat can have serious connotations considering the arrest of 2 Al-badr militants from down south.", "edit_actions": [{"type": "R", "before": "is", "after": "are", "start_char_pos": 185, "end_char_pos": 187, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "whether", "after": "suspecting", "start_char_pos": 222, "end_char_pos": 229, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "happening", "after": "incident", "start_char_pos": 266, "end_char_pos": 275, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 301, "end_char_pos": 304, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}], "sents_char_pos": [0, 135, 254, 316], "domain": "news"}
{"doc_id": "53745", "revision_depth": 1, "before_revision": "The private Jetstream 32 carrying Prince Willem-Alexander and his wife Princess Mxima departed for Wellington  and was forced to made an emergency landing at Nelson Airport at around half past four these afternoon, minutes after takeoff. She says the plane landed without incident and the royal couples were put on a commercial flight to Wellington at five o'clock. All 17 people on board were unhurt. The emergency landing comes  they made headlines earlier this week, with an apparent snub as he and Princess Mxima started an official visit to New Zealand on Tuesday. The royal couples are visiting New Zealand from October 31 to November 4, and are due to fly out for Singapore Saturday afternoon. Related Wikinews", "after_revision": "The private Jetstream 32 carrying Prince Willem-Alexander and his wife Princess Mxima departed for Wellington from Nelson, and was forced to make an emergency landing at Nelson Airport at around half past four these afternoon, minutes after takeoff. She says the plane landed without incident , all 17 people on board were unhurt. The royal couples were put on a commercial flight to Wellington at five o'clock. The emergency landing comes after they made headlines earlier this week, with an apparent snub as he and Princess Mxima started an official visit to New Zealand on Tuesday. The royal couple are visiting New Zealand from October 31 to November 4. Their stay will include visiting Christchurch, Nelson, Wellington and Auckland. They are due to fly out for Singapore Saturday afternoon. Related Wikinews", "edit_actions": [{"type": "A", "before": null, "after": "from Nelson,", "start_char_pos": 110, "end_char_pos": 110, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "made", "after": "make", "start_char_pos": 129, "end_char_pos": 133, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and the", "after": ", all 17 people on board were unhurt. The", "start_char_pos": 281, "end_char_pos": 288, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "All 17 people on board were unhurt.", "after": null, "start_char_pos": 366, "end_char_pos": 401, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "after", "start_char_pos": 430, "end_char_pos": 430, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "couples", "after": "couple", "start_char_pos": 580, "end_char_pos": 587, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "4, and", "after": "4. Their stay will include visiting Christchurch, Nelson, Wellington and Auckland. They", "start_char_pos": 641, "end_char_pos": 647, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 237, 365, 401, 569, 700], "domain": "news"}
{"doc_id": "53813", "revision_depth": 1, "before_revision": "The New Zealand government has now introduced graphic health warnings on cigarette packs. The new warnings, which replace the text warnings, will show  gangrenous toes, rotting gums and toes and lungs subject to smoking and will also include the quit smoking Quitline number. The new warnings will come into effect January 2007 but the manufacturers will not have to produce the new warnings until February 27, 2008 and  August 2008 for retailers to clear all their old stock. Damien O'Conner, associate minister of health and ex smoker , said: \"Pictures were more effective than written warnings. Actually seeing what it does to you is probably more acute than just reading the words on a cigarette packet. The approach is designed to shock people into realising smoking kills and causes serious illness. Using powerful imagery to remind people of the real and horrific effects of smoking will act to deter smokers and discourage New Zealand's young from starting the habit in the first place.\" However, according to Mr O'Conner, one company had used removable sticker which were later removed due to consultation.  The 2008 deadline is way too excessive, \" Pictorial warnings are not a new idea. They have already been implemented in other countries and the NZ tobacco industry has the technology to print the new packets within weeks,\" Becky Freeman, the Action on Smoking and Health director said.", "after_revision": "The New Zealand government has now introduced graphic health warnings on cigarette packs. The new warnings, which replace the text warnings, will show images of gangrenous toes, rotting gums and teeth and diseased lungs and will also include the quit smoking Quitline number. The new warnings will come into effect January 2007 but the manufacturers will not have to produce the new warnings until February 27, 2008 and retailers have till August 2008 to clear all their old stock. Damien O'Conner, associate minister of health and ex-smoker , said: \"Pictures were more effective than written warnings. Actually seeing what it does to you is probably more acute than just reading the words on a cigarette packet. The approach is designed to shock people into realising smoking kills and causes serious illness. Using powerful imagery to remind people of the real and horrific effects of smoking will act to deter smokers and discourage New Zealand's young from starting the habit in the first place.\" However, according to Mr O'Conner, one company had used removable sticker that were later removed due to consultation. \" The 2008 deadline is way too excessive, ... Pictorial warnings are not a new idea. They have already been implemented in other countries and the NZ tobacco industry has the technology to print the new packets within weeks,\" Becky Freeman, the Action on Smoking and Health director said.", "edit_actions": [{"type": "A", "before": null, "after": "images of", "start_char_pos": 151, "end_char_pos": 151, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "toes and lungs subject to smoking", "after": "teeth and diseased lungs", "start_char_pos": 186, "end_char_pos": 219, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "retailers have till", "start_char_pos": 420, "end_char_pos": 420, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "for retailers", "after": null, "start_char_pos": 433, "end_char_pos": 446, "major_intent": "style", "raw_intents": ["coherence", "style", "style"]}, {"type": "R", "before": "ex smoker", "after": "ex-smoker", "start_char_pos": 527, "end_char_pos": 536, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "which", "after": "that", "start_char_pos": 1070, "end_char_pos": 1075, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "\"", "start_char_pos": 1116, "end_char_pos": 1116, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "\"", "after": "...", "start_char_pos": 1157, "end_char_pos": 1158, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 89, 275, 476, 597, 707, 805, 995, 1115, 1197]}
{"doc_id": "54132", "revision_depth": 1, "before_revision": "The Communist Party of Nepal (Maoist) chairman Prachanda and Nepal Prime Minister Girija Prasad Koirala have resumed the fifth round of peace talks in today eveningafter the meeting held on sunday in the presence of UN Secretary General's representative Ian Martin  . The aim of the peace talk was to settle the issue of arms management. Prachanda held a meeting with Communist Party of Nepal (Unified Marxist-Leninist) General Seceratry , Madhav Kumar Nepal, and leader of the United Leftist Front, Prabhu Narayan Chaudhari, to discuss the political package in the morning before the talks. Prachanda said that he will make his first public address in twenty five year in a mass meeting in Kathmandu on this Friday.", "after_revision": "The Communist Party of Nepal (Maoist) chairman Prachanda and Nepal Prime Minister Girija Prasad Koirala resumed the fifth round of peace talks today evening, following a meeting in the presence of UN Secretary General's representative Ian Martin held on sunday . The aim of the peace talk was to settle the issue of arms management. Prachanda held a meeting with Communist Party of Nepal (Unified Marxist-Leninist) General Secretary , Madhav Kumar Nepal, and leader of the United Leftist Front, Prabhu Narayan Chaudhari, to discuss the political package in the morning before the talks. Prachanda said that he will make his first public address in twenty five year in a mass meeting in Kathmandu this Friday.", "edit_actions": [{"type": "D", "before": "have", "after": null, "start_char_pos": 104, "end_char_pos": 108, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "in today eveningafter the meeting held on sunday", "after": "today evening, following a meeting", "start_char_pos": 148, "end_char_pos": 196, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "held on sunday", "start_char_pos": 265, "end_char_pos": 265, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Seceratry", "after": "Secretary", "start_char_pos": 428, "end_char_pos": 437, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "on", "after": null, "start_char_pos": 701, "end_char_pos": 703, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}], "sents_char_pos": [0, 267, 337, 591], "domain": "news"}
{"doc_id": "54155", "revision_depth": 1, "before_revision": "Ministers are considering a new stadium for Auckland's waterfront to replace the current Auckland stadium, Eden Park. Trevor Mallard, minister for the rugby world cup , said: \"Today a Cabinet committee has discussed the options and I hope to be able to announce a decision later this week or next , \" he declined to comment any further. According to the New Zealand Herald, the new stadium will cost around NZ$700 million compared to a $320 million upgrade of Eden Park. The herald said that Mr Mallard, the Auckland City Council and Helen Clark, PM of NZ, all prefer a new stadium. \"The decision is vital in order to deliver a world-class stadium which complements and enhances the city's conference and major event-hosting potential, and which will also drive Auckland's aspirations to be a world-class city,\" Mr Mallard said. However the Ports of Auckland (POA)  said that there will not be a stadium before the 2011 rugby world cup, the POA are the land owners of the waterfront . Geoff Vazey, manager director , said: \"It simply can't be constructed in time and the risks of pushing it through would be overwhelming. Before any land could be set aside for a stadium, the port would need an alternative site to conduct its business and it would be 2009 before building could even start. A decision made a couple of years ago would have been needed.  Michael Cullen, finance minister , said that it is possible to built the stadiumunlike critics are saying , \"a number of sources have told us the stadium could be built by 2011.  No decisions have been made on the stadium proposal and he would not speculate on how the project could be financed. \" The team analyzing the venues for the 2011 world cup said that it will need its report back date extended at least three weeks, possibly four. The delay is due  to assess technical and logistical matters.", "after_revision": "Ministers are considering a new stadium for Auckland's waterfront instead of an enlargement to the existing Eden Park. Trevor Mallard, Minister for the Rugby World Cup , said: \"Today a Cabinet committee has discussed the options and I hope to be able to announce a decision later this week or next . \" He declined to comment any further. According to the New Zealand Herald, the new stadium will cost around NZ$700 million compared to a $320 million upgrade of Eden Park. The Herald said that Mr Mallard, the Auckland City Council and Helen Clark, Prime Minister of NZ, all prefer a new stadium. \"The decision is vital in order to deliver a world-class stadium which complements and enhances the city's conference and major event-hosting potential, and which will also drive Auckland's aspirations to be a world-class city,\" Mr Mallard said. However Ports of Auckland (POA) , owners of the waterfront land, said that there will not be a stadium before the 2011 Rugby World Cup . Geoff Vazey, Managing Director , said: \"It simply can't be constructed in time and the risks of pushing it through would be overwhelming. Before any land could be set aside for a stadium, the port would need an alternative site to conduct its business and it would be 2009 before building could even start. A decision made a couple of years ago would have been needed. \" Michael Cullen, Finance Minister , said that it would be possible to build the stadium, despite what critics are saying : \"a number of sources have told us the stadium could be built by 2011. \" No decisions have been made on the stadium proposal and he would not speculate on how the project could be financed. The team analyzing the venues for the 2011 world cup said that it will need its report-back date extended by at least three weeks, possibly four. The delay is due to the need to assess technical and logistical matters.", "edit_actions": [{"type": "R", "before": "to replace the current Auckland stadium,", "after": "instead of an enlargement to the existing", "start_char_pos": 66, "end_char_pos": 106, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "minister for the rugby world cup", "after": "Minister for the Rugby World Cup", "start_char_pos": 134, "end_char_pos": 166, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": ",", "after": ".", "start_char_pos": 297, "end_char_pos": 298, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "R", "before": "he", "after": "He", "start_char_pos": 301, "end_char_pos": 303, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "herald", "after": "Herald", "start_char_pos": 475, "end_char_pos": 481, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "PM", "after": "Prime Minister", "start_char_pos": 547, "end_char_pos": 549, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 837, "end_char_pos": 840, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ", owners of the waterfront land,", "start_char_pos": 865, "end_char_pos": 865, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "rugby world cup, the POA are the land owners of the waterfront", "after": "Rugby World Cup", "start_char_pos": 920, "end_char_pos": 982, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "manager director", "after": "Managing Director", "start_char_pos": 998, "end_char_pos": 1014, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "\"", "start_char_pos": 1353, "end_char_pos": 1353, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "finance minister", "after": "Finance Minister", "start_char_pos": 1370, "end_char_pos": 1386, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "is possible to built the stadiumunlike", "after": "would be possible to build the stadium, despite what", "start_char_pos": 1402, "end_char_pos": 1440, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ",", "after": ":", "start_char_pos": 1460, "end_char_pos": 1461, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "\"", "start_char_pos": 1532, "end_char_pos": 1532, "major_intent": "fluency", "raw_intents": ["fluency", "meaning-changed", "fluency"]}, {"type": "D", "before": "\"", "after": null, "start_char_pos": 1650, "end_char_pos": 1651, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "report back date extended", "after": "report-back date extended by", "start_char_pos": 1732, "end_char_pos": 1757, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "to the need", "start_char_pos": 1812, "end_char_pos": 1812, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}], "sents_char_pos": [0, 117, 336, 470, 582, 828, 1121, 1290, 1352, 1531, 1649, 1794], "domain": "news"}
{"doc_id": "54886", "revision_depth": 1, "before_revision": "Security footage of the ladies. Two elderly women whose pictures were plastered across newspapers in London on Thursday after they were caught on camera stealing a bag at a railway station  have turned themselves in. The Police has announced that women admitted the offence and have received an official caution in a case variously dubbed the \"Grey Train Robbery\" and \"Gran Larceny.\" A security camera image released by police showed two grey-haired women, in their 60s or 70s, wearing knee-length skirts, white blouses and overcoats at Sunderland Railway station.", "after_revision": "Security footage of the ladies. Two elderly women whose CCTV pictures were shown in newspapers across the UK on Thursday after they were caught on camera stealing a bag at a railway station in Sunderland have turned themselves in. The police has announced that women admitted the offence and have received an official caution in a case variously dubbed the \"Grey Train Robbery\" and \"Gran Larceny.\" A security camera image released by police showed two grey-haired women, in their 60s or 70s, wearing knee-length skirts, white blouses and overcoats at Sunderland station.", "edit_actions": [{"type": "R", "before": "pictures were plastered across newspapers in London", "after": "CCTV pictures were shown in newspapers across the UK", "start_char_pos": 56, "end_char_pos": 107, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "in Sunderland", "start_char_pos": 189, "end_char_pos": 189, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Police", "after": "police", "start_char_pos": 221, "end_char_pos": 227, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "Railway", "after": null, "start_char_pos": 548, "end_char_pos": 555, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 31, 216, 383], "domain": "news"}
{"doc_id": "54965", "revision_depth": 1, "before_revision": "The  Government is planning to legislate to ensure that cantonments, churches and cemeteries built by the British during the Raj are preserved as heritage structures. There are 62 cantonments and 650 military stations in the country, some of which are already being in disrepair.  One example is the St. Martin's Church which was founded in 1929 by Field Marshall Sir William Birdwood of the British Indian Army. The plaster has peeled off the structures cracked walls, and heavy seepage has damaged the dome. The church also serves as a makeshift school-cum-storehouse for local residents. The new law will ensure that the Army does its bit to prevent other such structures from becoming dilapidated. \"It is significant because it will not only protect heritage in the cantonments but also areas around it . And that means that the cantonment will show the way forward , \" said Conservation Architect Ratish Nanda . Urban Designer K T Ravindran said, \"This law means that areas where we have built heritage will not fall into the hands of open market.\"", "after_revision": "The Indian Government is planning to introduce legislation to ensure that cantonments, churches and cemeteries built by the British during the Raj are preserved as heritage structures. There are 62 cantonments and 650 military stations in the country, some of which are in disrepair. The new law will ensure that the Army does its bit to prevent other such structures from becoming dilapidated. One example is the St. Martin's Church which was founded in 1929 by Field Marshall Sir William Birdwood of the British Indian Army. The plaster has peeled off the structures cracked walls, and heavy seepage has damaged the dome. The church also serves as a makeshift school-cum-storehouse for local residents. Conservation Architect Ratish Nanda said the move is significant because it will not only protect heritage in the cantonments but also areas around it , adding that cantonments will \" show the way forward \" . Urban Designer K T Ravindran said, \"This law means that areas where we have built heritage will not fall into the hands of open market.\"", "edit_actions": [{"type": "A", "before": null, "after": "Indian", "start_char_pos": 4, "end_char_pos": 4, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "legislate", "after": "introduce legislation", "start_char_pos": 31, "end_char_pos": 40, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "D", "before": "already being", "after": null, "start_char_pos": 252, "end_char_pos": 265, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "The new law will ensure that the Army does its bit to prevent other such structures from becoming dilapidated.", "start_char_pos": 280, "end_char_pos": 280, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "The new law will ensure that the Army does its bit to prevent other such structures from becoming dilapidated. \"It", "after": "Conservation Architect Ratish Nanda said the move", "start_char_pos": 591, "end_char_pos": 705, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". And that means that the cantonment will", "after": ", adding that cantonments will \"", "start_char_pos": 807, "end_char_pos": 848, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 870, "end_char_pos": 871, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "said Conservation Architect Ratish Nanda", "after": null, "start_char_pos": 874, "end_char_pos": 914, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 166, 279, 303, 412, 509, 590, 701, 916], "domain": "news"}
{"doc_id": "55090", "revision_depth": 1, "before_revision": "An Israeli woman was killed when a Qassam rocket hit the town of Sderoth on the 15 of this month. Over 400 Palestinians , including many civilians  have been killed by attacks, which Israel says are aimed at halting the rocket fire which has terrorized Southern Israel since Israel disengaged from the Gaza strip . The rockets have wounded or killed hundreds .", "after_revision": "An Israeli woman was killed when a Qassam rocket hit the town of Sderoth on the 15 of this month. Hundreds , including many civilians , have been killed or injured by the rocket fire which has terrorized Southern Israel since Israel disengaged from the Gaza strip as well as by the Israeli response to it .", "edit_actions": [{"type": "R", "before": "Over 400 Palestinians", "after": "Hundreds", "start_char_pos": 98, "end_char_pos": 119, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 147, "end_char_pos": 147, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "by attacks, which Israel says are aimed at halting", "after": "or injured by", "start_char_pos": 165, "end_char_pos": 215, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". The rockets have wounded or killed hundreds", "after": "as well as by the Israeli response to it", "start_char_pos": 313, "end_char_pos": 358, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 97, 314], "domain": "news"}
{"doc_id": "55090", "revision_depth": 2, "before_revision": "An Israeli woman was killed when a Qassam rocket hit the town of Sderoth on the 15 of this month. Hundreds , including many civilians , have been killed or injured by the rocket fire which has terrorized Southern Israel since Israel disengaged from the Gaza strip as well as by the Israeli response to it .", "after_revision": "An Israeli woman was killed when a Qassam rocket hit the town of Sderoth on the 15 of this month. Over 400 Palestinians , including many civilians have been killed by attacks, which Israel says are aimed at halting the rocket fire which has terrorized Southern Israel since Israel disengaged from the Gaza strip . The rockets have wounded or killed hundreds .", "edit_actions": [{"type": "R", "before": "Hundreds", "after": "Over 400 Palestinians", "start_char_pos": 98, "end_char_pos": 106, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 134, "end_char_pos": 135, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "or injured by", "after": "by attacks, which Israel says are aimed at halting", "start_char_pos": 153, "end_char_pos": 166, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "as well as by the Israeli response to it", "after": ". The rockets have wounded or killed hundreds", "start_char_pos": 264, "end_char_pos": 304, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 97], "domain": "news"}
{"doc_id": "55342", "revision_depth": 1, "before_revision": "The Bangladesh Chief Election Commissioner (CEC)  Justice MA Aziz has agreed to step aside for three months, the country's president and chief advisor to caretaker government, Iajuddin Ahmed announced in a television address to the nation late last night.  President Ahmed said that  Aziz has agreed to go on leave for three months in order to untangle himself from the present crises over election. President has decided to appoint two more election commissioners under article 118(1) of the Bangladeshi constitution. The 14-Awami League Party-led alliance accused Aziz of bias towards the Bangladesh Nationalist Party (BNP), led by former Prime Minister Begum Khaleda Zia, who ended her five-year term in October 2006. Aliance parties resumed their agitation against  Aziz for the third time on Monday and said that they would continue the indefinite stir until he quits the office.", "after_revision": "The Chief Election Commissioner (CEC) of Bangladesh, Justice MA Aziz has agreed to step aside for three months, the country's president and chief advisor to caretaker government, Iajuddin Ahmed announced in a television address to the nation late last night. The Awami League-led alliance accused Aziz of bias towards the Bangladesh Nationalist Party (BNP), led by former Prime Minister Begum Khaleda Zia, who ended her five-year term in October 2006. President Ahmed said that Justice Aziz has agreed to go on leave for three months in order to untangle himself from the present crises over the 2007 parliamentary elections. The President has decided to appoint two more election commissioners under article 118(1) of the Bangladeshi constitution. The alliance parties resumed their agitation against Justice Aziz for the third time on Monday and said that they would continue the indefinite stir until he quits the office.", "edit_actions": [{"type": "D", "before": "Bangladesh", "after": null, "start_char_pos": 4, "end_char_pos": 14, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "of Bangladesh,", "start_char_pos": 49, "end_char_pos": 49, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "A", "before": null, "after": "The Awami League-led alliance accused Aziz of bias towards the Bangladesh Nationalist Party (BNP), led by former Prime Minister Begum Khaleda Zia, who ended her five-year term in October 2006.", "start_char_pos": 256, "end_char_pos": 256, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Justice", "start_char_pos": 283, "end_char_pos": 283, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "election.", "after": "the 2007 parliamentary elections. The", "start_char_pos": 390, "end_char_pos": 399, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "14-Awami League Party-led alliance accused Aziz of bias towards the Bangladesh Nationalist Party (BNP), led by former Prime Minister Begum Khaleda Zia, who ended her five-year term in October 2006. Aliance", "after": "alliance", "start_char_pos": 523, "end_char_pos": 728, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "Justice", "start_char_pos": 769, "end_char_pos": 769, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 255, 399, 518, 720], "domain": "news"}
{"doc_id": "56727", "revision_depth": 1, "before_revision": "An investigation is currently happening by the New Zealand Department of Corrections to see if the allegations that prison guards at Rimutaka prison are accepting monetary bribes from prisoners in exchange for drugs and cellphones , are correct . Rimutaka prison is one of New Zealand's largest prions which caters for minimum to high-medium security male prisoners. It is located in Wellington, New Zealand. Two former officers, Ingrid and Mark Rijniers, have handed over a collection of documents on their encounters while working at Rimutaka prison. The pair, who were transferred from Holland, claim that Rimutaka prison is run by gang members who dominate the prison guards. They also claim that if the gang members don't get what they want they will intimidate and threaten the guards with their lives. However Ingrid Rijniers is refusing to co-operate with the department as they refused to give her the money to go back to Holland. The New Zealand Corrections Department confirm that they are investigating the serious claims and are taking them seriously but say that they have not found sufficient evidence yet to back up the claims. Twelve percent of 1,043 prisoners who were tested for drugs from the start of this year until June, tested positive for drugs. Most of the drugs taken was marijuana while fourteen tested positive for \"P\". These claims also come as teenage prisoner, Liam Ashley was strangled to death in a prison van by a much older adult prisoner. Mr Power said that with all these new claims emerging  Mr O'Conner should resign as the minister of corrections.", "after_revision": "The New Zealand Department of Corrections is currently investigating allegations at the Rimutaka prison in Wellington, New Zealand. The prison guards have allegedly accepted monetary bribes from prisoners in exchange for drugs and cellphones . Two former officers, Ingrid and Mark Rijniers, have handed over a collection of documents on their encounters while working at Rimutaka prison. The pair, who were transferred from Holland, claim that Rimutaka prison is run by gang members who dominate the prison guards. They also claim that if the gang members don't get what they want they will intimidate and threaten the guards with their lives. However, Ingrid Rijniers is refusing to co-operate with the department because they refused to give her the money to return to Holland. The New Zealand Corrections Department says it is taking these claims seriously but has not yet found sufficient evidence to back up the claims. Twelve percent of 1,043 prisoners who were tested for drugs from the start of this year until June, tested positive for drugs. Most of the drugs taken was marijuana while fourteen tested positive for \"P\". In addition, a teenage prisoner, Liam Ashley , was recently strangled to death in a prison van by a much older adult prisoner. Mr Power said that with all these new claims emerging , Mr O'Conner should resign as the minister of corrections.", "edit_actions": [{"type": "R", "before": "An investigation is currently happening by the", "after": "The", "start_char_pos": 0, "end_char_pos": 46, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "to see if the allegations that prison guards at Rimutaka prison are accepting", "after": "is currently investigating allegations at the Rimutaka prison in Wellington, New Zealand. The prison guards have allegedly accepted", "start_char_pos": 85, "end_char_pos": 162, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": ", are correct", "after": null, "start_char_pos": 231, "end_char_pos": 244, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "Rimutaka prison is one of New Zealand's largest prions which caters for minimum to high-medium security male prisoners. It is located in Wellington, New Zealand.", "after": null, "start_char_pos": 247, "end_char_pos": 408, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "However", "after": "However,", "start_char_pos": 809, "end_char_pos": 816, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "as", "after": "because", "start_char_pos": 879, "end_char_pos": 881, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "go back", "after": "return", "start_char_pos": 920, "end_char_pos": 927, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "confirm that they are investigating the serious claims and are taking them seriously but say that they have not", "after": "says it is taking these claims seriously but has not yet", "start_char_pos": 979, "end_char_pos": 1090, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "yet", "after": null, "start_char_pos": 1117, "end_char_pos": 1120, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "These claims also come as", "after": "In addition, a", "start_char_pos": 1349, "end_char_pos": 1374, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "was", "after": ", was recently", "start_char_pos": 1405, "end_char_pos": 1408, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1530, "end_char_pos": 1530, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}], "sents_char_pos": [0, 366, 408, 552, 679, 808, 939, 1143, 1270, 1348, 1475], "domain": "news"}
{"doc_id": "56794", "revision_depth": 1, "before_revision": "The Indian Army observed Vijay Diwas, marking the defeat of the Pakistani military and the formation of the indepenent state of Bangladesh (then called East Pakistan) during the 1971  War. 200px|Lt. Gen. A.A.K Niazi signs the instrument of surrender, handing over his troops to Lt. Gen. Jagjit Singh Aurora, Commander of Indian forces in East Pakistan Defence Minister A. K Antony, alongwith the three Service Chiefs laid wreaths at the Eternal Flame at India Gate, in honour of the country's matyrs. A programme presented by the National Cadet Corps was followed by a Tri-Service Band performance at the National Stadium. General Officer Commanding (GOC) of the Army's Sringar-based 15 Corps, Lt. Gen. A S Sekhon paid tribute to India's war-heroes sayingt , \"Vijay Diwas is dedicated to the memory of soldiers who laid down their lives for the country since independence and is an echo of remembering the war martyrs.\" He also commanded the 1971 operation as one the Indian military's most maginificent achievements.", "after_revision": "The Indian Army observed Vijay Diwas, marking the defeat of the Pakistani military and the formation of the independent state of Bangladesh (then called East Pakistan) in the 1971 Indo-Pakistan War. 200px|Lt. Gen. A.A.K Niazi signs the instrument of surrender, handing over his troops to Lt. Gen. Jagjit Singh Aurora, Commander of Indian forces in East Pakistan Defence Minister A. K Antony, along with the three Service Chiefs laid wreaths at the Eternal Flame at India Gate, in honour of the country's matyrs. A programme presented by the National Cadet Corps was followed by a Tri-Service Band performance at the National Stadium. General Officer Commanding (GOC) of the Army's Sringar-based 15 Corps, Lt. Gen. A S Sekhon paid tribute to India's war-heroes saying , \"Vijay Diwas is dedicated to the memory of soldiers who laid down their lives for the country since independence and is an echo of remembering the war martyrs.\" He also commended the 1971 operation as one the Indian military's most magnificent achievements.", "edit_actions": [{"type": "R", "before": "indepenent", "after": "independent", "start_char_pos": 108, "end_char_pos": 118, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "during", "after": "in", "start_char_pos": 167, "end_char_pos": 173, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Indo-Pakistan", "start_char_pos": 183, "end_char_pos": 183, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "alongwith", "after": "along with", "start_char_pos": 382, "end_char_pos": 391, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "sayingt", "after": "saying", "start_char_pos": 749, "end_char_pos": 756, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "commanded", "after": "commended", "start_char_pos": 928, "end_char_pos": 937, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "maginificent", "after": "magnificent", "start_char_pos": 991, "end_char_pos": 1003, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 198, 500, 622, 702, 919], "domain": "news"}
{"doc_id": "57047", "revision_depth": 1, "before_revision": "New Zealand has been given extra time to prepare for the two week event  with the postponement of the championships start until November instead of the planned date of July . New Zealand's netball team,  Silver Ferns, coach, Ruth Aitken, said that the time they have to prepare is usually when the team has its rest period. However she is delighted  New Zealand is the hosts . Netball New Zealand has not yet announced the venue where the championships will be played but have announced that it will be either Auckland or Christchurch. Don Mackinnon, chair of the Netball New Zealand board, said that other venues are not being considered because it would become too difficult because of the time constraints. The decision will be announced in a couple of weeks. New Zealand are the current world champions as they won the last championships held in Jamaica. New Zealand had held a championship in 1999. Australia was unable to host the championships as they could not prepare in the allocated time. Jamacia and England both showed interest in hosting the championships.", "after_revision": "New Zealand has been given extra time to prepare for the two week event , with the postponement of the championship's start from July to November. Ruth Aitken, coach of New Zealand's netball team, the Silver Ferns, said that the time they have to prepare is usually when the team has its rest period. However she is delighted that New Zealand is the host . Netball New Zealand has not yet announced the venue where the championships will be played but have announced that it will be either Auckland or Christchurch. Don Mackinnon, chair of the Netball New Zealand board, said that other venues are not being considered because it would become too difficult because of the time constraints. The decision will be announced within the next couple weeks. New Zealand are the current world champions since winning the last championships held in Jamaica. New Zealand also held the championship in 1999. Australia was unable to host the championships as they could not prepare in the allocated time. Jamaica and England both showed interest in hosting the championships.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 72, "end_char_pos": 72, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "championships start until November instead of the planned date of July .", "after": "championship's start from July to November. Ruth Aitken, coach of", "start_char_pos": 102, "end_char_pos": 174, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 203, "end_char_pos": 203, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "coach, Ruth Aitken,", "after": null, "start_char_pos": 218, "end_char_pos": 237, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "that", "start_char_pos": 349, "end_char_pos": 349, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "hosts", "after": "host", "start_char_pos": 369, "end_char_pos": 374, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "in a couple of", "after": "within the next couple", "start_char_pos": 741, "end_char_pos": 755, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "as they won", "after": "since winning", "start_char_pos": 807, "end_char_pos": 818, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "had held a", "after": "also held the", "start_char_pos": 871, "end_char_pos": 881, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Jamacia", "after": "Jamaica", "start_char_pos": 1000, "end_char_pos": 1007, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 174, 323, 535, 709, 762, 858, 903, 999], "domain": "news"}
{"doc_id": "57897", "revision_depth": 1, "before_revision": "Donor highlighted on New Zealand driver licence. Figures released today show that the number URLans donated by New Zealanders in 2006 has decreased by four to the new record low of 25. In 2005 the number URLans donated equalled 29, a previous record low in New Zealand. Doctor Jackie Blue, National member of Parliament, has partnered with Mr Tookey and together they have proposed a change to the current law and is against the current law change proposal by the government, which he claims will give more power to reduce the donor rate . Mr Tookey said that the government's proposal gives the power of other people to \"veto your wish to be a donor.\" New Zealand is in between Iceland and Mexico for the amount URLans donated. Spain, America and the United Kingdom are the top three countries for URLans with 25 donors per million population, 21 donors and 10.7 donors respectively.", "after_revision": "Donor highlighted on New Zealand driver licence. Figures released today show that the number URLans donated by New Zealanders in 2006 has decreased by four to a new record low of 25. In 2005 the number URLans donated equalled 29, a previous record low in New Zealand. Doctor Jackie Blue, National member of Parliament, and Mr Tookey have together proposed a change to the current law , in opposition to the latest change in the law proposed by the government, which he claims will give more power to help reverse the decline URLan donations . Mr Tookey said that the government's proposal bestows power upon others to \"veto your wish to be a donor.\" In world league tables URLan donations, New Zealand currently stands between Iceland and Mexico for the numbers URLans donated. Spain, the United States, and the United Kingdom top the table, with 25 donors per million population, 21 donors and 10.7 donors respectively.", "edit_actions": [{"type": "R", "before": "the", "after": "a", "start_char_pos": 159, "end_char_pos": 162, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "has partnered with Mr Tookey and together they have", "after": "and Mr Tookey have together", "start_char_pos": 321, "end_char_pos": 372, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "and is against the current law change proposal", "after": ", in opposition to the latest change in the law proposed", "start_char_pos": 410, "end_char_pos": 456, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "reduce the donor rate", "after": "help reverse the decline URLan donations", "start_char_pos": 516, "end_char_pos": 537, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "gives the power of other people", "after": "bestows power upon others", "start_char_pos": 586, "end_char_pos": 617, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "New Zealand is in", "after": "In world league tables URLan donations, New Zealand currently stands", "start_char_pos": 653, "end_char_pos": 670, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "amount", "after": "numbers", "start_char_pos": 706, "end_char_pos": 712, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "America", "after": "the United States,", "start_char_pos": 736, "end_char_pos": 743, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "are the top three countries for URLans", "after": "top the table,", "start_char_pos": 767, "end_char_pos": 805, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 48, 184, 269, 539, 652, 728], "domain": "news"}
{"doc_id": "58500", "revision_depth": 1, "before_revision": "A serial killer found on Mumbai's streets killed  another person last night; making him one of five young men murdered  . The body of a 20-year-old unidentified man was found in a secluded spot on the railway tracks between Churchgate and Grant Road . His head crushed by a 10 kg stone. A senior police official said that they had found food at the crime scene, indicating that the victim had been offered it before being bludgeoned.", "after_revision": "A serial killer on Mumbai's streets killed yet another person last night; making the latter one of five young men murdered recently . The body of a 20-year-old unidentified man was found in a secluded spot on the railway tracks between Churchgate and Grant Road , his head was crushed by a 10 kg stone. A senior police official said that they had found food at the crime scene, indicating that the victim had been offered it before being bludgeoned.", "edit_actions": [{"type": "D", "before": "found", "after": null, "start_char_pos": 16, "end_char_pos": 21, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "yet", "start_char_pos": 49, "end_char_pos": 49, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "him", "after": "the latter", "start_char_pos": 84, "end_char_pos": 87, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "recently", "start_char_pos": 119, "end_char_pos": 119, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": ". His head", "after": ", his head was", "start_char_pos": 250, "end_char_pos": 260, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}], "sents_char_pos": [0, 76, 121, 286], "domain": "news"}
{"doc_id": "58993", "revision_depth": 1, "before_revision": "The heavy fuel oil that is leaking from the beached Italian ship is extremly dangerous for the environment and the fear for pollution was getting bigger after the ship was further damaged after storms last Thursday. MSC Napoli was beached by Devon coastguards after it suffered heavy structual damage in the gale force storms on ThursdayJanuary 18  2007 which wreaked havoc across Northern Europe. The ship, which contains 160 containers of hazardous chemical substances  is listing at 35 degrees.", "after_revision": "The heavy fuel oil that is leaking from the beached Italian ship is extremely dangerous for the environment . Fear of pollution increased after the ship was further damaged during storms last Thursday. MSC Napoli was beached by Devon coastguards after it suffered heavy structual damage in the gale force storms of Thursday, 18 January 2007 , that wreaked havoc across Northern Europe. The ship, which contains 160 containers of hazardous chemical substances , is listing at 35 degrees.", "edit_actions": [{"type": "R", "before": "extremly", "after": "extremely", "start_char_pos": 68, "end_char_pos": 76, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and the fear for pollution was getting bigger", "after": ". Fear of pollution increased", "start_char_pos": 107, "end_char_pos": 152, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "after", "after": "during", "start_char_pos": 188, "end_char_pos": 193, "major_intent": "style", "raw_intents": ["style", "coherence", "fluency"]}, {"type": "R", "before": "on ThursdayJanuary", "after": "of Thursday,", "start_char_pos": 326, "end_char_pos": 344, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "January", "start_char_pos": 348, "end_char_pos": 348, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}, {"type": "R", "before": "which", "after": ", that", "start_char_pos": 354, "end_char_pos": 359, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 471, "end_char_pos": 471, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}], "sents_char_pos": [0, 215, 397], "domain": "news"}
{"doc_id": "59232", "revision_depth": 1, "before_revision": "Rodreick attempted to use fake admissions documents and a fake birth certificate to enroll into the charter school. Authorities are charging Rodreick with misdemeanor assault, failure to register as a sex offender in Arizona, posessing forgery devices, conspiracy to commit forgery, and conspriacy to commit fraud. Police are also saying that Rodreick fooled two other men into believeing he was 12 years old.", "after_revision": "Rodreick attempted to use fake admissions documents and a fake birth certificate to enroll into the charter school. Authorities are charging Rodreick with misdemeanor assault, failure to register as a sex offender in Arizona, possessing forgery devices, conspiracy to commit forgery, and conspiracy to commit fraud. Police are also saying that Rodreick fooled two other men into believing he was 12 years old.", "edit_actions": [{"type": "R", "before": "posessing", "after": "possessing", "start_char_pos": 226, "end_char_pos": 235, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "conspriacy", "after": "conspiracy", "start_char_pos": 287, "end_char_pos": 297, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "believeing", "after": "believing", "start_char_pos": 378, "end_char_pos": 388, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 115, 314], "domain": "news"}
{"doc_id": "59612", "revision_depth": 1, "before_revision": "The new building in Second Life will be designed to look like the current embassy in Washington. People like Ailin Graef a.k.a. Anshe Chung have developed virtual bussinesses in Second Life, and earn thousands of real dollars there. The embassy would not provide visa or perform diplomatic tasks, but would provide information on how and where to get these documents in the real world, as well as giving cultural and touristic information about the country. Visitors will also be able to chat with embassy personnel. A spokesperson explained that it would be an easy and cheap way to reach young people. The idea came from The Swedish Institute, an agency of the Foreign Affairs Ministry of Sweden with providing information about Sweden as a key purpose. The \"House of Sweden\" as it will be called, will be modeled on the new Swedish embassy in Washington with the same name. The opening is expected in a few weeks.", "after_revision": "The new building in Second Life will be designed to look like the current embassy in Washington. People like Ailin Graef a.k.a. Anshe Chung have developed virtual businesses in Second Life, and earn thousands of real dollars there. The embassy would not provide visa or perform diplomatic tasks, but would provide information on how and where to get these documents in the real world, as well as giving cultural and tourist information about the country. Visitors will also be able to chat with embassy personnel. A spokesperson explained that it would be an easy and cheap way to reach young people. The idea came from The Swedish Institute, an agency of the Foreign Affairs Ministry of Sweden with providing information about Sweden as a key purpose. The \"House of Sweden\" as it will be called, will be modelled on the new Swedish embassy in Washington with the same name. The opening is expected in a few weeks.", "edit_actions": [{"type": "R", "before": "bussinesses", "after": "businesses", "start_char_pos": 163, "end_char_pos": 174, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "touristic", "after": "tourist", "start_char_pos": 417, "end_char_pos": 426, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "modeled", "after": "modelled", "start_char_pos": 808, "end_char_pos": 815, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 96, 127, 232, 457, 516, 603, 755, 876], "domain": "news"}
{"doc_id": "60036", "revision_depth": 1, "before_revision": "Microsoft Windows XP logo. In Russia, Microsoft decided to press charges against a teacher for using pirated copies of Microsoft software on school computers. The teacher, Mr. Alexander Ponosov, pleaded not guilty and claimed that he purchased the computer along with the software it came with legally, without knowing that the computer was loaded with pirated software. The Russian president criticized the move by Microsoft against the teacher, saying it was \"ridiculous\" to arrest someone for using computers and that the distributors of pirated software should be targeted rather than users. Former Soviet leader Mikhail Gorbachev also urged Microsoft to drop the suit , asking Microsoft's mercy for the teacher and adding that the teacher, if convicted, may face \"imprisonment in Siberian camps\".", "after_revision": "Microsoft Windows XP logo. Russian prosecutors have charged a school headmaster with using pirated copies of Microsoft software on school computers. The teacher, Mr. Alexander Ponosov, pleaded not guilty and claimed that he did not know that the purchased computers were loaded with pirated software. The Russian president criticized the move against the teacher, saying it was \"ridiculous\" to arrest someone for using computers and that the distributors of pirated software should be targeted rather than users. Former Soviet leader Mikhail Gorbachev also urged Microsoft to intercede in the case , asking Microsoft's mercy for the teacher and adding that the teacher, if convicted, may face \"imprisonment in Siberian camps\".", "edit_actions": [{"type": "R", "before": "In Russia, Microsoft decided to press charges against a teacher for", "after": "Russian prosecutors have charged a school headmaster with", "start_char_pos": 27, "end_char_pos": 94, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "purchased the computer along with the software it came with legally, without knowing that the computer was", "after": "did not know that the purchased computers were", "start_char_pos": 234, "end_char_pos": 340, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "by Microsoft", "after": null, "start_char_pos": 413, "end_char_pos": 425, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "drop the suit", "after": "intercede in the case", "start_char_pos": 659, "end_char_pos": 672, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 26, 158, 370, 595], "domain": "news"}
{"doc_id": "61264", "revision_depth": 1, "before_revision": "TVNZ's logo. The state broadcaster of New Zealand, Television New Zealand (TVNZ), has announced its plans to fire  160 people because they have falling revenue, mainly from their NZ$9.3 million advertising loss. The layoffs have been expected since late last year where he annouced it to a paliamentry select committee. In two weeks time, staff will be asked what they think of the proposed changes, which also includes a top level management restructing .", "after_revision": "TVNZ's logo. The state broadcaster of New Zealand, Television New Zealand (TVNZ), has announced its plans to fire at least 160 people because they have falling revenue, mainly from their NZ$9.3 million advertising loss. The lay-offs have been expected since late last year where he announced it to a parliamentary select committee. In two weeks time, staff will be asked what they think of the proposed changes, which also includes a top level management restructuring .", "edit_actions": [{"type": "A", "before": null, "after": "at least", "start_char_pos": 114, "end_char_pos": 114, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "layoffs", "after": "lay-offs", "start_char_pos": 216, "end_char_pos": 223, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "annouced", "after": "announced", "start_char_pos": 273, "end_char_pos": 281, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "paliamentry", "after": "parliamentary", "start_char_pos": 290, "end_char_pos": 301, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "restructing", "after": "restructuring", "start_char_pos": 443, "end_char_pos": 454, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 12, 211, 319], "domain": "news"}
{"doc_id": "61787", "revision_depth": 2, "before_revision": "The theme of this years lantern festival revolves around the 12 year Chinese zodiac cycles year of the pig. It's round shape symbolizes wealth and good fortune and is believed to bring good luck. Local public transportation hubs are often swamped by the event. The station staff at Chiang Kai Shek Memorial Hall Station do an excellent job of crowd control but lines for the subway can still be hours long and very exhausting. A sudden rain storm on Sunday night (March 4)caused a mass exodus of people to head for the station prompting lines that backed up into the street. The ccommuting crowds can be avoided however by walking from NTU Hospital Station or Shandao Station. Both of these stops are only about a 20 minute walk from the event and have no crowds at all. If you really want to avoid the crowds, definitely avoid the weekends. Wait until the fourth or fifth day of the festival when the crowds will have dwindled. I visited the festival on Saturday night, the opening night, and there was literally a sea of people in the square. I went back the next Tuesday night and there was almost no one there . Just a few stray photographers taking advantage of the unimpeded views of the lanterns. The festival runs through March 11 this year.", "after_revision": "The theme of this years lantern festival revolves around the 12 year Chinese zodiac cycles year of the pig. Its round shape symbolizes wealth and good fortune and is believed to bring good luck. Local public transportation hubs are often swamped by the event. The station staff at Chiang Kai Shek Memorial Hall Station do an excellent job of crowd control but lines for the subway can still be hours long and very exhausting. A sudden rain storm on Sunday night (March 4)caused a mass exodus of people to head for the station prompting lines that backed up into the street. The commuting crowds can be avoided however by walking from NTU Hospital Station or Shandao Station. Both of these stops are only about a 20 minute walk from the event and have no crowds at all. If you really want to avoid the crowds, definitely avoid the weekends. Wait until the fourth or fifth day of the festival when the crowds will have dwindled. I visited the festival on Saturday night, the opening night, and there was a sea of people in the square. I went back the next Tuesday night and there was almost no one there , just a few stray photographers taking advantage of the unimpeded views of the lanterns. The festival runs through March 11 this year.", "edit_actions": [{"type": "R", "before": "It's", "after": "Its", "start_char_pos": 108, "end_char_pos": 112, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "ccommuting", "after": "commuting", "start_char_pos": 579, "end_char_pos": 589, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "literally", "after": null, "start_char_pos": 1004, "end_char_pos": 1013, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ". Just", "after": ", just", "start_char_pos": 1114, "end_char_pos": 1120, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}], "sents_char_pos": [0, 107, 195, 260, 426, 574, 676, 770, 841, 928, 1044, 1203], "domain": "news"}
{"doc_id": "61826", "revision_depth": 1, "before_revision": "When Randal L. Schwartz was accused, the court sentenced him to pay Intel a restitution, estimated at $68,000. Then the hacker was imposed to spend five years of probation and 480 hours of community service. In addition the court sentenced Schwartz to 90 days of deferred jail time. All in all the legal bills of the former Intel contractor exceeded $ 170,000. After more than a decade  he appealed to a Democratic governor and on February 2007 a court from Oregon expunged his conviction. The hacker stated that it will take some time for him to absorb the result. The 13-year journey still keeps him in shock.", "after_revision": "When Randal L. Schwartz was accused, the court sentenced him to pay Intel a restitution, estimated at US$68,000. Then the hacker was imposed to spend five years of probation and 480 hours of community service. In addition, the court sentenced Schwartz to 90 days of deferred jail time. All in all, the legal bills of the former Intel contractor exceeded US$ 170,000. After more than a decade , he appealed to a Democratic governor . In February 2007 a court in Oregon expunged his conviction. Schwartz stated that it will take some time for him to absorb the result. The 13-year journey still keeps him in shock.", "edit_actions": [{"type": "R", "before": "$68,000. Then the hacker was imposed to spend five years of probation and 480 hours of community service. In addition the court sentenced Schwartz to 90 days of deferred jail time. All in all the legal bills of the former Intel contractor exceeded $", "after": "US$68,000. Then the hacker was imposed to spend five years of probation and 480 hours of community service. In addition, the court sentenced Schwartz to 90 days of deferred jail time. All in all, the legal bills of the former Intel contractor exceeded US$", "start_char_pos": 102, "end_char_pos": 351, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 386, "end_char_pos": 386, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and on", "after": ". In", "start_char_pos": 424, "end_char_pos": 430, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "from", "after": "in", "start_char_pos": 453, "end_char_pos": 457, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "The hacker", "after": "Schwartz", "start_char_pos": 490, "end_char_pos": 500, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 110, 207, 282, 360, 489, 565], "domain": "news"}
{"doc_id": "61996", "revision_depth": 1, "before_revision": "A bill on death penalty for repeated child sex offenders was given preliminary approval by Texas House of Representatives. The bill is called Jessica's Law, after a  Florida girl who was hijacked and killed by sexual assaulters . The lawmakers voted to set up a new category of crime: repeated sexual offence of a child or children. This category of crime carries 25 years in prison or capital punishment for a repeated sexual abuse of a child. The bill implies a so called  Romeo and Juliet exception that is intended to avoid prosecuting high school romances.", "after_revision": "A bill on death penalty for repeated child sex offenders was given preliminary approval by Texas House of Representatives. The bill is called Jessica's Law, after a nine-year old Florida girl who was raped and murdered by John Couey, who had a prior history of sexual assault on minors . The lawmakers voted to set up a new category of crime: repeated sexual assault on a child or children. This category of crime carries 25 years in prison or capital punishment for a repeated sexual abuse of a child. The bill implies a so called \" Romeo and Juliet exception that is intended to avoid prosecuting high school romances.", "edit_actions": [{"type": "A", "before": null, "after": "nine-year old", "start_char_pos": 165, "end_char_pos": 165, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "hijacked and killed by sexual assaulters", "after": "raped and murdered by John Couey, who had a prior history of sexual assault on minors", "start_char_pos": 187, "end_char_pos": 227, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "offence of", "after": "assault on", "start_char_pos": 301, "end_char_pos": 311, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "\"", "start_char_pos": 474, "end_char_pos": 474, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 122, 229, 332, 444], "domain": "news"}
{"doc_id": "62050", "revision_depth": 1, "before_revision": "Again, while this may imply that the agency is unable to process all returns - even ones filed on paper - that is not explicitly stated, e.g . don't get your hopes up that you'll get away with a long delay in filing. An article in The Globe and Mail", "after_revision": "Again, while this may imply that the agency is unable to process all returns - even ones filed on paper - that is not explicitly stated, i.e . don't get your hopes up that you'll get away with a long delay in filing. An article in The Globe and Mail", "edit_actions": [{"type": "R", "before": "e.g", "after": "i.e", "start_char_pos": 137, "end_char_pos": 140, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 216], "domain": "news"}
{"doc_id": "62050", "revision_depth": 2, "before_revision": "The Agency's electronic systems apparently transposed the birth date and the user's Social Insurance Number (the Canadian equivalent to the U.S. Social Security Number) and thus corrupted all electronic databases. A reference to the incident in the news aggregator site Slashdot%DIFDELCMD < %DIFDELCMD < %%% states that no returns - not even paper ones - can be accepted, \" paper based returns will be stacking up in the mail room, as returns cannot be filed at all until the problem is fixed. \" This could be inferred from the first paragraph of the article in the Star, which reads \"a problem with electronic filing is making it impossible even to submit tax returns to the Canada Revenue Agency.\" The remainder of the article in the Star says nothing about their system for accepting paperreturns, only about the on-line and telephone systems. A check of the taxing authority's website regarding the issue states \"We have temporarily shut down public access to electronic services to ensure the integrity of taxpayer information.\" and that \"We have now traced the source of the problem to software maintenance conducted on March 4, 2007. We are currently working to bring all systems back online gradually.\" Again, while this may imply that the agency is unable to process all returns - even ones filed on paper - that is not explicitly stated, i.e. don't get your hopes up that you'll get away with a long delay in filing. An article in The Globe and Mail states that taxpayers \"can wait for Netfile to return to service, or they can print their returns and mail them to the CRA\" which indicates that the paper-based systems are unaffected . The deadline for Canadians to submit tax returns is April 30. CRA IT specialists are working on the computer system and have traced the source of the problem to software maintenance conducted on March 4, 2007. The CRA indicates it is too soon to speculate on whether the filing deadline will be extended. They expect to restore all services, including EFile and Netfile, well in advance of the filing deadline.", "after_revision": " %DIFDELCMD < %DIFDELCMD < %%% A fact sheet from the CRA states that \" until the problem is resolved, we cannot process returns filed on paper, or returns filed electronically before the system interruption. Refunds will be delayed until processing is resumed \" . A check of the taxing authority's website regarding the issue states \"We have temporarily shut down public access to electronic services to ensure the integrity of taxpayer information.\" and that \"We have now traced the source of the problem to software maintenance conducted on March 4, 2007. We are currently working to bring all systems back online gradually.\" An article in The Globe and Mail states that taxpayers \"can wait for Netfile to return to service, or they can print their returns and mail them to the CRA\" , which will be processed when the computer glitch is resolved . The deadline for Canadians to submit tax returns is April 30. The CRA indicates it is too soon to speculate on whether the filing deadline will be extended. They expect to restore all services, including EFile and Netfile, well in advance of the filing deadline.", "edit_actions": [{"type": "D", "before": "The Agency's electronic systems apparently transposed the birth date and the user's Social Insurance Number (the Canadian equivalent to the U.S. Social Security Number) and thus corrupted all electronic databases.", "after": null, "start_char_pos": 0, "end_char_pos": 213, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "D", "before": "A reference to the incident in the news aggregator site Slashdot", "after": null, "start_char_pos": 214, "end_char_pos": 278, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "states that no returns - not even paper ones - can be accepted,", "after": "A fact sheet from the CRA states that", "start_char_pos": 308, "end_char_pos": 371, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "paper based returns will be stacking up in the mail room, as returns cannot be filed at all", "after": null, "start_char_pos": 374, "end_char_pos": 465, "major_intent": "style", "raw_intents": ["style", "coherence", "clarity"]}, {"type": "R", "before": "fixed.", "after": "resolved, we cannot process returns filed on paper, or returns filed electronically before the system interruption. Refunds will be delayed until processing is resumed", "start_char_pos": 487, "end_char_pos": 493, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "This could be inferred from the first paragraph of the article in the Star, which reads \"a problem with electronic filing is making it impossible even to submit tax returns to the Canada Revenue Agency.\"", "after": ".", "start_char_pos": 496, "end_char_pos": 699, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "The remainder of the article in the Star says nothing about their system for accepting paperreturns, only about the on-line and telephone systems.", "after": null, "start_char_pos": 700, "end_char_pos": 846, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "others"]}, {"type": "D", "before": "Again, while this may imply that the agency is unable to process all returns - even ones filed on paper - that is not explicitly stated, i.e. don't get your hopes up that you'll get away with a long delay in filing.", "after": null, "start_char_pos": 1211, "end_char_pos": 1426, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "which indicates that the paper-based systems are unaffected", "after": ", which will be processed when the computer glitch is resolved", "start_char_pos": 1584, "end_char_pos": 1643, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "CRA IT specialists are working on the computer system and have traced the source of the problem to software maintenance conducted on March 4, 2007.", "after": null, "start_char_pos": 1708, "end_char_pos": 1855, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 213, 493, 699, 846, 1140, 1210, 1426, 1645, 1855, 1950], "domain": "news"}
{"doc_id": "62907", "revision_depth": 1, "before_revision": "A 34 year old obstetrician from Forestville on Sydney's Northern Beaches has escaped after he was robbed, car-jacked and locked in the boot (trunk) while his BMW was set alight. Police say the Doctor was in locked in the boot while his car was set alight, but was allowed to escape before flames engulfed the vehicle. Police allege that the men drove to several locations to withdraw cash and buy things before driving to Carisbook Street in Linley Point at around 4 AM where they set the car alight with the doctor still in the boot. Crime Manager of the Harbourside Local Area Command, Detective Inspector Houlahan said  \" He heard a click in the boot area and he heard someone call out: 'Get out of the car'\" Det Insp Houlahan said the doctor told police he did not unlock the boot himself, and it appeared the man's captors unlocked it before fleeing. He could only provide police with a vague description of his captors, but said the men were both caucasian males aged in their 20s, and that one of them was about six feet tall (183cm) with short blond hair and medium build said Det Insp Houlahan.", "after_revision": "A 34 year old obstetrician from Forestville on Sydney's Northern Beaches has escaped after he was robbed, car-jacked and locked in the boot (trunk) of his BMW, which was set alight. Police say the doctor was locked in the boot while his car was set alight, but was allowed to escape before flames engulfed the vehicle. Police allege that the men drove to several locations to withdraw cash and buy things before driving to Carisbook Street in Linley Point at around 4 a.m. local time where they set the car alight with the doctor still in the boot. Crime Manager of the Harbourside Local Area Command, Detective Inspector Houlahan said that the doctor then \" heard a click in the boot area and he heard someone call out: 'Get out of the car'\" Det . Insp. Houlahan said the doctor told police he did not unlock the boot himself, and it appeared the man's captors unlocked it before fleeing. He could only provide police with a vague description of his captors, but said the men were both Caucasian males aged in their 20s, and that one of them was about six feet tall (183cm) with short blond hair and medium build said Det . Insp. Houlahan.", "edit_actions": [{"type": "R", "before": "while his BMW", "after": "of his BMW, which", "start_char_pos": 148, "end_char_pos": 161, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "Doctor was in", "after": "doctor was", "start_char_pos": 193, "end_char_pos": 206, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "R", "before": "AM", "after": "a.m. local time", "start_char_pos": 467, "end_char_pos": 469, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "that the doctor then", "start_char_pos": 622, "end_char_pos": 622, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "D", "before": "He", "after": null, "start_char_pos": 625, "end_char_pos": 627, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "R", "before": "Insp", "after": ". Insp.", "start_char_pos": 716, "end_char_pos": 720, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "caucasian", "after": "Caucasian", "start_char_pos": 953, "end_char_pos": 962, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Insp", "after": ". Insp.", "start_char_pos": 1089, "end_char_pos": 1093, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 177, 317, 534, 855], "domain": "news"}
{"doc_id": "65593", "revision_depth": 1, "before_revision": "Campus map with important locations highlited . The shootings on Monday were in two separate locations, and occurred two hours apart. The first took place at 07:15 (12:15 UTC), at West Ambler Johnston Hall dormitory. Then, at about 09:15 (14:15 UTC), 30 people were killed in the second shooting, about half a mile (800 metres) from the first at the Norris Hall classroom building, on the same campus. Bodies were found in four classrooms and a stairwell, according to Virginia state police superintendent Col. Steve Flaherty. Some students complained that they had received no warning from the university until an e-mail alert sent more than two hours after the first incident. The university has suspended classroom courses for one week, and Norris Hall will remain closed for the rest of the semester. University offficials have expressed an interest in developing a text messaging alert system where persons can add their phone number to a list for notification in the event that incidents may warrant the use of caution or extra attention. Witness reports from the Norris Hall classroom building Norris Hall, the building at Virginia Polytechnic Institute and State University, where most of the shooting occured .Image courtesy of Christopher First-year student Erin Sheehan was one of the only 4 students coming out of the shooting in the German class unharmed. Erin explained how \"(The gunman) peeked in twice, earlier in the lesson, like he was looking for someone, somebody, before he started shooting.\"", "after_revision": "Campus map with important locations highlighted . The shootings on Monday were in two separate locations, and occurred two hours apart. The first took place at 07:15 (12:15 UTC), at West Ambler Johnston Hall dormitory. Then, at about 09:15 (14:15 UTC), 30 people were killed in the second shooting, about half a mile (800 metres) from the first at the Norris Hall classroom building, on the same campus. Bodies were found in four classrooms and a stairwell, according to Virginia state police superintendent Col. Steve Flaherty. Some students complained that they had received no warning from the university until an e-mail alert sent more than two hours after the first incident. The university has suspended classroom courses for one week, and Norris Hall will remain closed for the rest of the semester. University officials have expressed an interest in developing a text messaging alert system where persons can add their phone number to a list for notification in the event that incidents may warrant the use of caution or extra attention. Witness reports from the Norris Hall classroom building Norris Hall, the building at Virginia Polytechnic Institute and State University, where most of the shooting occurred .Image courtesy of Christopher First-year student Erin Sheehan was one of the only 4 students coming out of the shooting in the German class unharmed. Erin explained how \"(The gunman) peeked in twice, earlier in the lesson, like he was looking for someone, somebody, before he started shooting.\"", "edit_actions": [{"type": "R", "before": "highlited", "after": "highlighted", "start_char_pos": 36, "end_char_pos": 45, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "offficials", "after": "officials", "start_char_pos": 816, "end_char_pos": 826, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "occured", "after": "occurred", "start_char_pos": 1210, "end_char_pos": 1217, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 47, 133, 216, 401, 510, 526, 678, 804, 1044, 1219, 1368], "domain": "news"}
{"doc_id": "66780", "revision_depth": 1, "before_revision": "New Zealanders will have more time for sunbathing with the extension of daylight savings. The New Zealand Government has announced that the daylight savings period has been extended by three weeks to run for 27 weeks . Clocks will be brought forward one hour one week earlier on the last Sunday of September , and two weeks later on the first Sunday of April  .", "after_revision": "New Zealanders will have more time for sunbathing with the extension of daylight savings. The New Zealand Government has announced that the daylight savings period has been extended by three weeks . Clocks will be brought forward by one hour for 27 weeks in a year beginning on the last Sunday of September ( a week earlier than before), and ending on the first Sunday of April (two weeks later) .", "edit_actions": [{"type": "D", "before": "to run for 27 weeks", "after": null, "start_char_pos": 197, "end_char_pos": 216, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "one hour one week earlier", "after": "by one hour for 27 weeks in a year beginning", "start_char_pos": 250, "end_char_pos": 275, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": ", and two weeks later", "after": "( a week earlier than before), and ending", "start_char_pos": 308, "end_char_pos": 329, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "(two weeks later)", "start_char_pos": 359, "end_char_pos": 359, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 89], "domain": "news"}
{"doc_id": "66780", "revision_depth": 3, "before_revision": "New Zealanders will have more time for sunbathing with the extension of daylight savings . The New Zealand Government has announced that the daylight savings period has been extended by three weeks. Clocks will be brought forward by one hour for 27 weeks in a year beginning a week earlier on the last Sunday of September and ending on the first Sunday of April. The Department of Internal Affairs and the Internal Affairs minister, Rick Barker, also received a lot of support for a change, including from the Federated Farmers. Farmers had usually opposed a change to the daylight savings period because it hampered with operations. Helen Clark, the Prime Minister of New Zealand, stated that she personally did not support a change to the start of daylight savings in March. However the main reason that the start was altered was to avoid clashes with the fourth school term. Rick Barker says that they will be working with various companies in the information technology sector to make sure computers are updated and ready for the first new daylight savings period change since 1990. The United States and Canada have both also recently extended their daylight savings to a 34-week period.", "after_revision": "New Zealanders will have more time for sunbathing with the extension of daylight saving . The New Zealand Government has announced that the daylight saving period has been extended by three weeks. Clocks will be brought forward by one hour for 27 weeks in a year beginning a week earlier on the last Sunday of September and ending on the first Sunday of April. The Department of Internal Affairs and the Internal Affairs minister, Rick Barker, also received a lot of support for a change, including from the Federated Farmers. Farmers had usually opposed a change to the daylight saving period because it hampered with operations. Helen Clark, the Prime Minister of New Zealand, stated that she personally did not support a change to the start of daylight saving in March. However the main reason that the start was altered was to avoid clashes with the fourth school term. Rick Barker says that they will be working with various companies in the information technology sector to make sure computers are updated and ready for the first new daylight saving period change since 1990. The United States and Canada have both also recently extended their daylight saving to a 34-week period.", "edit_actions": [{"type": "R", "before": "savings", "after": "saving", "start_char_pos": 81, "end_char_pos": 88, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "savings", "after": "saving", "start_char_pos": 150, "end_char_pos": 157, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "R", "before": "savings", "after": "saving", "start_char_pos": 582, "end_char_pos": 589, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "R", "before": "savings", "after": "saving", "start_char_pos": 759, "end_char_pos": 766, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "savings", "after": "saving", "start_char_pos": 1053, "end_char_pos": 1060, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "savings", "after": "saving", "start_char_pos": 1164, "end_char_pos": 1171, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}], "sents_char_pos": [0, 90, 198, 362, 528, 633, 776, 877, 1086], "domain": "news"}
{"doc_id": "70914", "revision_depth": 1, "before_revision": "\"The dead included a 4-year-old boy, who was sleeping with his mother when the earth-built walls of their house collapsed in the quake and buried him and half of the body of his mother  ,\" the Xinhua News Agency said. \"The mother was rescued but the boy was killed.\" \"The earthquake has left two people dead, 15 seriously injured and 192 slightly hurt in ( Ning'er ) county. Some houses near the county collapsed and roads are destroyed . The communication network, and water and power supplies to the county, are also affected,\" Reuters quoted rescue officials as saying. Ning'er county is home to 190,000 people.", "after_revision": "\"The dead included a 4-year-old boy, who was sleeping with his mother when the earth-built walls of their house collapsed in the quake and buried him and half of the body of his mother the boy is dead now. ,\" the Xinhua News Agency said. \"The mother was rescued but the boy was killed.\" \"The earthquake has left two people dead, seriously injured and 192 slightly hurt in ( Ninger ) county. Some houses near the county collapsed and roads are destroyed , and the communication network, and water and power supplies to that county, are also affected,\" Reuters quoted rescue officials as saying. (Ninger)country is home to 190,000 people.", "edit_actions": [{"type": "A", "before": null, "after": "the boy is dead now.", "start_char_pos": 185, "end_char_pos": 185, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "15", "after": null, "start_char_pos": 309, "end_char_pos": 311, "major_intent": "style", "raw_intents": ["style", "style", "others"]}, {"type": "R", "before": "Ning'er", "after": "Ninger", "start_char_pos": 357, "end_char_pos": 364, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ". The", "after": ", and the", "start_char_pos": 437, "end_char_pos": 442, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "R", "before": "the", "after": "that", "start_char_pos": 498, "end_char_pos": 501, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "Ning'er county", "after": "(Ninger)country", "start_char_pos": 573, "end_char_pos": 587, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 217, 374, 438, 572], "domain": "news"}
{"doc_id": "71098", "revision_depth": 1, "before_revision": "President Bush and President Putin sign the Strategic Offensive Reductions Treaty in 2002. Just one day before the G8 summit, a testy situation involving Russian President Vladimir Putin and United States President URLe W. Bush has flared up yet again with President Bush criticizing Russia for \"derailing\" its promised democratic reforms. President Bush also attacked China in this speech, saying that \"[China's leaders] believe that they can continue to open the nation's economy without also opening its political system\". Kremlin spokesman Dmitry Peskovsaid that Russia \"cannot agree on his declaration about the derailing of democratic reforms\", adding that \"Russia is a democratic country that shares common world and common European values.\" Bush has stated that Russia is not an enemy and that \"the Cold War is over.\"", "after_revision": "President Bush and President Putin sign the Strategic Offensive Reductions Treaty in 2002. Just one day before the G8 summit, a battle in the media between Russian President Vladimir Putin and United States President URLe W. Bush continued as President Bush criticized Russia for \"derailing\" its promised democratic reforms. President Bush also criticized China in this speech, saying that \"[China's leaders] believe that they can continue to open the nation's economy without also opening its political system\". Kremlin spokesperson, Dmitry Peskov, said Russia \"cannot agree on his declaration about the derailing of democratic reforms\", adding that \"Russia is a democratic country that shares common world and common European values.\" Bush stated that Russia is not an enemy and that \"the Cold War is over.\"", "edit_actions": [{"type": "R", "before": "testy situation involving", "after": "battle in the media between", "start_char_pos": 128, "end_char_pos": 153, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "has flared up yet again with President Bush criticizing", "after": "continued as President Bush criticized", "start_char_pos": 228, "end_char_pos": 283, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "attacked", "after": "criticized", "start_char_pos": 360, "end_char_pos": 368, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "spokesman Dmitry Peskovsaid that", "after": "spokesperson, Dmitry Peskov, said", "start_char_pos": 534, "end_char_pos": 566, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "has", "after": null, "start_char_pos": 754, "end_char_pos": 757, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "coherence"]}], "sents_char_pos": [0, 90, 339, 525, 748], "domain": "news"}
{"doc_id": "71310", "revision_depth": 1, "before_revision": "Debbie Gerbich, who was involved in consensual group sex sessions with former and serving New Zealand police officers in 2001, has been found dead. One of those former police officers on a videotape made during the group sex sessions was convicted of rape but acquitted in two high profile trials dating back to historic sex allegations involving police officers serving in Rotorua in the 1980s. The New Zealand Herald revealed the identity of Mrs Gerbich on Sunday 31 March after she gave a rival newspaper access to the videotape made in 2001. The Sunday News had published information revealing the existence of the videotape but did not reveal her identity. Mrs Gerbich was reported by the Sunday News to have been devastated after her name was subsequently published by the New Zealand Herald. Mrs Gerbich claimed to have received intimidating emails from a New Zealand Herald journalist  after she refused to speak with them.  It is not known how exactly the New Zealand Herald learned her identity and that she was being interviewed as part of police operation Austin investigating the historic sex allegations . The cause of death has not been released although it is believed to be suicide .", "after_revision": "Debbie Gerbich, who was involved in consensual group sex sessions with former and serving New Zealand police officers in 2001, committed suicide on June 6, 2007. One of those former police officers on a videotape made during the group sex sessions was convicted of rape but acquitted in two high profile trials dating back to historic sex allegations involving police officers serving in Rotorua in the 1980s. The New Zealand Herald revealed the identity of Mrs Gerbich on Sunday 31 March after she gave a rival newspaper access to the videotape made in 2001. The Sunday News had published information revealing the existence of the videotape but did not reveal her identity. Mrs Gerbich was reported by the Sunday News to have been devastated after her name was subsequently published by the New Zealand Herald. Mrs Gerbich claimed to have received intimidating emails from New Zealand Herald journalist Steven Cook after she refused to speak with them. Two emails sent from Steven Cooks email account were presented and read out to a reporter on the New Zealand current affairs program Close Up, one of them stated; It is not known how exactly the New Zealand Herald learned her identity and that she was being interviewed as part of police operation Austin investigating the historic sex allegations .", "edit_actions": [{"type": "R", "before": "has been found dead.", "after": "committed suicide on June 6, 2007.", "start_char_pos": 127, "end_char_pos": 147, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "a", "after": null, "start_char_pos": 861, "end_char_pos": 862, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "A", "before": null, "after": "Steven Cook", "start_char_pos": 893, "end_char_pos": 893, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Two emails sent from Steven Cooks email account were presented and read out to a reporter on the New Zealand current affairs program Close Up, one of them stated;", "start_char_pos": 932, "end_char_pos": 932, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 1118, "end_char_pos": 1119, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "D", "before": "The cause of death has not been released although it is believed to be suicide", "after": null, "start_char_pos": 1120, "end_char_pos": 1198, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 147, 395, 545, 661, 798, 931, 1119], "domain": "news"}
{"doc_id": "71313", "revision_depth": 1, "before_revision": "Myanmar authorities have detained 11 HIV/AIDS patients  in a Yangon hospital following the group's recent demonstrations calling for the release of detained HIV/AIDS activist Phyu Phyu Thinn. The patients are being detained at Waybargi Infectious Diseases Hospital on the outskirts of Yangon. \"The authorities made necessary preparations in advance at the hospital to confine them in isolation as 'political patients ' ,\" Myint Thein of the National League for Democracy (NLD) was quoted as saying by Reuters . \" These 11 people had been holding prayers at pagodas for the release of Phyu Phyu Thinn since May 23. \"  Another HIV/AIDS activist and colleague of Phyu Phyu Thinn, Yazar , said the patients were taken from their homes to the hospital on Tuesday after being told they would receive proper treatment. ", "after_revision": "Myanmar authorities detained 11 HIV/AIDS patients for three days in a Yangon hospital following the group's recent demonstrations calling for the release of detained HIV/AIDS activist Phyu Phyu Thinn. The patients were being detained at Waybargi Infectious Diseases Hospital on the outskirts of Yangon. \"The patients were given a series of tests and all of them were freed Friday ,\" Yazar, a colleague of Phyu Phyu Thinn and a supporter of the National League for Democracy , was quoted as saying by Associated Press . \" Three of the patients were given ARV (anti-retroviral) treatment and all of them were provided with cash to return to their homes in their districts, \" said Yazar, who like many Myanmar people only has one name. Yazar said the patients were taken from their homes to the hospital on Tuesday after being told they would receive proper treatment. \"The authorities made necessary preparations in advance at the hospital to confine them in isolation as 'political patients',\" Myint Thein of the National League for Democracy (NLD) was quoted as saying by Reuters. \"These 11 people had been holding prayers at pagodas for the release of Phyu Phyu Thinn since May 23.\"", "edit_actions": [{"type": "D", "before": "have", "after": null, "start_char_pos": 20, "end_char_pos": 24, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "for three days", "start_char_pos": 55, "end_char_pos": 55, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "are", "after": "were", "start_char_pos": 205, "end_char_pos": 208, "major_intent": "style", "raw_intents": ["style", "style", "fluency"]}, {"type": "R", "before": "authorities made necessary preparations in advance at the hospital to confine them in isolation as 'political patients '", "after": "patients were given a series of tests and all of them were freed Friday", "start_char_pos": 298, "end_char_pos": 418, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Myint Thein of", "after": "Yazar, a colleague of Phyu Phyu Thinn and a supporter of", "start_char_pos": 422, "end_char_pos": 436, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "(NLD)", "after": ",", "start_char_pos": 471, "end_char_pos": 476, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "Reuters", "after": "Associated Press", "start_char_pos": 501, "end_char_pos": 508, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "These 11 people had been holding prayers at pagodas for the release of Phyu Phyu Thinn since May 23.", "after": "Three of the patients were given ARV (anti-retroviral) treatment and all of them were provided with cash to return to their homes in their districts,", "start_char_pos": 513, "end_char_pos": 613, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "said Yazar, who like many Myanmar people only has one name.", "start_char_pos": 616, "end_char_pos": 616, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Another HIV/AIDS activist and colleague of Phyu Phyu Thinn, Yazar ,", "after": "Yazar", "start_char_pos": 617, "end_char_pos": 684, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "\"The authorities made necessary preparations in advance at the hospital to confine them in isolation as 'political patients',\" Myint Thein of the National League for Democracy (NLD) was quoted as saying by Reuters. \"These 11 people had been holding prayers at pagodas for the release of Phyu Phyu Thinn since May 23.\"", "start_char_pos": 812, "end_char_pos": 812, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}], "sents_char_pos": [0, 191, 292, 613], "domain": "news"}
{"doc_id": "71412", "revision_depth": 1, "before_revision": "Manchester Cathedral, England The Church of England has threatened legal action against Sony Entertainment over the inclusion of the Manchester Cathedral in the PS3 launch title Resistance: Fall of Man. The game ( set in various English locations during an alternate history ) includes a gun fight within the building. The scene in question involves a small group of humans (which include the player) taking refuge in the cathedral and attempting to fight off aliens who are attacking them. Meanwhile  a spokesman for Sony defended the game, telling the Daily Telegraph, \"[the game] is not based on reality at all\" and comparing it to the fictional series Doctor Who. The Church will send a letter directly to Sony on Monday requesting an apology and that they remove the game from sale.", "after_revision": "Manchester Cathedral, England The Church of England has threatened legal action against Sony Entertainment over the inclusion of the Manchester Cathedral in the PS3 launch title Resistance: Fall of Man. The game , set in various English locations during an alternate history , includes a gun fight within the building. The scene in question involves a small group of humans , including the player, taking refuge in the cathedral and attempting to fight off aliens who are attacking them. Meanwhile , a spokesman for Sony defended the game, telling the Daily Telegraph, \"[the game] is not based on reality at all\" and compared it to the fictional series Doctor Who. The Church will send a letter directly to Sony on Monday requesting an apology and withdrawal of the game from sale.", "edit_actions": [{"type": "R", "before": "(", "after": ",", "start_char_pos": 212, "end_char_pos": 213, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ")", "after": ",", "start_char_pos": 275, "end_char_pos": 276, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "(which include the player)", "after": ", including the player,", "start_char_pos": 374, "end_char_pos": 400, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 501, "end_char_pos": 501, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "comparing", "after": "compared", "start_char_pos": 619, "end_char_pos": 628, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "that they remove", "after": "withdrawal of", "start_char_pos": 751, "end_char_pos": 767, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 202, 318, 490, 667], "domain": "news"}
{"doc_id": "71477", "revision_depth": 1, "before_revision": " Memorial's gray panels rise above the crowd. The Minnesota World War II Veterans Memorial was dedicated Saturday on in front of the State Capitol. The memorial features ten glass panels etched with text and scenes from World War II, within a plaza area at the foot of a mall leading to the front entrance to the Capitol. The glass panels are three feet wide, eight feet tall, and mounted on Minnesota granite. Well over twelve thousand people attended. Stan Turner was the Master of Ceremonies during the dedication. The keynote speaker was General John Vessey(Ret. ). Among the many veterans honored were Stan Kowalski and \"Chuck\" Lindberg; the latter was one of those who raised the first flag over Iwo Jima. Contruction was authorized by the State Legislature in 2000.", "after_revision": "300px| Memorial's gray panels rise above the crowd. The Minnesota World War II Veterans Memorial was dedicated Saturday in front of the State Capitol. The memorial features ten glass panels etched with text and scenes from World War II, within a plaza area at the foot of a mall leading to the front entrance to the Capitol. The glass panels are three feet wide, eight feet tall, and mounted on Minnesota granite. Well over twelve thousand people attended. Stan Turner was the Master of Ceremonies during the dedication. The keynote speaker was retired general John Vessey. Among the many veterans honored were Stan Kowalski and \"Chuck\" Lindberg; the latter was one of those who raised the first flag over Iwo Jima. Construction was authorized by the State Legislature in 2000.", "edit_actions": [{"type": "A", "before": null, "after": "300px|", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "on", "after": null, "start_char_pos": 114, "end_char_pos": 116, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "General John Vessey(Ret. ).", "after": "retired general John Vessey.", "start_char_pos": 542, "end_char_pos": 569, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Contruction", "after": "Construction", "start_char_pos": 712, "end_char_pos": 723, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 45, 147, 321, 410, 453, 517, 569, 642, 711], "domain": "news"}
{"doc_id": "71577", "revision_depth": 1, "before_revision": "Map highlighting Sevenoaks, England. A woman, reported to be an expert in martial arts, has been shot dead by police in a car park behind M&Co, a women's clothing shop in Sevenoaks, Kent, England. According to reports, the woman pulled a weapon on police who then shot her. A bus station in Sevenoaks has been shut down and closed off to the public while police investigate the incident. A tent has been placed over the crime scene to ensure that evidence is kept intact.", "after_revision": "Map highlighting Sevenoaks, England. A woman, reported to be an expert in martial arts, has been shot dead by police in a car park behind M&Co, a women's clothing shop in Sevenoaks, Kent, England. According to reports, the woman pulled a gun on police who then shot her. A bus station in Sevenoaks was shut down and closed off to the public while police investigated the incident. A tent was placed over the crime scene to ensure that evidence was kept intact.", "edit_actions": [{"type": "R", "before": "weapon", "after": "gun", "start_char_pos": 238, "end_char_pos": 244, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "has been", "after": "was", "start_char_pos": 301, "end_char_pos": 309, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "investigate", "after": "investigated", "start_char_pos": 362, "end_char_pos": 373, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "has been", "after": "was", "start_char_pos": 395, "end_char_pos": 403, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "is", "after": "was", "start_char_pos": 456, "end_char_pos": 458, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}], "sents_char_pos": [0, 36, 196, 273, 387], "domain": "news"}
{"doc_id": "71813", "revision_depth": 1, "before_revision": "Graham grew noticeably weaker over the course of the past several weeks, suffering from degenerative osteoarthritis of the back and neck. She also underwent treatment for pneumonia two weeks ago. She asked for her feeding tube to be removed Wednesday , after her request, and consultation with family . Billy Graham made a mention of her during the dedication of his library, saying that she was bedridden, and that she was \"getting closer to meeting God in heaven.\" He also announced that the couple will be buried at the library.", "after_revision": "Graham had grown noticeably weaker over the course of the past several weeks, suffering from degenerative osteoarthritis of the back and neck. She also underwent treatment for pneumonia two weeks ago. She asked for her feeding tube to be removed Wednesday ; following her request, and after consultation with her family, she ceased to receive further nutrients . Billy Graham made a mention of her during the dedication of his library, saying that she was bedridden, and that she was \"getting closer to meeting God in heaven.\" He also announced that the couple will be buried at the library.", "edit_actions": [{"type": "R", "before": "grew", "after": "had grown", "start_char_pos": 7, "end_char_pos": 11, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": ", after", "after": "; following", "start_char_pos": 251, "end_char_pos": 258, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "consultation with family", "after": "after consultation with her family, she ceased to receive further nutrients", "start_char_pos": 276, "end_char_pos": 300, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 137, 195, 466], "domain": "news"}
{"doc_id": "71813", "revision_depth": 2, "before_revision": "Graham had grown noticeably weaker over the course of the past several weeks, suffering from degenerative osteoarthritis of the back and neck . She also underwent treatment for pneumonia two weeks ago. She asked for her feeding tube to be removed Wednesday; following her request, and after consultation with her family, she ceased to receive further nutrients. Billy Graham made a mention of her during the dedication of his library, saying that she was bedridden, and that she was \"getting closer to meeting God in heaven .\" He also announced that the couple will be buried at the library.", "after_revision": "She had grown noticeably weaker over the course of the past several weeks, suffering from degenerative osteoarthritis of the back and neck , and also underwent treatment for pneumonia two weeks ago. She asked for her feeding tube to be removed Wednesday; following her request, and after consultation with her family, she ceased to receive further nutrients. Billy Graham made a mention of her during the dedication of his library, saying that she was bedridden, and that she was \"getting closer to meeting God in Heaven .\" He also announced that the couple will be buried at the library.", "edit_actions": [{"type": "R", "before": "Graham", "after": "She", "start_char_pos": 0, "end_char_pos": 6, "major_intent": "style", "raw_intents": ["meaning-changed", "style", "style"]}, {"type": "R", "before": ". She", "after": ", and", "start_char_pos": 142, "end_char_pos": 147, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "R", "before": "heaven", "after": "Heaven", "start_char_pos": 517, "end_char_pos": 523, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 143, 201, 257, 361], "domain": "news"}
{"doc_id": "72776", "revision_depth": 1, "before_revision": "The EU would like to make it easier to dismiss employees. At the same time, social security should be improved. Then, thanks to the improved flexibility of the market, it should be simpler to find a job. Surveys showes that Europeans mostly seem to agree with  need to adapt and change. The concept of flexicurity, preached by the European Commission, is aimed to encourage EU governments to unfasten their employment policies. According to Vladimr pidla, Member of the European Commission,  a too protective employment policy may paralyze the labour-market. Such a policy \"extends the time needed for jobseeking and the period of unemployment\", writes the Commission in its report. On its paper, the Commission is inspired by  social model of the Scandinavian countries - namely by the social system established in Denmark in the late 1990's.", "after_revision": "The European Union would like to make it easier to dismiss employees. At the same time, social security should be improved. Then, thanks to the improved flexibility of the market, it should be simpler to find a job. Surveys show that Europeans mostly seem to agree with the need to adapt and change. The concept of flexicurity, preached by the European Commission, is aimed to encourage EU governments to unfasten their employment policies. According to Vladimr pidla, the European Commissioner for Employment, Social Affairs Equal Opportunities, a too protective employment policy may paralyze the labour-market. Such a policy \"extends the time needed for jobseeking and the period of unemployment\", writes the Commission in its report. On its paper, the Commission is inspired by the social model of the Scandinavian countries - namely by the social system established in Denmark in the late 1990's.", "edit_actions": [{"type": "R", "before": "EU", "after": "European Union", "start_char_pos": 4, "end_char_pos": 6, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "showes", "after": "show", "start_char_pos": 212, "end_char_pos": 218, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 260, "end_char_pos": 260, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "Member of the European Commission,", "after": "the European Commissioner for Employment, Social Affairs", "start_char_pos": 456, "end_char_pos": 490, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "Equal Opportunities,", "start_char_pos": 491, "end_char_pos": 491, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 727, "end_char_pos": 727, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}], "sents_char_pos": [0, 57, 111, 203, 286, 427, 558, 682], "domain": "news"}
{"doc_id": "73114", "revision_depth": 1, "before_revision": "During Hurricane Camille in 1969, a 72 foot tugboat named the East Point washed ashore on Gulfport's West Beach. Lucille Moody, a beachfront property owner renamed the boat the S.S. Hurricane Camille and built a gift shop behind it, after moving it on specially built railroad tracks. The boat and gift shop became a popular attraction for tourists visiting the Mississippi Gulf Coast. But on August 29, 2005 Hurricane Katrina came ashore pushing a 20 plus foot storm surge leaving mostly everything on the beachfront destroyed. Only a few things were standing including the Biloxi Lighthouse, the Friendship Oak, and the S.S. Hurricane Camille. Lucille Moody decided not to rebuild the gift shop, so there the S.S. Camille stands, a rusted hulk on a beachfront that is quickly rebuilding. People want it gone, after all it is an eyesore as they say, but then there are those who want to save thier history. After all, we have lost so much, Grass Lawn , Tullis-Toledo Manor , and other things, we should try to save the history that we still have. After a new coat of paint and a little work the S.S. Hurricane Camille could look a lot better. You can't expect something thats survived two hurricanes, the latter one of the worst on record not to be a little beat up. The dents and scratches just add to its charm .", "after_revision": "During Hurricane Camille in 1969, a 72 foot tugboat named the East Point washed ashore on Gulfport's West Beach. Lucille Moody, a beachfront property owner renamed the boat the S.S. Hurricane Camille and built a gift shop behind it, after moving it on specially built railroad tracks. The boat and gift shop became a popular attraction for tourists visiting the Mississippi Gulf Coast. But on August 29, 2005 Hurricane Katrina came ashore pushing a 20 plus foot storm surge leaving mostly everything on the beachfront destroyed. Only a few things were standing including the Biloxi Lighthouse, the Friendship Oak, and the S.S. Hurricane Camille. Lucille Moody decided not to rebuild the gift shop, so there the S.S. Camille stands, a rusted hulk on a beachfront that is quickly rebuilding. People want it gone, suggesting it is an \"eyesore\", but others suggest this historic object be preserved, especially after the loss of historic sites including Grass Lawn and Tullis-Toledo Manor . They suggest a coat of paint and some repairs could make the boat much more appealing .", "edit_actions": [{"type": "R", "before": "after all", "after": "suggesting", "start_char_pos": 811, "end_char_pos": 820, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "eyesore as they say, but then there are those who want to save thier history. After all, we have lost so much, Grass Lawn ,", "after": "\"eyesore\", but others suggest this historic object be preserved, especially after the loss of historic sites including Grass Lawn and", "start_char_pos": 830, "end_char_pos": 953, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", and other things, we should try to save the history that we still have. After a new", "after": ". They suggest a", "start_char_pos": 974, "end_char_pos": 1059, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "a little work the S.S. Hurricane Camille could look a lot better. You can't expect something thats survived two hurricanes, the latter one of the worst on record not to be a little beat up. The dents and scratches just add to its charm", "after": "some repairs could make the boat much more appealing", "start_char_pos": 1078, "end_char_pos": 1313, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 112, 284, 385, 528, 645, 789, 907, 1047, 1143, 1267], "domain": "news"}
{"doc_id": "73814", "revision_depth": 1, "before_revision": "1500 people in northern Nevada town of Winnemucca had to be evacuate this weekend as a 8,000-acre 55 sqaure miles wildfire burned danger close to the towns southern end. The fire threatened homes,business and a power substation the cause of the wildfire is believed to lighting strikes", "after_revision": "1500 people in northern Nevada town of Winnemucca had to be evacuated this weekend as a 8,000-acre 55 square mile wildfire burned dangerously close to the town's southern end. The fire threatened homes,business and a power substation . Lightning strikes are believed to have started the blaze.", "edit_actions": [{"type": "R", "before": "evacuate", "after": "evacuated", "start_char_pos": 60, "end_char_pos": 68, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "sqaure miles wildfire burned danger", "after": "square mile wildfire burned dangerously", "start_char_pos": 101, "end_char_pos": 136, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "towns", "after": "town's", "start_char_pos": 150, "end_char_pos": 155, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "the cause of the wildfire is believed to lighting strikes", "after": ". Lightning strikes are believed to have started the blaze.", "start_char_pos": 228, "end_char_pos": 285, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 169], "domain": "news"}
{"doc_id": "73816", "revision_depth": 1, "before_revision": "The mosque has been accused of enforcing it's own Islamic law in the capital, notable the illegal detention of Chinese foreign nationals working in massage parlors. The mosques leaders follow the Hanafi school of thought, believing that the Koran is the authority by which justice should by metered. Historically the mosque has had close links with the Pakistan leadership, and members of the current administration have resisted calls for a crackdown on the mosques antagonistic behaviour.", "after_revision": "The mosque has been accused of enforcing it's own Islamic law in the capital, notably the illegal detention of Chinese foreign nationals working in massage parlors. The mosques leaders follow the Hanafi school of thought, believing that the Koran is the authority by which justice should be metered. Historically the mosque has had close links with the Pakistan leadership, and members of the current administration have resisted calls for a crackdown on the mosques provocative behaviour.", "edit_actions": [{"type": "R", "before": "notable", "after": "notably", "start_char_pos": 78, "end_char_pos": 85, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "by", "after": "be", "start_char_pos": 288, "end_char_pos": 290, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "antagonistic", "after": "provocative", "start_char_pos": 467, "end_char_pos": 479, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}], "sents_char_pos": [0, 164, 299], "domain": "news"}
{"doc_id": "74638", "revision_depth": 1, "before_revision": "Map of Ukraine highlighting the location of Lviv Oblast and with red dot marking location of Lviv. A freight train in Lviv, Ukraine derailed, caught fire  and spilled a toxic chemical, releasing the fumes into the air early Tuesday morning (local time)  and people who live near the site of the crash are still getting sick. Officials say that the air that surrounds at least 14 villages near the scene  is at least 23\\% above normal limitations . So far 70 people,  19 of those are children, have been hospitalized after they were poisoned by the fumes from train cars filled with yellow phosphorous. That number is up compared to 15 which were seeking medical attention on Tuesday. The European Union has stated that they will continue to monitor the gas cloud saying that toxic gas clouds do not \"respect borders\" of countries and will travel. The train was shipping cargo from Kazakhstan to Poland when it derailed. Authorities have ruled out sabotage and terrorism as playing a role in the crash and are still investigating the accident. At least 15 cars, of the 58 the train was pulling, derailed and or tipped over .", "after_revision": "Map of Ukraine highlighting the location of Lviv Oblast and with red dot marking location of Lviv. A freight train in Lviv, Ukraine derailed, caught fire , and spilled a toxic chemical, releasing dangerous fumes into the air early Tuesday morning (local time) , and people who live near the site of the crash are still becoming sick. Officials say that the toxic cloud, surrounding at least 14 villages near the scene , is at least 23\\% above benign levels . So far 70 people, including 19 children, have been hospitalized after being poisoned by the fumes from train cars filled with yellow phosphorous. That number is up from the 15 seeking medical attention on Tuesday. The European Union has stated that they will continue to monitor the gas cloud because toxic gas clouds do not \"respect borders\" of countries and will travel. The train was shipping cargo from Kazakhstan to Poland when it derailed. Authorities have ruled out sabotage and terrorism from playing a role in the crash and are still investigating the accident. At least 15 cars, of the 58 the train was pulling, derailed or were overturned .", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 154, "end_char_pos": 154, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "the", "after": "dangerous", "start_char_pos": 195, "end_char_pos": 198, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 253, "end_char_pos": 253, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "getting", "after": "becoming", "start_char_pos": 311, "end_char_pos": 318, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "air that surrounds", "after": "toxic cloud, surrounding", "start_char_pos": 348, "end_char_pos": 366, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 403, "end_char_pos": 403, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "normal limitations", "after": "benign levels", "start_char_pos": 427, "end_char_pos": 445, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "including", "start_char_pos": 466, "end_char_pos": 466, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "of those are", "after": null, "start_char_pos": 470, "end_char_pos": 482, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "they were", "after": "being", "start_char_pos": 522, "end_char_pos": 531, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "compared to", "after": "from the", "start_char_pos": 620, "end_char_pos": 631, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "which were", "after": null, "start_char_pos": 635, "end_char_pos": 645, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "saying that", "after": "because", "start_char_pos": 763, "end_char_pos": 774, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "as", "after": "from", "start_char_pos": 970, "end_char_pos": 972, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "and or tipped over", "after": "or were overturned", "start_char_pos": 1103, "end_char_pos": 1121, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 98, 324, 447, 601, 683, 846, 919, 1042], "domain": "news"}
{"doc_id": "7557", "revision_depth": 1, "before_revision": " 400px|In the above file illustration, the long yellow portion at the upper right represents the skyway of the new $1.5 billion Bay Bridge span. The \"T\"-shaped objects in the lower right corner represent the piers with the alleged substandard welds. Source: Caltrans The Federal Bureau of Investigation is investigating allegations of wrongdoing in the building of the new $ 1.5 billion \"skyway\" portion of the San Francisco-Oakland Bay Bridge spanning the San Francisco Bay between Oakland and Yerba Buena Island , according to a report first published in the Oakland Tribune. According to reports, no criminal charges have been filed. The paper reported that welders for KFM, the company under contract to build the bridge, told authorities that they were pressured and even paid monthly bonuses of between $200 and $600 for performing substandard welds on the support piers for the bridge if it meant the project proceeded faster. But Caltrans, the state agency in charge of building the bridge said their inspectors checked the welds and determined that the work, although rushed, met safety standards. According to a report by Oakland television station, KTVU, the Office of the Inspector General of the U.S. Department of Labor and the U.S. Department of Transportation are joining the FBI in the investigation of the matter. The federal government is getting involved in the investigation because more than $1 million in U.S. taxpayer money was already spent during construction. If faulty welds are found in the support piers, according to reports, much of the new construction may have to be torn down to be rebuilt. The current 68-year-old span was damaged in the 1989 Loma Prieta earthquake. At $ 6.2 billion total, the new bridge qualifies as the largest public works project in California's history.", "after_revision": "April 7, 2005 Category:April 7, 2005 400px|In the above file illustration, the long yellow portion at the upper right represents the skyway of a new $1.5 billion Bay Bridge span. The \"T\"-shaped objects in the lower right corner represent the piers with the alleged substandard welds. Source: Caltrans The Federal Bureau of Investigation is investigating allegations of wrongdoing in the building of a new $ 1.5 billion \"skyway\" portion of the San Francisco-Oakland Bay Bridge , according to a report published in the Oakland Tribune. The bridge spans the San Francisco Bay between Oakland and Yerba Buena Island. There are so far no criminal charges filed. The paper reported that welders for KFM, the company under contract to build the bridge, told authorities they were pressured and even paid monthly bonuses of between $200 and $600 for performing substandard welds on the support piers for the bridge if it meant the project proceeded faster. But Caltrans, the state agency in charge of building the bridge said their inspectors checked the welds and determined the work, although rushed, met safety standards. According to a report by Oakland television station, KTVU, the Office of the Inspector General of the U.S. Department of Labor and the U.S. Department of Transportation are joining the FBI in the investigation of the matter. The federal government is getting involved in the investigation because more than $1 million in U.S. taxpayer money was spent during its construction. If questionable welds are found in the support piers, according to reports, much of the new construction may have to be torn down to be rebuilt. The current 68-year-old span was damaged in the 1989 Loma Prieta earthquake. At $ 6.2 billion total, the new bridge qualifies as the largest public works project in California's history.", "edit_actions": [{"type": "A", "before": null, "after": "April 7, 2005 Category:April 7, 2005", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the new $1.5 billion Bay Bridge span. The \"T\"-shaped objects in the lower right corner represent the piers with the alleged substandard welds. Source: Caltrans The Federal Bureau of Investigation is investigating allegations of wrongdoing in the building of the new $", "after": "a new $1.5 billion Bay Bridge span. The \"T\"-shaped objects in the lower right corner represent the piers with the alleged substandard welds. Source: Caltrans The Federal Bureau of Investigation is investigating allegations of wrongdoing in the building of a new $", "start_char_pos": 107, "end_char_pos": 374, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "D", "before": "spanning the San Francisco Bay between Oakland and Yerba Buena Island", "after": null, "start_char_pos": 444, "end_char_pos": 513, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "first", "after": null, "start_char_pos": 538, "end_char_pos": 543, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "According to reports,", "after": "The bridge spans the San Francisco Bay between Oakland and Yerba Buena Island. There are so far", "start_char_pos": 578, "end_char_pos": 599, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "D", "before": "have been", "after": null, "start_char_pos": 620, "end_char_pos": 629, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "that", "after": null, "start_char_pos": 743, "end_char_pos": 747, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "D", "before": "that", "after": null, "start_char_pos": 1053, "end_char_pos": 1057, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "$1 million in U.S. taxpayer money was already spent during construction. If faulty welds are found in the support piers, according to reports, much of the new construction may have to be torn down to be rebuilt. The current 68-year-old span was damaged in the 1989 Loma Prieta earthquake. At $", "after": "$1 million in U.S. taxpayer money was spent during its construction. If questionable welds are found in the support piers, according to reports, much of the new construction may have to be torn down to be rebuilt. The current 68-year-old span was damaged in the 1989 Loma Prieta earthquake. At $", "start_char_pos": 1414, "end_char_pos": 1707, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 144, 249, 577, 636, 933, 1106, 1331, 1486, 1625, 1702], "domain": "news"}
{"doc_id": "75851", "revision_depth": 1, "before_revision": "Kevin Garnett speaking at a press conference in Taiwan in 2006. Kevin Garnett, known as \" Big Ticket\", is going to the Boston Celtics in a blockbuster trade involving at least 6 players. The forward , who has spent his entire twelve season career playing for the Minnesota Timberwolves, will join fellow all-stars Paul Pierce and Ray Allen to give Boston a boost in the Eastern Conference.", "after_revision": "Kevin Garnett speaking at a press conference in Taiwan in 2006. Citing \" league sources,\" both CNN/SI and the Boston Globe are reporting that Kevin Garnett is going to the Boston Celtics in a trade involving at least 6 players. The forward, known as \"Big Ticket\" , who has spent his entire twelve season career playing for the Minnesota Timberwolves, will join fellow all-stars Paul Pierce and Ray Allen to give Boston a boost in the Eastern Conference.", "edit_actions": [{"type": "R", "before": "Kevin Garnett, known as", "after": "Citing", "start_char_pos": 64, "end_char_pos": 87, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "Big Ticket\",", "after": "league sources,\" both CNN/SI and the Boston Globe are reporting that Kevin Garnett", "start_char_pos": 90, "end_char_pos": 102, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "blockbuster", "after": null, "start_char_pos": 139, "end_char_pos": 150, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "The forward", "after": "The forward, known as \"Big Ticket\"", "start_char_pos": 187, "end_char_pos": 198, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 63, 186]}
{"doc_id": "77591", "revision_depth": 1, "before_revision": "Hacked iPhone for auction on eBay A second phone, which was also hacked by Hotz, was placed for sale on eBay , with bids now reaching US$50,100 after starting at $540, although bids up to $99,999,999 were canceled . He states at the auction site \"Soon, you will be able to buy unlocked iPhones everywhere, in fact probably before this auction ends. But this is the one that started it all. The one unlocked while the steps were posted live on the blog. This isn't just an unlocked iPhone, this is a piece of history.\" However,Hotz also commented on the excessive bids showing up on eBay. \"I'm sure these most recent bids are fake.\" ", "after_revision": "Hacked iPhone for auction on eBay A second phone, which was also hacked by Hotz, was placed for sale on eBay . He states at the auction site \"Soon, you will be able to buy unlocked iPhones everywhere, in fact probably before this auction ends. But this is the one that started it all. The one unlocked while the steps were posted live on the blog. This isn't just an unlocked iPhone, this is a piece of history.\" Starting at $540, bids reached $99,999,999, although many bids were canceled as fraudulent. Before their removal, Hotz commented on the excessive bids , stating \"I'm sure these most recent bids are fake.\" The auction was canceled early by Hotz due to \"an error in the listing\".", "edit_actions": [{"type": "D", "before": ", with bids now reaching US$50,100 after starting at $540, although bids up to $99,999,999 were canceled", "after": null, "start_char_pos": 109, "end_char_pos": 213, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "However,Hotz also", "after": "Starting at $540, bids reached $99,999,999, although many bids were canceled as fraudulent. Before their removal, Hotz", "start_char_pos": 518, "end_char_pos": 535, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "showing up on eBay.", "after": ", stating", "start_char_pos": 568, "end_char_pos": 587, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "The auction was canceled early by Hotz due to \"an error in the listing\".", "start_char_pos": 632, "end_char_pos": 632, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 215, 348, 389, 452, 517, 587], "domain": "news"}
{"doc_id": "78712", "revision_depth": 1, "before_revision": "According to the Canadian Press, the recent tadpoles also escaped from the local wetlands into the highway instead of using the fence that guides them. An interesting fact is that when the tadpoles come back in the winter they do not use the highway, they cross at streams and rivers instead. The Ministry of Transportation has now put buckets beside a fence along the highway so the tadpoles jump into the buckets and local volunteers and conservation groups bring the tadpoles across the road into the Millar Creek area. The tadpoles were described as brownish green or black and were camouflaged.", "after_revision": "According to the Canadian Press, the recent young toads also escaped from the local wetlands into the highway instead of using the fence that guides them. An interesting fact is that when the toads come back in the winter they do not use the highway, they cross at streams and rivers instead. The Ministry of Transportation has now put buckets beside a fence along the highway so the toadlets jump into the buckets and local volunteers and conservation groups bring them across the road into the Millar Creek area. The toadlets were described as brownish green or black and were camouflaged.", "edit_actions": [{"type": "R", "before": "tadpoles", "after": "young toads", "start_char_pos": 44, "end_char_pos": 52, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "tadpoles", "after": "toads", "start_char_pos": 189, "end_char_pos": 197, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": "tadpoles", "after": "toadlets", "start_char_pos": 384, "end_char_pos": 392, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "the tadpoles", "after": "them", "start_char_pos": 466, "end_char_pos": 478, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "tadpoles", "after": "toadlets", "start_char_pos": 527, "end_char_pos": 535, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "others"]}], "sents_char_pos": [0, 151, 292, 522], "domain": "news"}
{"doc_id": "78730", "revision_depth": 1, "before_revision": "At the annual European Society of Cardiology Summit in Vienna, Austria, a female cardiologist has suffered a heart attack in front of her colleagues . The 46-year-old Italian woman collapsed on Sunday and was immediately revived by one of her fellow heart specialists . The woman was transferred to the general hospital in Vienna, where she  died later that day . Reinhard Krepler, chief doctor of the hospital, stated that her death was due to a congenital heart problem .", "after_revision": "At the annual European Society of Cardiology Summit in Vienna, Austria, the wife of a cardiologist suffered a heart attack while queueing with her husband to register for the congress . 46-year-old Silvia Celeste from Italy, the wife of Dr. Fabrizio Celeste, collapsed on Sunday and was immediately revived using a defibrillator . The woman was transferred to the general hospital in Vienna, where she was operated upon. She died later that night . Reinhard Krepler, chief doctor of the hospital, stated that her death was due to an underlying congenital heart disease .", "edit_actions": [{"type": "R", "before": "a female cardiologist has", "after": "the wife of a cardiologist", "start_char_pos": 72, "end_char_pos": 97, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in front of her colleagues", "after": "while queueing with her husband to register for the congress", "start_char_pos": 122, "end_char_pos": 148, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "D", "before": "The", "after": null, "start_char_pos": 151, "end_char_pos": 154, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "Italian woman", "after": "Silvia Celeste from Italy, the wife of Dr. Fabrizio Celeste,", "start_char_pos": 167, "end_char_pos": 180, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "by one of her fellow heart specialists", "after": "using a defibrillator", "start_char_pos": 229, "end_char_pos": 267, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "A", "before": null, "after": "was operated upon. She", "start_char_pos": 341, "end_char_pos": 341, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "day", "after": "night", "start_char_pos": 358, "end_char_pos": 361, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "a congenital heart problem", "after": "an underlying congenital heart disease", "start_char_pos": 445, "end_char_pos": 471, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 150, 269], "domain": "news"}
{"doc_id": "80174", "revision_depth": 1, "before_revision": "Detainees, or individuals in United States custody considered to be suspects connected to terrorism will no longer be able to protest their being held in custody or their treatment while in custody, other wise known as the \"writ of Habeas Corpus  \" at least not in U.S. federal courtrooms ; striking a victory for Republicans . Democrats  needed at least 60 votes in order for the bill to allow detainees to fight and they only received 56, which included six Republicans. 43 Republicans voted against the bill. \"[The bills failure to pass] calls into question the United States' historic role of defender of human rights in the world. It accomplishes what opponents could never accomplish on the battlefield, whittling away our own liberties. This is America? , \" said Democratic Senator for Vermont, Patrick J. Leahy. In 2006, the U.S. Military Commissions Act severely limited the rights of detainees at Guantanamo Bay, the U.S. run prison in Cuba for the most dangerous individuals in the world, allegedly accused of horrific acts of terrorism. The act also included smaller U.S. prisons around the globe. Republicans claim that to allow detainees those rights, the U.S. would have the potential to be engulfed in an endless line of lawsuits.", "after_revision": "Detainees, or individuals in United States custody considered to be suspects connected to terrorism will no longer be able to protest their being held in custody or their treatment while in custody, other wise known as the \"writ of Habeas Corpus , \" at least not in U.S. federal courtrooms . Democrats , who sponsored the bill, needed at least 60 votes in order for the bill to allow detainees to fight but received only 56, which included six Republicans. 43 Republicans voted against the bill. \"[The bill's failure to pass] calls into question the United States' historic role of defender of human rights in the world. It accomplishes what opponents could never accomplish on the battlefield, whittling away our own liberties. This is America? \" said Democratic Senator for Vermont, Patrick J. Leahy. In 2006, the U.S. Military Commissions Act limited the rights of detainees at Guantanamo Bay, the U.S. run dentention facility in Cuba for the most dangerous individuals in the world, accused of acts of terrorism. The act also included smaller U.S. prisons around the globe. Opponents of the bill claim that to allow detainees those rights, the U.S. would potentially be engulfed in an endless line of lawsuits.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 246, "end_char_pos": 246, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "; striking a victory for Republicans", "after": null, "start_char_pos": 289, "end_char_pos": 325, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": ", who sponsored the bill,", "start_char_pos": 338, "end_char_pos": 338, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "and they only received", "after": "but received only", "start_char_pos": 414, "end_char_pos": 436, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "bills", "after": "bill's", "start_char_pos": 518, "end_char_pos": 523, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 761, "end_char_pos": 762, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "D", "before": "severely", "after": null, "start_char_pos": 863, "end_char_pos": 871, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "prison", "after": "dentention facility", "start_char_pos": 936, "end_char_pos": 942, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "allegedly accused of horrific", "after": "accused of", "start_char_pos": 1000, "end_char_pos": 1029, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Republicans", "after": "Opponents of the bill", "start_char_pos": 1110, "end_char_pos": 1121, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "have the potential to", "after": "potentially", "start_char_pos": 1181, "end_char_pos": 1202, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}], "sents_char_pos": [0, 290, 472, 511, 635, 743, 819, 1048, 1109], "domain": "news"}
{"doc_id": "80486", "revision_depth": 1, "before_revision": "200px|Britney Spears performing in Los Angeles, CA in 2004. The Los Angeles City Attorney's Office said yesterday that singing pop music super-star Britney Spears was facing charges for a hit and run  and driving without a valid drivers' license. Charges were filed following an accident in which police allege Spears smashed her car into another vehicle in a parking lot back in August of 2007. Spears could face a maximum of six months in jail and $2,000 in fines if convicted. Some paparazzi caught a video recording of Spears steering her car into another vehicle on August 6 as Spears tried to pull into a parking spot in a Studio City lot.", "after_revision": "200px|Britney Spears performing in Los Angeles, CA in 2004. The Los Angeles City Attorney's Office said yesterday that pop music super-star Britney Spears was facing charges for a hit and run offence and driving without a valid drivers' license. Charges were filed following an accident in which police allege Spears smashed her car into another vehicle in a parking area in August of 2007. She could face a maximum of six months in jail and $2,000 in fines if convicted. A group of paparazzi recorded a video Britney steering her car into another vehicle on August 6 as she tried to pull into a parking spot in a Studio City lot.", "edit_actions": [{"type": "D", "before": "singing", "after": null, "start_char_pos": 119, "end_char_pos": 126, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "offence", "start_char_pos": 200, "end_char_pos": 200, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "lot back", "after": "area", "start_char_pos": 368, "end_char_pos": 376, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Spears", "after": "She", "start_char_pos": 396, "end_char_pos": 402, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Some paparazzi caught a video recording of Spears", "after": "A group of paparazzi recorded a video Britney", "start_char_pos": 480, "end_char_pos": 529, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Spears", "after": "she", "start_char_pos": 583, "end_char_pos": 589, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 59, 246, 395, 479], "domain": "news"}
{"doc_id": "85084", "revision_depth": 1, "before_revision": "Location of Iran A top US commander says that the United States is not planning a preemptive attack on Iran despite increasing rhetoric from Washington, D.C. On Sunday, The Observer ran a report saying that US interrogators are under pressure to find evidence in Iran. The United States Secretary of State Condoleezza Rice also stated that there is no  authorization for military action against Iran. Citing Micah Brose, a privately contracted interrogator working for the US military in Iraq, the British newspaper, The Observer reported there is \" huge \" pressure on interrogators to uncover evidence pointing to Iran . \"They push a lot for us to establish a link with Iran when interrogating Iraqi insurgents \", Brose claimed in the interview, adding that \"it feels a lot like, if you get something and Iran's not involved, it's a let down.\" He further claimed that people have said to him that \"they're really pushing the Iran thing.\" Brose denied being asked to manufacture evidence, but stated that \"if a detainee wants to tell me what I want to hear so he can get out of jail ... you know what I'm saying.\" The article by The Observer  said most military intelligence officers refused to comment but said that one said \"The message is, 'Got to find a link with Iran, got to find a link with Iran.' It's sickening.\" A strike is \"not in the offing\" - William J. Fallon  Admiral William J. Fallon, the commanding officer of United States Central Command which is responsible for the Middle East, East Africa and Central Asia, speaking in Monday's Financial Times, said that a strike againt Iran is \"not in the offing.\" \"None of this is helped by the continuing stories that just keep going around and around and around that any day now there will be another war which is just not where we want to go,\" Fallon continued. \"Getting Iranian behavior to change and finding ways to get them to come to their senses and do that is the real objective. Attacking them as a means to get to that spot strikes me as being not the first choice in my book.\"  \"The question is, why will Tehran not talk to us?\" - Condoleezza Rice Meanwhile, US Secretary of State Condoleezza Rice was on ABC's This Week on Sunday, saying that a Senate resolution designating Iran's Islamic Revolutionary Guard as a URLanization does not authorize military action. \"This resolution is saying that there needs to be strong measures taken against Iran, which we have definitely done,\" Rice continued. \"And if the Iranians suspend their enrichment and reprocessing, I'm prepared to meet my counterpart anyplace, anytime, anywhere. So the question isn't why will we not talk to Tehran. The question is, why will Tehran not talk to us?\" \"No congressional authority exists for unilateral military action against Iran\" - Hillary Clinton Senator Hillary Clinton, who is running for President, voted for the resolution. She has said that the resolution could not be used to justify war, but rather it seeks to escalate diplomatic efforts. \"We wish to emphasize that no congressional authority exists for unilateral military action against Iran,\" she said in an open letter on November 1, 2007. Attack on Iran would be a \"nightmare\" - Joschka Fischer Last week, former Foreign Minister of Germany, Joschka Fischer said that he was worried that the upcoming US Presidential election could heighten tensions between the US and Iran. He found particularly worrisome a statement by Presidential candidate Rudy Giuliani in October that \"If I am President of the United States, I guarantee you, we will never find out what they will do if they get nuclear weapons, because they are not going to get a nuclear weapon .  \" The United States government accuses Iran of arming insurgents in neighbouring Iraq.It has also refused to rule out military action against Iran for its alleged attempts to build nuclear weapons. The US is accused of using exaggerated and fabricated evidence to build its case against Iraq prior to the war.", "after_revision": "Location of Iran A top US commander said yesterday that the United States is not planning a preemptive attack on Iran despite increasing rhetoric from Washington, D.C. The United States Secretary of State Condoleezza Rice also denied that an attack was imminent, stating that there is no Congressional authorization for military action against Iran. The denials came as The Observer ran a report saying that US personnel involved in interrogating insurgents in Iraq are under \"huge\" pressure to find evidence against Iran. It quoted Micah Brose, a privately contracted interrogator working for the US military there, as saying that information on Iran is \" gold \" . \"They push a lot for us to establish a link with Iran \", Brose claimed in the interview, adding that \"it feels a lot like, if you get something and Iran's not involved, it's a let down.\" He further claimed that people have said to him that \"they're really pushing the Iran thing.\" Brose denied being asked to manufacture evidence, but stated that \"if a detainee wants to tell me what I want to hear so he can get out of jail ... you know what I'm saying.\" The US has in the past been accused of using exaggerated and fabricated evidence to build its case against Iraq prior to the war. The article by the British based newspaper said most military intelligence officers refused to comment but that one said \"The message is, 'Got to find a link with Iran, got to find a link with Iran.' It's sickening.\" A strike is \"not in the offing\" - William J. Fallon On Monday, Admiral William J. Fallon, the commanding officer of United States Central Command which is responsible for the Middle East, East Africa and Central Asia, speaking in Financial Times, said that a strike against Iran is \"not in the offing.\" \"None of this is helped by the continuing stories that just keep going around and around and around that any day now there will be another war which is just not where we want to go,\" Fallon continued. \"Getting Iranian behavior to change and finding ways to get them to come to their senses and do that is the real objective. Attacking them as a means to get to that spot strikes me as being not the first choice in my book.\" Despite saying this, he refused to rule out a strike against Iran in the future. \"The question is, why will Tehran not talk to us?\" - Condoleezza Rice Meanwhile, US Secretary of State Condoleezza Rice was on ABC's This Week on Sunday, saying that a Senate resolution designating Iran's Islamic Revolutionary Guard as a URLanization does not authorize military action. \"This resolution is saying that there needs to be strong measures taken against Iran, which we have definitely done,\" Rice continued. \"And if the Iranians suspend their enrichment and reprocessing, I'm prepared to meet my counterpart anyplace, anytime, anywhere. So the question isn't why will we not talk to Tehran. The question is, why will Tehran not talk to us?\" \"No congressional authority exists for unilateral military action against Iran\" - Hillary Clinton Senator Hillary Clinton, who is running for President, voted for the resolution. She has said that the resolution could not be used to justify war, but rather it seeks to escalate diplomatic efforts. \"We wish to emphasize that no congressional authority exists for unilateral military action against Iran,\" she said in an open letter on November 1, 2007. Attack on Iran would be a \"nightmare\" - Joschka Fischer Last week, former Foreign Minister of Germany, Joschka Fischer said that he was worried that the upcoming US Presidential election could heighten tensions between the US and Iran. He found particularly worrisome a statement by Presidential candidate Rudy Giuliani in October that \"If I am President of the United States, I guarantee you, we will never find out what they will do if they get nuclear weapons, because they are not going to get a nuclear weapon . .. the military option is not off the table. \" ", "edit_actions": [{"type": "R", "before": "says", "after": "said yesterday", "start_char_pos": 36, "end_char_pos": 40, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "On Sunday, The Observer ran a report saying that US interrogators are under pressure to find evidence in Iran. The", "after": "The", "start_char_pos": 158, "end_char_pos": 272, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "stated that", "after": "denied that an attack was imminent, stating that", "start_char_pos": 328, "end_char_pos": 339, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Congressional", "start_char_pos": 352, "end_char_pos": 352, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Citing", "after": "The denials came as The Observer ran a report saying that US personnel involved in interrogating insurgents in Iraq are under \"huge\" pressure to find evidence against Iran. It quoted", "start_char_pos": 401, "end_char_pos": 407, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "in Iraq, the British newspaper, The Observer reported there", "after": "there, as saying that information on Iran", "start_char_pos": 485, "end_char_pos": 544, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "huge", "after": "gold", "start_char_pos": 550, "end_char_pos": 554, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "pressure on interrogators to uncover evidence pointing to Iran", "after": null, "start_char_pos": 557, "end_char_pos": 619, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "D", "before": "when interrogating Iraqi insurgents", "after": null, "start_char_pos": 676, "end_char_pos": 711, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "article by The Observer", "after": "US has in the past been accused of using exaggerated and fabricated evidence to build its case against Iraq prior to the war.", "start_char_pos": 1118, "end_char_pos": 1141, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "The article by the British based newspaper", "start_char_pos": 1142, "end_char_pos": 1142, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "said", "after": null, "start_char_pos": 1207, "end_char_pos": 1211, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "On Monday,", "start_char_pos": 1374, "end_char_pos": 1374, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "Monday's", "after": null, "start_char_pos": 1542, "end_char_pos": 1550, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "againt", "after": "against", "start_char_pos": 1587, "end_char_pos": 1593, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "Despite saying this, he refused to rule out a strike against Iran in the future.", "start_char_pos": 2048, "end_char_pos": 2048, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ".. the military option is not off the table.", "start_char_pos": 3673, "end_char_pos": 3673, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "The United States government accuses Iran of arming insurgents in neighbouring Iraq.It has also refused to rule out military action against Iran for its alleged attempts to build nuclear weapons. The US is accused of using exaggerated and fabricated evidence to build its case against Iraq prior to the war.", "after": null, "start_char_pos": 3676, "end_char_pos": 3983, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 268, 400, 844, 938, 1113, 1304, 1321, 1622, 1823, 1947, 2335, 2469, 2598, 2652, 2881, 3000, 3155, 3391, 3760, 3871], "domain": "news"}
{"doc_id": "85106", "revision_depth": 1, "before_revision": "In the final of  Konami Cup 2007, Chunichi Dragons battled SK Wyverns again, many local (Japan) spectators hoped Dragons revenge on Wyverns and win the champion. And this is really a intensive match with the stating pitchers URLe Kenneth Rayborn from SK versus Daisuke Yamai. The key innings took place at the 9th inning, Hirokazu Ibata (Dragons) hit a key run back to make their team lead with 6-5. Finally, the relief pitcher Hitoki Iwase successfully shut down SK's offenses and tied the score to the end to win the champion. After this game, Hiromitsu Ochiai (Coach of Chunichi Dragons ) remarked: \"Even though we ever lose to SK Wyverns, but participants in this year proved their actual strength more than last year. In this game, after we led with 6-5, Hitoki Iwase told me he want to pitch at the bottom of 9th inning, then I accepted, he promised his wish and finally helped us win the champion. I'll not URLet the contributions by Ibata and Iwase.\"", "after_revision": "In the final of the Konami Cup 2007, Chunichi Dragons battled SK Wyverns again, many Japanese spectators hoped for Dragons to take revenge on Wyverns and win the championship. This was really a intensive match with starting pitchers URLe Kenneth Rayborn from SK versus Daisuke Yamai. The key innings took place at the 9th inning, Hirokazu Ibata (Dragons) hit a key run back to make their team lead with 6-5. Finally, the relief pitcher Hitoki Iwase successfully shut down SK's offense and tied the score to the end to win the champion. After this game, Chunichi Dragons coach Hiromitsu Ochiai remarked: \"Even though we ever lose to SK Wyverns, but participants in this year proved their actual strength more than last year. In this game, after we led with 6-5, Hitoki Iwase told me he want to pitch at the bottom of 9th inning, then I accepted, he promised his wish and finally helped us win the champion. I'll not URLet the contributions by Ibata and Iwase.\"", "edit_actions": [{"type": "A", "before": null, "after": "the", "start_char_pos": 16, "end_char_pos": 16, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "local (Japan) spectators hoped Dragons", "after": "Japanese spectators hoped for Dragons to take", "start_char_pos": 82, "end_char_pos": 120, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "champion. And this is", "after": "championship. This was", "start_char_pos": 152, "end_char_pos": 173, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the stating", "after": "starting", "start_char_pos": 204, "end_char_pos": 215, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "offenses", "after": "offense", "start_char_pos": 469, "end_char_pos": 477, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Hiromitsu Ochiai (Coach of Chunichi Dragons )", "after": "Chunichi Dragons coach Hiromitsu Ochiai", "start_char_pos": 546, "end_char_pos": 591, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 161, 275, 399, 528, 722, 904], "domain": "news"}
{"doc_id": "86080", "revision_depth": 1, "before_revision": "Gordon Brown, in the United States on July 30, 2007. The Prime Minister of the United Kingdom, Gordon Brown has announced that donations made to his political party, the Labour Party were \"not lawfully declared\" as they were donated through a middleman. According to UK law, all donations over 50,000 must be declared along with the name of the donator and all donations made through middlemen must declare the original donator. David Abraham officially donated the money, which totalled almost 600,000, through 3 third parties. They have been named as; Ray Ruddick who donated 196,850 since 2003, Janet Kidd, who has donated 185,000 since 2003 and John McCarthy, who has donated 202,125 since 2004. The acting leader of the Liberal Democrats, Vince Cable, whose party suffered from the increase in Labour's popularity said \"an enourmous can of worms had been opened up\" and that the UK government should have introduced new regulations on the fuding of political parties.", "after_revision": "Gordon Brown, in the United States on July 30, 2007. The Prime Minister of the United Kingdom, Gordon Brown has announced that donations made to his political party, the Labour Party were \"not lawfully declared\" as they were donated through a middleman. According to UK law, all donations over 50,000 must be declared along with the name of the donator and all donations made through middlemen must declare the original donator. David Abraham officially donated the money, which totaled almost 600,000, through 3 third parties. They have been named as; Ray Ruddick who donated 196,850 since 2003, Janet Kidd, who has donated 185,000 since 2003 and John McCarthy, who has donated 202,125 since 2004. The acting leader of the Liberal Democrats, Vince Cable, whose party suffered from the increase in Labour's popularity said \"an enormous can of worms had been opened up\" and that the UK government should have introduced new regulations on the funding of political parties.", "edit_actions": [{"type": "R", "before": "totalled", "after": "totaled", "start_char_pos": 479, "end_char_pos": 487, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "enourmous", "after": "enormous", "start_char_pos": 828, "end_char_pos": 837, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "fuding", "after": "funding", "start_char_pos": 944, "end_char_pos": 950, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 52, 253, 428, 528, 553, 699], "domain": "news"}
{"doc_id": "86541", "revision_depth": 1, "before_revision": "As the two rivals postured over factors like performance and high-definition, Wikinews reporter Rico Shen interviewed some of their partners. BenQ and Micro-Star International both said: \"No matter whether the consumer chooses a computer with AMD or Intel, the consumer should consider it based on functionality and practically before purchasing. Some low-price NBs (EeePC, OLPC) were launched before the show , NB manufacturers had considered and evaluated what a consumer really want. Sincerely, I recommend a consumer not be fooled by a NB's price but to consider its functionality first!\"", "after_revision": "As the two rivals postured over factors like performance and high-definition, Wikinews reporter Rico Shen interviewed some of their partners. BenQ and Micro-Star International both said: \"No matter whether the consumer chooses a computer with AMD or Intel, the consumer should consider it based on functionality and practically before purchasing. For example, some low-price notebook computers (EeePC, OLPC) were launched before the show because notebook computer manufacturers had considered and evaluated what a consumer really want. Sincerely, I recommend a consumer not be fooled by a its price but to consider its functionality first!\"", "edit_actions": [{"type": "R", "before": "Some", "after": "For example, some", "start_char_pos": 347, "end_char_pos": 351, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "NBs", "after": "notebook computers", "start_char_pos": 362, "end_char_pos": 365, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", NB", "after": "because notebook computer", "start_char_pos": 410, "end_char_pos": 414, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "NB's", "after": "its", "start_char_pos": 540, "end_char_pos": 544, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}], "sents_char_pos": [0, 346, 486], "domain": "news"}
{"doc_id": "86561", "revision_depth": 1, "before_revision": "The former Mayor of Sefton in the United Kingdom has been gaoled for fraudulantly obtaining state benefits. He was gaoled for 15 months for conspiring to defraud 32,871.85 in Disability Living Allowance together with a six months concurrently sentence for conspiracy to defraud 4,000.25 in Income Support. His wife, Catie Walker, was gaoled for eight months for the DLA offence and a concurrent four months for the Income Support swindle.", "after_revision": "The former Mayor of Sefton in the United Kingdom has been jailed for fraudulantly obtaining state benefits. He was jailed for 15 months for conspiring to defraud 32,871.85 in Disability Living Allowance together with a six months concurrently sentence for conspiracy to defraud 4,000.25 in Income Support. His wife, Catie Walker, was jailed for eight months for the DLA offence and a concurrent four months for the Income Support swindle.", "edit_actions": [{"type": "R", "before": "gaoled", "after": "jailed", "start_char_pos": 58, "end_char_pos": 64, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "gaoled", "after": "jailed", "start_char_pos": 115, "end_char_pos": 121, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "gaoled", "after": "jailed", "start_char_pos": 334, "end_char_pos": 340, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}], "sents_char_pos": [0, 107, 305], "domain": "news"}
{"doc_id": "87132", "revision_depth": 1, "before_revision": "Former American Idol finalist Jessica Sierra, 22, has had her invitation to perform at an upcoming concert for United States troops revoked , following her recent arrest for disorderly conduct while intoxicated. Sierra has been in prison since the start of the month after an altercation with police officers outside a Tampa nightclub, which she had been ejected from. She is charged with disorderly intoxication and resisting arrest. Patterson said URL , who URLanizing the concert, had received comments from patrons who were concerned about her participation. John Fitzgibbons, attorney for Jessica, refused to comment.", "after_revision": "Former American Idol finalist Jessica Sierra, 22, has had her invitation to perform at an upcoming concert for United States troops retracted , following her recent arrest for disorderly conduct while intoxicated. Sierra has been in jail since the start of the month after an altercation with police officers outside a Tampa nightclub, which she had been ejected from. She is charged with disorderly intoxication and resisting arrest. Patterson said Move America Forward , who URLanizing the concert, had received comments from patrons who were concerned about her participation. John Fitzgibbons, attorney for Jessica, refused to comment.", "edit_actions": [{"type": "R", "before": "revoked", "after": "retracted", "start_char_pos": 132, "end_char_pos": 139, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "prison", "after": "jail", "start_char_pos": 231, "end_char_pos": 237, "major_intent": "clarity", "raw_intents": ["clarity", "style", "meaning-changed"]}, {"type": "R", "before": "URL", "after": "Move America Forward", "start_char_pos": 450, "end_char_pos": 453, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}], "sents_char_pos": [0, 211, 368, 434, 562], "domain": "news"}
{"doc_id": "89094", "revision_depth": 1, "before_revision": "First official sunspot belonging to the new Solar Cycle 24. The first sunspot on The Sun of solar cycle 24 was announced on January 4, 2008. This sunspot is a precursor for the normal increase in activity which takes place during the 11-year solar cycles .", "after_revision": "First official sunspot belonging to the new Solar Cycle 24. The first sunspot marking a new solar cycle has been identified, physicists at the NOAA announced on January 4. This sunspot is a precursor for the normal increase in activity which takes place during the 11-year solar cycle .", "edit_actions": [{"type": "R", "before": "on The Sun of solar cycle 24 was", "after": "marking a new solar cycle has been identified, physicists at the NOAA", "start_char_pos": 78, "end_char_pos": 110, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "4, 2008.", "after": "4.", "start_char_pos": 132, "end_char_pos": 140, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "cycles", "after": "cycle", "start_char_pos": 248, "end_char_pos": 254, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 59, 140], "domain": "news"}
{"doc_id": "96430", "revision_depth": 1, "before_revision": "The third Test  commenced on Wednesday at the WACA in Perth , between Australia and India . India won the toss, and Indian captain Anil Kumble elected to bat. India quickly made a half century in nine overs , then shortly afterwards the first wicket fell with Virender Sehwag caught behind by Adam Gilchrist for 29 runs. Three overs later, about about half an hour before lunch  , the next wicket fell when Gilchrist also caught Wasim Jaffer. India were 2/74 at lunch, and played well during the afternoon, not losing any wickets. Rahul Dravid was dropped on 11 runs by Michael Clarke, and went on to share a partnership of 139 runs with Sachin Tendulkar as India reached 2/177 at tea. In the last session India put on 120 runs, but lost four wickets, including two just before stumps.  Tendulkar (71) was trapped lbw by Brett Lee , while Sourav Ganguly fell for just nine runs soon afterwards. Dravid reached 93 runs before being caught by Ricky Ponting, and V. V. S. Laxman was dismissed for 27 runs shortly before the close of play. India finished the day on 6/297.", "after_revision": "The third Test between Australia and India commenced on Wednesday at the WACA in Perth . India won the toss, and Indian captain Anil Kumble elected to bat. India quickly made a half century in nine overs . Shortly after, the first wicket fell with Virender Sehwag caught behind by Adam Gilchrist for 29 runs. About half an hour before lunch with three more overs , the next wicket fell when Gilchrist also caught Wasim Jaffer. India were 2/74 at lunch, and played well during the afternoon, losing no wickets. Rahul Dravid was dropped on 11 runs by Michael Clarke, and went on to share a partnership of 139 runs with Sachin Tendulkar as India reached 2/177 at tea. In the last session India put on 120 runs, losing four wickets, including two just before stumps. Brett Lee trapped lbw Tendulkar (71) , while Sourav Ganguly fell for just nine runs soon afterwards. Dravid reached 93 runs before being caught by Ricky Ponting, and V. V. S. Laxman was dismissed for 27 runs shortly before the close of play. India finished the day on 6/297.", "edit_actions": [{"type": "A", "before": null, "after": "between Australia and India", "start_char_pos": 15, "end_char_pos": 15, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ", between Australia and India", "after": null, "start_char_pos": 60, "end_char_pos": 89, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": ", then shortly afterwards", "after": ". Shortly after,", "start_char_pos": 207, "end_char_pos": 232, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "Three overs later, about about", "after": "About", "start_char_pos": 321, "end_char_pos": 351, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "with three more overs", "start_char_pos": 378, "end_char_pos": 378, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "others", "meaning-changed"]}, {"type": "R", "before": "not losing any", "after": "losing no", "start_char_pos": 507, "end_char_pos": 521, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "but lost", "after": "losing", "start_char_pos": 729, "end_char_pos": 737, "major_intent": "coherence", "raw_intents": ["coherence", "style", "fluency"]}, {"type": "A", "before": null, "after": "Brett Lee trapped lbw", "start_char_pos": 786, "end_char_pos": 786, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "was trapped lbw by Brett Lee", "after": null, "start_char_pos": 802, "end_char_pos": 830, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 91, 158, 320, 442, 530, 685, 785, 894, 1035], "domain": "news"}
{"doc_id": "96656", "revision_depth": 1, "before_revision": "240px|Executives from left to right: Jean-Claude Poimboeuf (French Representative to Taiwan), Alain Jaubert, Sylvia Feng (President of PTS Taiwan), and Mingchuan Huang. Alain Jaubert, director from the famous art program \"Palettes\" by Association Relative la Tlvision Europenne (ARTE  was invited by Taiwan Public Television Service (PTS Taiwan) and the French Institute in Taipei to talk about productions on his notable art program \"Palettes\" on January 19 in Taipei City. He also talked about some experiences and about functionalities on art programs with the Chairman of National Culture and Arts Foundation Mingchuan Huang. According to PTS Taiwan, Jaubert failed at his first attempt to produce \"Palettes\" in 1984, but finally succeeded in 1987 after his proposal was recalled and Jaubery tried another production style to introduce history and background on establishments of a notable picture to Europe. This program was finally recognized and encouraged by administration centers from three ancient monuments (Louvre, Muse d'Orsay, and Pompidou Centre). This also echoed Jaubert's words in his speech: \"Without supports by art comrades and those administrations centers, the 'Palettes' will no longer exist, eventually, I won't become notable in Europe.\" Before the Executive Panel, the PTS Taiwan screened highlights from some art programs included \"Pelettes\" by ARTE and \"Avant Guard Liberation\" by Mingchuan Huang. After the screening, executives talked about specifications on art programs and forecasted the future of directions on art programs world-wide.", "after_revision": "240px|Executives from left to right: Jean-Claude Poimboeuf (French Representative to Taiwan), Alain Jaubert, Sylvia Feng (President of PTS Taiwan), and Mingchuan Huang. Alain Jaubert, director of the famous art program \"Palettes\" by Association Relative la Tlvision Europenne (ARTE ) was invited by Taiwan Public Television Service (PTS Taiwan) and the French Institute in Taipei to talk about productions on his art program on January 19 in Taipei City. He also talked about some experiences and about functionalities on art programs with the Chairman of National Culture and Arts Foundation Mingchuan Huang. According to PTS Taiwan, Jaubert failed at his first attempt to produce \"Palettes\" in 1984, but finally succeeded in 1987 after his proposal was recalled and Jaubery tried another production style to introduce history and background on establishments of notable pictures to Europe. This program was finally recognized and encouraged by administration centers from three ancient monuments (Louvre, Muse d'Orsay, and Pompidou Centre). This also echoed Jaubert's words in his speech: \"Without support by art comrades and those administrations centers, 'Palettes' would not exist; I wouldn't have become notable in Europe.\" Before the Executive Panel, the PTS Taiwan screened highlights from some art programs including \"Pelettes\" by ARTE and \"Avant Guard Liberation\" by Mingchuan Huang. After the screening, executives talked about specifications on art programs and forecasted the future of directions on art programs world-wide.", "edit_actions": [{"type": "R", "before": "from", "after": "of", "start_char_pos": 193, "end_char_pos": 197, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ")", "start_char_pos": 284, "end_char_pos": 284, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "notable art program \"Palettes\"", "after": "art program", "start_char_pos": 414, "end_char_pos": 444, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "a notable picture", "after": "notable pictures", "start_char_pos": 884, "end_char_pos": 901, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "supports", "after": "support", "start_char_pos": 1121, "end_char_pos": 1129, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 1181, "end_char_pos": 1184, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "will no longer exist, eventually, I won't", "after": "would not exist; I wouldn't have", "start_char_pos": 1196, "end_char_pos": 1237, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "included", "after": "including", "start_char_pos": 1351, "end_char_pos": 1359, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}], "sents_char_pos": [0, 168, 474, 629, 912, 1063, 1264, 1427], "domain": "news"}
{"doc_id": "98301", "revision_depth": 1, "before_revision": "Map showing the North and South of Waziristan, al-Libi was killed in the Northern part. Abu Laith al-Libi, an al-Qaeda leader in Afghanistan, was reported killed on Islamist website URL, saying: \"[Libi] was martyred with a group of his brothers in the land of Muslim Pakistan....Though we are sad for his loss, he left a legacy that will inflame the enemy nation and religion.\" It is speculated that he was hit by a U.S. Missile , which is now identified as  a Predator drone, in the North Waziristan of Pakistan, and a dozen more militants were also reported dead. The News reports the attack was aimed at the leader as well as Obaidah al-Masri. \"But it's not bin Laden and Zawahri. They do have an ability to regenerate and replace these guys,\" said a Western intelligence official. Libi is said to be the number 3 person in al-Qaeda.", "after_revision": "Map showing the North and South of Waziristan, al-Libi was killed in the Northern part. Abu Laith al-Libi, an al-Qaeda leader in Afghanistan, was reported killed on Islamist website URL, saying: \"[Libi] was martyred with a group of his brothers in the land of Muslim Pakistan....Though we are sad for his loss, he left a legacy that will inflame the enemy nation and religion.\" It is speculated that he was hit by a United States missile , which is now identified as being fired from a Predator drone, in the North Waziristan of Pakistan, and a dozen more militants were also reported dead. The News reports the attack was aimed at al-Libi, as well as Obaidah al-Masri. \"But it's not bin Laden and Zawahri. They do have an ability to regenerate and replace these guys,\" said a Western intelligence official. Al-Libi is said to have been the third highest ranking member of al-Qaeda.", "edit_actions": [{"type": "R", "before": "U.S. Missile", "after": "United States missile", "start_char_pos": 416, "end_char_pos": 428, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "A", "before": null, "after": "being fired from", "start_char_pos": 458, "end_char_pos": 458, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "others", "meaning-changed"]}, {"type": "R", "before": "the leader", "after": "al-Libi,", "start_char_pos": 607, "end_char_pos": 617, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Libi", "after": "Al-Libi", "start_char_pos": 785, "end_char_pos": 789, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "be the number 3 person in", "after": "have been the third highest ranking member of", "start_char_pos": 801, "end_char_pos": 826, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 87, 279, 377, 565, 646, 683, 784], "domain": "news"}
{"doc_id": "99200", "revision_depth": 1, "before_revision": "The ruins of a Palestinian village. The town of Kiryat Yam in Israel is suing Internet corporation Google over an alleged inaccuracy in its Google Earth software, which claims the town was built over the ruins of an Arab village. However, town official Naty Keyzilberman says this \"obviously cannot be true, because Kiryat Yam was founded in 1945\". The Palestinian exodus, in which hundreds of thousands of Arabs abandoned their villages due to the Arab-Israeli War, occurred in 1948, three years after Keyzilberman says the town was founded. The town filed a slander complaint with Israel police, he says. The town 's claim is supported by Yossi Ben-Artzi, a history professor at the University of Haifa, who said, \"Kiryat Yam was built on sand dunes, and there wasn't any Palestinian village in the area. The lands were bought in 1939 by the Gav Yam construction company.\"", "after_revision": "The ruins of a Palestinian village. The city of Kiryat Yam in Israel is suing Internet corporation Google over an alleged inaccuracy in its Google Earth software, which claims the city was built over the ruins of an Arab village. However, town 's spokesman Naty Key Zilberman says this \"obviously cannot be true, because Kiryat Yam was founded in 1945\". The Palestinian exodus, in which hundreds of thousands of Arabs abandoned their villages due to the Arab-Israeli War, occurred in 1948, three years after Key zilberman says the city was founded. The city filed a slander complaint with Israel police, he says. The city 's claim is supported by Yossi Ben-Artzi, a history professor at the University of Haifa, who said, \"Kiryat Yam was built on sand dunes, and there wasn't any Palestinian village in the area. The lands were bought in 1939 by the Gav Yam construction company.\"", "edit_actions": [{"type": "R", "before": "town", "after": "city", "start_char_pos": 40, "end_char_pos": 44, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "town", "after": "city", "start_char_pos": 180, "end_char_pos": 184, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "official Naty Keyzilberman", "after": "'s spokesman Naty Key Zilberman", "start_char_pos": 244, "end_char_pos": 270, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Keyzilberman says the town", "after": "Key zilberman says the city", "start_char_pos": 503, "end_char_pos": 529, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "town", "after": "city", "start_char_pos": 547, "end_char_pos": 551, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "town", "after": "city", "start_char_pos": 611, "end_char_pos": 615, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 35, 229, 348, 542, 606, 806], "domain": "news"}
{"doc_id": "9961", "revision_depth": 1, "before_revision": "Leading off is the week-long Harley rally, which then will be followed in the next week by BikeFest. In and around town, both day and night are punctured by the sounds of bike engines gunned and revved at stop lights and in parkings lots, while groups of cycle riders dominate the streets. The highly accesorised bikes, decked with chome and polished to show it, flashed the townscape. Choppers made a showing, but road hogs dominated the ridership, often going twosome. Many rally goers arrived on the scene with SUV\u2019s or big pickup trucks towing cargo trailers loaded with cycles. 240px Growth in the sheer size of the two rallies led police to make changes in the handling of traffic flow. During BikeFest last year, the mostly black crowd that came in on the heels of the largely white Harley rally the week earlier, were faced with confusion when the two-lane Ocean Blvd was made one-way. A branch of the NAACP in Conway, the next town over from Myrtle Beach, alledged discrimination by Horry County and Myrtle Beach Police. They claimed authorities and police used an overwhelming and aggressive police presence, combined with a restrictive one-way traffic pattern, to intimidate and discourage the participants in the rally. A plan to submit an opposition to the notice has already been announced by Michael Navarre, an attorney for Steptoe & Johnson, who represents the NAACP civil rights group. \"We certainly don't think the judge has ruled erroneously,\" Navarre said, according to \u2018\u2019 The Sun News \u2019\u2019 .", "after_revision": "Leading off is the week-long Harley rally, followed by the next week 's BikeFest. In and around town, both day and night are punctured by the sounds of bike engines gunned and revved at stop lights and in parking lots. Groups of cycle riders dominate the streets. The highly accesorised bikes, decked with chrome and polished to show it, flashed the townscape. Choppers made a showing, but road hogs dominated the ridership, often going twosome. Many rally goers arrived on the scene with SUV\u2019s or big pickup trucks towing cargo trailers loaded with cycles. 240px Growth in the sheer size of the two rallies led police to make changes in the handling of traffic flow. During BikeFest last year, the mostly black crowd that came in on the heels of the largely white Harley rally the week earlier, were faced with confusion when the two-lane Ocean Blvd was made one-way. A branch of the NAACP in Conway, the next town over from Myrtle Beach, alleged discrimination by Horry County and Myrtle Beach Police. They claimed authorities and police used an overwhelming and aggressive police presence, combined with a restrictive one-way traffic pattern, to intimidate and discourage the participants in the rally. A plan to submit an opposition to the notice has already been announced by Michael Navarre, an attorney for Steptoe & Johnson, who represents the NAACP civil rights group. \"We certainly don't think the judge has ruled erroneously,\" Navarre said, according to The Sun News .", "edit_actions": [{"type": "R", "before": "which then will be followed in", "after": "followed by", "start_char_pos": 43, "end_char_pos": 73, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "by", "after": "'s", "start_char_pos": 88, "end_char_pos": 90, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "parkings lots, while groups", "after": "parking lots. Groups", "start_char_pos": 224, "end_char_pos": 251, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "chome", "after": "chrome", "start_char_pos": 332, "end_char_pos": 337, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "alledged", "after": "alleged", "start_char_pos": 965, "end_char_pos": 973, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "\u2018\u2019", "after": null, "start_char_pos": 1491, "end_char_pos": 1493, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "\u2019\u2019", "after": null, "start_char_pos": 1507, "end_char_pos": 1509, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}], "sents_char_pos": [0, 100, 289, 385, 470, 582, 692, 893, 1029, 1231, 1403], "domain": "news"}
{"doc_id": "99987", "revision_depth": 2, "before_revision": "Dutch tennis player Raemon Sluiter played his last game as a professional at the ABN AMRO World Tennis Tournament in Rotterdam. Sluiter lost the first round game in two sets to Frenchman Michael Llodra. After the 82-minute game , Sluiter recieved a standing ovation from the spectators. \"I have never been of world-class,\" Sluiter said after the match, \"but this was a farewell worthy to a world-topper. One can only be grateful.\" Sluiter remarked that the results of the game supported his decision to retire, \"Today it was clear: I have to play at my best to compete with a sub-topper. In the past I could defeat guys like this and give the real world-class players a hard time. This was symbolic: I'm just not good enough anymore\", he said. Sluiter 's became a professional tennis player in 1996, his career highlights were mostly on home soil, reacing ATP finals in Amsterdam in 2000, and in Rotterdam and Amersfoort in 2003, as well as a semi-final appearance in Rosmalen. Another career highlight was reaching the semi-finals with the Dutch Davis Cup team in 2001. He announced his retirement in February 2008.", "after_revision": "Dutch tennis player Raemon Sluiter played his last match as a professional at the ABN AMRO World Tennis Tournament in Rotterdam. Sluiter lost the first round match in two sets to Frenchman Michael Llodra. After the 82-minute match , Sluiter recieved a standing ovation from the spectators. \"I have never been of world-class,\" Sluiter said after the match, \"but this was a farewell worthy to a world-topper. One can only be grateful.\" Sluiter remarked that the results of the match supported his decision to retire, \"Today it was clear: I have to play at my best to compete with a sub-topper. In the past I could defeat guys like this and give the real world-class players a hard time. This was symbolic: I'm just not good enough anymore\", he said. Sluiter became a professional tennis player in 1996, his career highlights were mostly on home soil, reacing ATP finals in Amsterdam in 2000, and in Rotterdam and Amersfoort in 2003, as well as a semi-final appearance in Rosmalen. Another career highlight was reaching the semi-finals with the Dutch Davis Cup team in 2001. He announced his retirement in February 2008.", "edit_actions": [{"type": "R", "before": "game", "after": "match", "start_char_pos": 51, "end_char_pos": 55, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "game", "after": "match", "start_char_pos": 157, "end_char_pos": 161, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "game", "after": "match", "start_char_pos": 223, "end_char_pos": 227, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "game", "after": "match", "start_char_pos": 472, "end_char_pos": 476, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "'s", "after": null, "start_char_pos": 752, "end_char_pos": 754, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 127, 202, 286, 403, 430, 587, 680, 743, 977, 1070], "domain": "news"}
{"doc_id": "100175", "revision_depth": 1, "before_revision": "In the Czech Republic The Czech Republic use the term Pta oblast (PO, bird area) for SPAs. There were declared 41 bird areas by the government directivesbetween 2004 and 2009. Pta oblasti nazen vldy They cover 9\\% of the state area.", "after_revision": "In the Czech Republic The Czech Republic uses the term Pta oblast (PO, bird area) for SPAs. Between 2004 and 2009, 41 bird areas were declared by government directives. Pta oblasti nazen vldy They cover 9\\% of the state area.", "edit_actions": [{"type": "R", "before": "use", "after": "uses", "start_char_pos": 41, "end_char_pos": 44, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "There were declared", "after": "Between 2004 and 2009,", "start_char_pos": 91, "end_char_pos": 110, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "by the government directivesbetween 2004 and 2009.", "after": "were declared by government directives.", "start_char_pos": 125, "end_char_pos": 175, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}], "sents_char_pos": [0, 90, 175], "domain": "wiki"}
{"doc_id": "10059597", "revision_depth": 1, "before_revision": " Global Positioning System (GPS) satellites broadcast microwave signals to enable GPS receivers on or near the Earth's surface to determine locationand time, and to derive velocity . The system is operated by the U.S. Department of Defense (DoD) for use by both the military and the general public.", "after_revision": "GPS signals are broadcast by Global Positioning System satellites to enable satellite navigation. Receivers on or near the Earth's surface can determine location, time, and velocity using this information . The system is operated by the U.S. Department of Defense (DoD) for use by both the military and the general public.", "edit_actions": [{"type": "A", "before": null, "after": "GPS signals are broadcast by", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "(GPS) satellites broadcast microwave signals to enable GPS receivers", "after": "satellites to enable satellite navigation. Receivers", "start_char_pos": 27, "end_char_pos": 95, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "to determine locationand", "after": "can determine location,", "start_char_pos": 127, "end_char_pos": 151, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "to derive velocity", "after": "velocity using this information", "start_char_pos": 162, "end_char_pos": 180, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 182], "domain": "wiki"}
{"doc_id": "10174927", "revision_depth": 1, "before_revision": "Expelled Asians in the Netherlands after leaving Uganda, 1972 In early August 1972, the President of Uganda, Idi Amin, ordered the expulsion of his country's Asian minority, giving them 90 days to leave the country. At the time of the expulsion, there were approximately 80,000 individuals of Indian descent (mostly Gujaratis) in Uganda, of whom 23,000 had their applications for citizenship both processed and accepted. Although the latter were ultimately exempted from the expulsion, many chose to leave voluntarily.. The expulsion took place against a backdrop of Indophobia in Uganda, with Amin accusing a minority of the Asian population of disloyalty, non-integration and commercial malpractice, claims  Indian leaders disputed. Amin defended the expulsion by arguing that he was \"giving Uganda back to ethnic Ugandans\". Many of the expellees were citizens of the United Kingdom and Colonies and 27, 200 subsequently emigrated to the United Kingdom. Of the other refugees who were accounted for, 6,000 went to Canada, 4,500 refugees ended up in India and 2,500 went to nearby Kenya or Pakistan. In total, some 5,655 firms, ranches, farms, and agricultural estates were reallocated, along with cars, homes and other household goods.", "after_revision": "Expelled Asians in the Netherlands after leaving Uganda, 1972 In early August 1972, the President of Uganda, Idi Amin, ordered the expulsion of his country's Asian minority, giving them 90 days to leave the country. At the time of the expulsion, there were about 80,000 individuals of Indian descent (mostly Gujaratis) in Uganda, of whom 23,000 had their applications for citizenship both processed and accepted. Although the latter were ultimately exempted from the expulsion, many chose to leave voluntarily.. The expulsion took place against a backdrop of Indophobia in Uganda, with Amin accusing a minority of the Asian population of disloyalty, nonintegration and commercial malpractice, claims that Indian leaders disputed. Amin defended the expulsion by arguing that he was \"giving Uganda back to ethnic Ugandans\". Many of those expelled were citizens of the United Kingdom and Colonies and 27, 2021 emigrated to the United Kingdom. Of the other refugees who were accounted for, 6,000 went to Canada, 4,500 refugees ended up in India and 2,500 went to nearby Kenya or Pakistan. In total, some 5,655 firms, ranches, farms, and agricultural estates were reallocated, along with cars, homes and other household goods.", "edit_actions": [{"type": "R", "before": "approximately", "after": "about", "start_char_pos": 257, "end_char_pos": 270, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": "non-integration", "after": "nonintegration", "start_char_pos": 658, "end_char_pos": 673, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "that", "start_char_pos": 709, "end_char_pos": 709, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "the expellees", "after": "those expelled", "start_char_pos": 835, "end_char_pos": 848, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "200 subsequently", "after": "2021", "start_char_pos": 906, "end_char_pos": 922, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 215, 420, 519, 734, 826, 955, 1100], "domain": "wiki"}
{"doc_id": "1021622", "revision_depth": 2, "before_revision": "Symptoms and signs The clinical condition denoted abulia was first described in 1838; however, since that time, a number of different  definitions have emerged , with some contradicting others .Vijayaraghavan, L., Krishnamoorthy, E. S., Brown, R. G., & Trimble, M. R. (2002). Abulia: A Delphi survey of British neurologists and psychiatrists. [Article]. Movement Disorders, 17(5), 1052-1057. Abulia has been described as a loss of drive, expression, loss of behavior and speech output,  slowing and prolonged speech latency, and reduction of spontaneous thought content and initiative.Jahanshahi, M., & Frith, C. D. (1998). Willed action and its impairments. [Review]. Cognitive Neuropsychology, 15(6-8), 483-533.. The clinical features most commonly associated with abulia are: Difficulty in initiating and sustaining purposeful movements Lack of spontaneous movement Reduced spontaneous speech Increased response-time to queries Passivity Reduced emotional responsiveness and spontaneity Reduced social interactions Reduced interest in usual pastimes A 2002 survey of two movement disorder experts, two neuropsychiatrists, and two rehabilitation experts  did not seem to shed any light on the matter of differentiating abulia from other DDMs. The experts used the terms apathy and abulia interchangeably and debated whether or not abulia was a discrete entity, or just a hazy gray area on a spectrum of more defined disorders. Four of the experts said abulia was a sign and a symptom, and the group was split on whether or not it was a syndrome. Another survey, which consisted of true and false questions about what abulia is distinct from, whether it is a sign, symptom, or syndrome, where lesions are present in cases of abulia, what diseases are commonly associated with abulia, and what current treatments are used for abulia, was sent to 15 neurologists and 10 psychiatrists. Most experts agreed that abulia is clinically distinct from depression, akinetic mutism, and alexithymia. However, only 32\\% believed abulia was different from apathy, while 44\\% said they were not different, and 24\\% were unsure. Yet again, there was disagreement about whether or not abulia is a sign, symptom, or syndrome.", "after_revision": "Symptoms and signs The clinical condition denoted abulia was first described in 1838; however, since that time, a number of different , some contradictory, definitions have emerged .Vijayaraghavan, L., Krishnamoorthy, E. S., Brown, R. G., & Trimble, M. R. (2002). Abulia: A Delphi survey of British neurologists and psychiatrists. [Article]. Movement Disorders, 17(5), 1052-1057. Abulia has been described as a loss of drive, expression, behavior and speech output, with slowing and prolonged speech latency, and reduction of spontaneous thought content and initiative.Jahanshahi, M., & Frith, C. D. (1998). Willed action and its impairments. [Review]. Cognitive Neuropsychology, 15(6-8), 483-533.. The clinical features most commonly associated with abulia are: Difficulty in initiating and sustaining purposeful movements Lack of spontaneous movement Reduced spontaneous speech Increased response-time to queries Passivity Reduced emotional responsiveness and spontaneity Reduced social interactions Reduced interest in usual pastimes A 2002 survey of two movement disorder experts, two neuropsychiatrists, and two rehabilitation experts , did not seem to shed any light on the matter of differentiating abulia from other DDMs. The experts used the terms \"apathy\" and \"abulia\" interchangeably and debated whether or not abulia was a discrete entity, or just a hazy gray area on a spectrum of more defined disorders. Four of the experts said abulia was a sign and a symptom, and the group was split on whether or not it was a syndrome. Another survey, which consisted of true and false questions about what abulia is distinct from, whether it is a sign, symptom, or syndrome, where lesions are present in cases of abulia, what diseases are commonly associated with abulia, and what current treatments are used for abulia, was sent to 15 neurologists and 10 psychiatrists. Most experts agreed that abulia is clinically distinct from depression, akinetic mutism, and alexithymia. However, only 32\\% believed abulia was different from apathy, while 44\\% said they were not different, and 24\\% were unsure. Yet again, there was disagreement about whether or not abulia is a sign, symptom, or syndrome.", "edit_actions": [{"type": "A", "before": null, "after": ", some contradictory,", "start_char_pos": 134, "end_char_pos": 134, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ", with some contradicting others", "after": null, "start_char_pos": 160, "end_char_pos": 192, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "loss of", "after": null, "start_char_pos": 450, "end_char_pos": 457, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "with", "start_char_pos": 486, "end_char_pos": 486, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1156, "end_char_pos": 1156, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "apathy and abulia", "after": "\"apathy\" and \"abulia\"", "start_char_pos": 1272, "end_char_pos": 1289, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 85, 275, 342, 353, 391, 585, 623, 658, 714, 778, 1244, 1428, 1547, 1883, 1989, 2114], "domain": "wiki"}
{"doc_id": "1056700", "revision_depth": 1, "before_revision": "An intimate relationship is an interpersonal relationship that involves physical or emotional intimacy . Although an intimate relationship is commonly a sexual relationship , it may also be a non-sexual relationship involving family, friends, or acquaintances. Emotional intimacy involves feelings of liking or loving one or more people, and may result in physical intimacy. Physical intimacy is characterized by romantic love, sexual activity, or other passionate attachment. These relationships play a central role in the overall human  experience.Miller, Rowland & Perlman, Daniel (2008). Intimate Relationships (5th ed.). McGraw-Hill. Humans have a general desire to belong and to love , which is usually satisfied within an intimate relationship.Perlman, D. (2007). The best of times, the worst of times: The place of close relationships in psychology and our daily lives. Canadian Psychology, 48, 718. Such relationships allow a social network for people to form strong emotional attachments.", "after_revision": "An intimate relationship is an interpersonal relationship that involves sexual intimacy, because romantic love is based on sex . Although an intimate relationship is commonly a sexual relationship Emotional intimacy involves feelings of sex or lustingone or more people, and may result in physical intimacy. Physical intimacy is characterized by sexual activity, or other sexual attachment. These relationships play a central role in the overall human sexual experience.Miller, Rowland & Perlman, Daniel (2008). Intimate Relationships (5th ed.). McGraw-Hill. Humans have a general desire to belong and to have sex , which is usually satisfied within a sexual relationship.Perlman, D. (2007). The best of times, the worst of times: The place of close relationships in psychology and our daily lives. Canadian Psychology, 48, 718. Such relationships allow a social network for people to form strong emotional attachments.", "edit_actions": [{"type": "R", "before": "physical or emotional intimacy", "after": "sexual intimacy, because romantic love is based on sex", "start_char_pos": 72, "end_char_pos": 102, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": ", it may also be a non-sexual relationship involving family, friends, or acquaintances.", "after": null, "start_char_pos": 173, "end_char_pos": 260, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "liking or loving one", "after": "sex or lustingone", "start_char_pos": 301, "end_char_pos": 321, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "romantic love,", "after": null, "start_char_pos": 413, "end_char_pos": 427, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "passionate", "after": "sexual", "start_char_pos": 454, "end_char_pos": 464, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "sexual", "start_char_pos": 538, "end_char_pos": 538, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "love", "after": "have sex", "start_char_pos": 685, "end_char_pos": 689, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "an intimate", "after": "a sexual", "start_char_pos": 726, "end_char_pos": 737, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 104, 260, 374, 476, 550, 591, 625, 638, 751, 770, 877, 907], "domain": "wiki"}
{"doc_id": "1056700", "revision_depth": 2, "before_revision": "An intimate relationship is an interpersonal relationship that involves sexual intimacy, because romantic love is based on sex . Although an intimate relationship is commonly a sexual relationship  Emotional intimacy involves feelings of sex or lustingone or more people, and may result in physical intimacy. Physical intimacy is characterized by  sexual activity, or other sexual attachment. These relationships play a central role in the overall human sexual experience.Miller, Rowland & Perlman, Daniel (2008). Intimate Relationships (5th ed.). McGraw-Hill. Humans have a general desire to belong and to have sex , which is usually satisfied within a sexual relationship.Perlman, D. (2007). The best of times, the worst of times: The place of close relationships in psychology and our daily lives. Canadian Psychology, 48, 7\u201318. Such relationships allow a social network for people to form strong emotional attachments.", "after_revision": "An intimate relationship is an interpersonal relationship that involves physical or emotional intimacy . Although an intimate relationship is commonly a sexual relationship , it may also be a non-sexual relationship involving family, friends, or acquaintances. Emotional intimacy involves feelings of liking or loving one or more people, and may result in physical intimacy. Physical intimacy is characterized by romantic love, sexual activity, or other passionate attachment. These relationships play a central role in the overall human experience.Miller, Rowland & Perlman, Daniel (2008). Intimate Relationships (5th ed.). McGraw-Hill. Humans have a general desire to belong and to love , which is usually satisfied within an intimate relationship.Perlman, D. (2007). The best of times, the worst of times: The place of close relationships in psychology and our daily lives. Canadian Psychology, 48, 7\u201318. Such relationships allow a social network for people to form strong emotional attachments.", "edit_actions": [{"type": "R", "before": "sexual intimacy, because romantic love is based on sex", "after": "physical or emotional intimacy", "start_char_pos": 72, "end_char_pos": 126, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ", it may also be a non-sexual relationship involving family, friends, or acquaintances.", "start_char_pos": 197, "end_char_pos": 197, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "sex or lustingone", "after": "liking or loving one", "start_char_pos": 238, "end_char_pos": 255, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "romantic love,", "start_char_pos": 347, "end_char_pos": 347, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "sexual", "after": "passionate", "start_char_pos": 374, "end_char_pos": 380, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "sexual", "after": null, "start_char_pos": 454, "end_char_pos": 460, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "have sex", "after": "love", "start_char_pos": 607, "end_char_pos": 615, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "a sexual", "after": "an intimate", "start_char_pos": 652, "end_char_pos": 660, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 128, 308, 392, 472, 513, 547, 560, 674, 693, 800, 831], "domain": "wiki"}
{"doc_id": "1115806", "revision_depth": 1, "before_revision": "ERG theory is a theory in psychology proposed by Clayton Alderfer. thumb Alderfer further developed Maslow's hierarchy of needs by categorizing the hierarchy into his ERG theory (Existence, Relatedness and Growth). The existence group is concerned with  providing the basic material existence requirements of humans. They include the items that Maslow considered to be physiological and safety needs. The second group of needs is those of relatedness the desire people have for maintaining important interpersonal relationships. These social and status desires require interaction with others if they are to be satisfied, and they align with Maslow's social need and the external component of Maslow's esteem classification. Finally, Alderfer isolates growth needs: an intrinsic desire for personal development. These include the intrinsic component from Maslow's esteem category and the characteristics included under self-actualization. Alderfer categorized the lower order needs (Physiological and Safety) into the Existence category. He fit Maslow's interpersonal love and esteem needs into the Relatedness category. The Growth category contained the self-actualization and self-esteem needs  . Alderfer also proposed a  regression theory to go along with the ERG theory . He said that when needs in a higher category are not met then individuals redouble the efforts invested in a lower category need . For example if self-actualization or self-esteem  is not met then individuals will invest more effort in the relatedness category in the hopes of achieving the higher need.Design in the Manufacturing Firm syllabus University of Washington Industrial Engineering course syllabus. Retrieved on 07-17-2011.", "after_revision": "ERG theory is a theory in psychology proposed by Clayton Alderfer. ERG theory. When needs in a category are satisfied, an individual will invest more efforts in the higher category. When needs in a category are frustrated, an individual will invest more efforts in the lower category. Alderfer further developed Maslow's hierarchy of needs by categorizing the hierarchy into his ERG theory (Existence, Relatedness and Growth). The existence category is concerned with the need for providing the basic material existence requirements of humans. The relatedness category is concerned about the desire for maintaining important interpersonal relationships. The growth category is concerned about the desire for personal development. These include the intrinsic component from Maslow's esteem category and the characteristics included under self-actualization. Alderfer categorized Maslow's physiological needs and Maslow's safety needs into the existence category, Maslow's social needs and Maslow's extrinsic component of self-esteem needs into the relatedness category, and Maslow's intrinsic component of self-esteem needs and Maslow's self-actualization needs into the growth category . Alderfer also proposed a progression and regression theory to go along with the ERG theory : he said that when needs in a lower category are satisfied, individuals invest more efforts in the higher category, and when needs in a higher category are frustrated, individuals invest more efforts in the lower category . For example if self-esteem or self-actualization is not met then individuals will invest more effort in the relatedness category in the hopes of achieving the higher need.Design in the Manufacturing Firm syllabus University of Washington Industrial Engineering course syllabus. Retrieved on 07-17-2011.", "edit_actions": [{"type": "R", "before": "thumb", "after": "ERG theory. When needs in a category are satisfied, an individual will invest more efforts in the higher category. When needs in a category are frustrated, an individual will invest more efforts in the lower category.", "start_char_pos": 67, "end_char_pos": 72, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "group", "after": "category", "start_char_pos": 229, "end_char_pos": 234, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "the need for", "start_char_pos": 253, "end_char_pos": 253, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "They include the items that Maslow considered to be physiological and safety needs. The second group of needs is those of relatedness the desire people have", "after": "The relatedness category is concerned about the desire", "start_char_pos": 317, "end_char_pos": 473, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "These social and status desires require interaction with others if they are to be satisfied, and they align with Maslow's social need and the external component of Maslow's esteem classification. Finally, Alderfer isolates growth needs: an intrinsic", "after": "The growth category is concerned about the", "start_char_pos": 529, "end_char_pos": 778, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Alderfer categorized the lower order needs (Physiological and Safety) into the Existence category. He fit", "after": "Alderfer categorized Maslow's physiological needs and Maslow's safety needs into the existence category,", "start_char_pos": 939, "end_char_pos": 1044, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "interpersonal love and esteem", "after": "social needs and Maslow's extrinsic component of self-esteem", "start_char_pos": 1054, "end_char_pos": 1083, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "Relatedness category. The Growth category contained the self-actualization and", "after": "relatedness category, and Maslow's intrinsic component of", "start_char_pos": 1099, "end_char_pos": 1177, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}, {"type": "A", "before": null, "after": "and Maslow's self-actualization needs into the growth category", "start_char_pos": 1196, "end_char_pos": 1196, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "progression and", "start_char_pos": 1224, "end_char_pos": 1224, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". He", "after": ": he", "start_char_pos": 1275, "end_char_pos": 1279, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "higher category are not met then individuals redouble the efforts invested in a lower category need", "after": "lower category are satisfied, individuals invest more efforts in the higher category, and when needs in a higher category are frustrated, individuals invest more efforts in the lower category", "start_char_pos": 1306, "end_char_pos": 1405, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "D", "before": "self-actualization or", "after": null, "start_char_pos": 1423, "end_char_pos": 1444, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "or self-actualization", "start_char_pos": 1457, "end_char_pos": 1457, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 214, 316, 400, 528, 724, 811, 938, 1037, 1120, 1198, 1276, 1407, 1580, 1686]}
{"doc_id": "1115806", "revision_depth": 2, "before_revision": "Alderfer categorized Maslow's physiological needs and Maslow's safety needs into the existence category, Maslow's social needs and Maslow's extrinsic component of self-esteem needs into the relatedness category, and Maslow's intrinsic component of self-esteem needs and Maslow's self-actualization needs into the growth category. Alderfer also proposed a progression and regression theory to go along with the ERG theory: he said that when needs in a lower category are satisfied, individuals invest more efforts in the higher category, and when needs in a higher category are frustrated, individuals invest more efforts in the lower category. For example if self-esteem or self-actualization is not met then individuals will invest more effort in the relatedness category in the hopes of achieving the higher need.Design in the Manufacturing Firm syllabus University of Washington Industrial Engineering course syllabus. Retrieved on 07-17-2011.", "after_revision": "Alderfer categorized Maslow's physiological needs and Maslow's safety needs into the existence category, Maslow's social needs and Maslow's extrinsic component of self-esteem needs into the relatedness category, and Maslow's intrinsic component of self-esteem needs and Maslow's self-actualization needs into the growth category. Alderfer also proposed a progression and regression theory to go along with the ERG theory: he said that when needs in a lower category are satisfied, an individual will invest more efforts in the higher category, and when needs in a higher category are frustrated, an individual will invest more efforts in the lower category. For example if self-esteem or self-actualization is not met then an individual will invest more effort in the relatedness category in the hopes of achieving the higher need.Design in the Manufacturing Firm syllabus University of Washington Industrial Engineering course syllabus. Retrieved on 07-17-2011.", "edit_actions": [{"type": "R", "before": "individuals", "after": "an individual will", "start_char_pos": 481, "end_char_pos": 492, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "individuals", "after": "an individual will", "start_char_pos": 589, "end_char_pos": 600, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "individuals", "after": "an individual", "start_char_pos": 709, "end_char_pos": 720, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 329, 643, 815, 921], "domain": "wiki"}
{"doc_id": "11164214", "revision_depth": 1, "before_revision": "The site is on a plain, at the foot of a basalt plateau, which forms a portion of the Hijaz mountains. The western and north-western portions of the site contain a water table that can be reached at a depth of . The setting is notable for its desert landscape, marked by sandstone outcroppings of various sizes and heights.", "after_revision": "The site is on a plain, at the foot of a basalt plateau, which forms a portion of the Hijaz mountains. Beneath the western and north-western parts of the site , the water table can be reached at a depth of . The setting is notable for its desert landscape, marked by sandstone outcroppings of various sizes and heights.", "edit_actions": [{"type": "R", "before": "The", "after": "Beneath the", "start_char_pos": 103, "end_char_pos": 106, "major_intent": "clarity", "raw_intents": ["clarity", "style", "others"]}, {"type": "R", "before": "portions", "after": "parts", "start_char_pos": 133, "end_char_pos": 141, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "contain a water table that", "after": ", the water table", "start_char_pos": 154, "end_char_pos": 180, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 102, 211], "domain": "wiki"}
{"doc_id": "1214033", "revision_depth": 1, "before_revision": "In the 19th century , the idea appeared in universities that a definitive history could be written of a major region of the world for a great span of time  in a similar manner to the way that an encyclopediawas written. The time period was subdivided into eras and one volume specified for each. Within each volume there would be a fixed number of topics. Either each volume would either be written by one historian on the faculty, or else each topic would be handled by a faculty member throughout the series of volumes, or perhaps another system of specialization would be prescribed. This procedure was similar to that undertaken on such campuses to produce encyclopedias of natural history, such as marine biology, for which different scholars would write about different phyla. Examples of the end result of this procedure include the series done by Cambridge on Greco-Roman history, and that of Oxford on British history, which may be found on the reserve stacks of many public libraries in the 21st century. What gives this concept of 'academic history' its own historicity, or \"cubbyhole in time\", superseded by progress, is that an academic history was intended to be definitive even though its subject matter, unlike the marine biology mentioned above, was not objective. When the volume on the Regency was published, for example, some may have thought that such would be the complete history of that era, and no one would need to do as much work in that field, because the best people with the best resources would already have written it down. Subsequent changes in scholarly perspective can alter that perception; for example the work of Lewis Namier on mid-18th century British politics caused one of the Oxford History volumes to appear outdated. It was not considered that entirely new viewpoints and methods would come into being, or that scholars would follow new threads of causality throughout stretches of time that differed from the canonical ones over a region which varied over time. And as each academic history was primarily a list of persons, places, things, and events, there was hardly any Marxian content to any of these projects. By the second half of the 20th , there weren't any more  academic histories. History is no longer subdivided  in such an assembly-line fashion with such an authoritative result expected. However, the project of globalization has brought with it the notion of writing a history that has no national center. All the projects above, allocated to the faculty of a university, had the viewpoint of their country or region in mind. These new histories are similar to the academic , in that they are large and done by many people by a similar process of allocation, but they do not have the same all-specifying concept of classification; instead, it is interrelation which is of concern.", "after_revision": "In the 19th century the idea appeared in universities that a definitive history could be written of a major region of the world for a great span of time - in a similar manner to the way in which people compiled an encyclopedia. The chosen time period was subdivided into eras and one volume specified for each. Within each volume there would be a fixed number of topics. Either each volume would be written by one historian on the faculty, or else each topic would be handled by a faculty member throughout the series of volumes, or perhaps another system of specialization would be prescribed. This procedure resembled that undertaken on such campuses to produce encyclopedias of natural history, such as marine biology, for which different scholars would write about different phyla. Examples of the end result of this procedure include the series produced by Cambridge on Greco-Roman history, and that of Oxford on British history, which may be found on the reserve stacks of many public libraries in the 21st century. What gives this concept of \"academic history\" its own historicity, or \"cubbyhole in time\", challenged by progress, is that an academic history was intended to be definitive even though its subject matter, unlike the marine biology mentioned above, was not objective. When the volume on the Regency was published, for example, some may have thought that such would be the complete history of that era, and no one would need to do as much work in that field, because the best people with the best resources would already have written it down. Subsequent changes in scholarly perspective can alter that perception; for example the work of Lewis Namier on mid-18th century British politics caused one of the Oxford History volumes to appear outdated. It was not considered that entirely new viewpoints and methods would come into being, or that scholars would follow new threads of causality throughout stretches of time that differed from the canonical ones over a region which varied over time. And as each academic history was primarily a list of persons, places, things, and events, there was hardly any Marxian content to any of these projects. By the second half of the 20th century there weren't any more new academic histories. Historians no longer subdivided their subject-matter in such an assembly-line fashion with such an authoritative result expected. However, the project of globalization has brought with it the notion of writing a history that has no national center. All the academic-history projects mentioned above, allocated to the faculty of a university, had the viewpoint of their country or region in mind. Some new global histories resemble the academic histories , in that they are large and done by many people using a similar process of allocation, but they do not have the same all-specifying concept of classification; instead, it is interrelation which is of concern.", "edit_actions": [{"type": "D", "before": ",", "after": null, "start_char_pos": 20, "end_char_pos": 21, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "-", "start_char_pos": 155, "end_char_pos": 155, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "that an encyclopediawas written. The", "after": "in which people compiled an encyclopedia. The chosen", "start_char_pos": 187, "end_char_pos": 223, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "D", "before": "either", "after": null, "start_char_pos": 381, "end_char_pos": 387, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "was similar to", "after": "resembled", "start_char_pos": 602, "end_char_pos": 616, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "done", "after": "produced", "start_char_pos": 847, "end_char_pos": 851, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "'academic history'", "after": "\"academic history\"", "start_char_pos": 1042, "end_char_pos": 1060, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "superseded", "after": "challenged", "start_char_pos": 1106, "end_char_pos": 1116, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ",", "after": "century", "start_char_pos": 2192, "end_char_pos": 2193, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "new", "start_char_pos": 2217, "end_char_pos": 2217, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "History is", "after": "Historians", "start_char_pos": 2238, "end_char_pos": 2248, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "their subject-matter", "start_char_pos": 2270, "end_char_pos": 2270, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "projects", "after": "academic-history projects mentioned", "start_char_pos": 2475, "end_char_pos": 2483, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "These new", "after": "Some new global histories resemble the academic", "start_char_pos": 2587, "end_char_pos": 2596, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "D", "before": "are similar to the academic", "after": null, "start_char_pos": 2607, "end_char_pos": 2634, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "by", "after": "using", "start_char_pos": 2684, "end_char_pos": 2686, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 219, 295, 355, 586, 782, 1014, 1281, 1555, 1626, 1761, 2007, 2160, 2237, 2347, 2466, 2586, 2791], "domain": "wiki"}
{"doc_id": "128608", "revision_depth": 1, "before_revision": "The world's population is around 7,800,000,000U.S. & World Population Clocks. Census.gov. Retrieved on December 10, 2011. and Earth's total area (including land and water) is 510,000,000km2 (197,000,000 sq. mi.).World. CIA World Handbook Therefore,  the worldwide human population density is around 7,500,000,000 510,000,000 = 14.7 per km2 (38 per sq. mi.). If only the Earth's land area of 150,000,000km2 (58,000,000 sq. mi.) is taken into account, then human population density is 50 per km2 (129 per sq. mi.). This includes all continental and island land area, including Antarctica. If Antarctica is also excluded, then population density rises to over 55 people per km2 (over 142 per sq. mi.). However, over half of the Earth's land mass consists of areas inhospitable to human habitation, such as deserts and high mountains, and population tends to cluster around seaports and fresh-water sources. Thus , additional criteria are needed to make simple population density values  useful. In comparison, based on a world population of 7.8 billion, the world's inhabitants,  as a loose crowd taking up almost 1 m2 (10 sq. ft) per person (cf. Jacobs Method), would occupy a space a little larger than Delaware's land area.", "after_revision": "The world's population is around 7,800,000,000U.S. & World Population Clocks. Census.gov. Retrieved on December 10, 2011. and Earth's total area (including land and water) is 510,000,000km2 (197,000,000 sq. mi.).World. CIA World Handbook Therefore, from this very crude type of calculation, the worldwide human population density is approximately 7,500,000,000 510,000,000 = 14.7 per km2 (38 per sq. mi.). However, if only the Earth's land area of 150,000,000km2 (58,000,000 sq. mi.) is taken into account, then human population density is 50 per km2 (129 per sq. mi.). This includes all continental and island land area, including Antarctica. Furthermore, if Antarctica is also excluded, then population density rises to over 55 people per km2 (over 142 per sq. mi.). However, over half of the Earth's land mass consists of areas inhospitable to human habitation, such as deserts and high mountains, and population tends to cluster around seaports and fresh-water sources. Therefore , additional criteria are needed to make simple population density values meaningful and useful. In comparison, based on a world population of 7.8 billion, the world's inhabitants, if conceptualized as a loose crowd occupying just under 1 m2 (10 sq. ft) per person (cf. Jacobs Method), would occupy a space a little larger than Delaware's land area.", "edit_actions": [{"type": "A", "before": null, "after": "from this very crude type of calculation,", "start_char_pos": 249, "end_char_pos": 249, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "around", "after": "approximately", "start_char_pos": 292, "end_char_pos": 298, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "If", "after": "However, if", "start_char_pos": 358, "end_char_pos": 360, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "style"]}, {"type": "R", "before": "If", "after": "Furthermore, if", "start_char_pos": 587, "end_char_pos": 589, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "style"]}, {"type": "R", "before": "Thus", "after": "Therefore", "start_char_pos": 904, "end_char_pos": 908, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "meaningful and", "start_char_pos": 983, "end_char_pos": 983, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "if conceptualized", "start_char_pos": 1076, "end_char_pos": 1076, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "taking up almost", "after": "occupying just under", "start_char_pos": 1094, "end_char_pos": 1110, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 77, 89, 212, 357, 512, 586, 698, 903, 991], "domain": "wiki"}
{"doc_id": "12940645", "revision_depth": 1, "before_revision": "Mor Themotheos Thomas has a significant role in the development of the present Jacobite Syrian Orthodox Church in India, especially in the outside dioceses. Mor Themotheos ordained 6 present metropolitans of the church into the priesthood \" Kassiso\". No other bishop in the Jacobite church has that record as of now. Priests ordained by Mor Themotheos serving as Bishops are Mor Osthatheos Pathrose (Bangalore), Mor Eusebious Kuriakose (Delhi), Mor Coorilose Geevargese (Niranam), Mor Phelexinos Zacharias (Idukki), Mor Alexandrios Thomas (Mumbai), and Mor Thimotheos Matthew  Mor Themotheos is the founder of a pre-marital counseling program in the Jacobite church. First, he started this in his diocese, and later it was added to the whole church. Mor Themotheos served as the President and resident metropolitan of M.S.O.T seminary for 9 years ( 19912000) and a faculty of the seminary.Later on, Dr. Mor Theophilose Kuriakose succeeded him.  Mor Themotheos is presently the \"Secretary\" to the Holy Episcopal Synod of the Jacobite Syrian Orthodox Church  by the regional synod  on 2018 December held at Puthencruz .", "after_revision": "Mor Themotheos Thomas has significant role in the development of the present Jacobite Syrian Orthodox Church in India, especially in the outside dioceses. The \"Greater Indian Outside Arch diocese\" under his jurisdiction is divided into 4 separate dioceses called Mumbai, Bangalore, Delhi and Mylapore (Chennai) which are administered by different metropolitans now. The efforts and labor of Mor Themotheos significantly helped the present growth of outside Kerala Jacobite dioceses into a worth position. Mor Themotheos ordained 6 present metropolitans of the church into the priesthood \" Kassiso\". No other bishop in the Jacobite church has that record as of now. Priests ordained by Mor Themotheos serving as Bishops are Mor Osthatheos Pathrose (Bangalore), Mor Eusebious Kuriakose (Delhi), Mor Coorilose Geevargese (Niranam), Mor Phelexinos Zacharias (Idukki), Mor Alexandrios Thomas (Mumbai), and Mor Thimotheos Matthew (Former Patriarchal Secretary). Mor Themotheos is the founder of pre-marital counseling programs in Jacobite Syrian church. First, he started this in his diocese, and later on it was implemented to the whole church. Mor Themotheos served as the President and Resident Metropolitan of M.S.O.T seminary for 9 years ( 19912000) .Later Dr. Mor Theophilose Kuriakose succeeded him. Mor Themotheos served as a Professor of MSOT Seminary as well. He has very well depth knowledge about the Christian church history and proved his gifted talent as a well known scholar in the Patristic studies. He is very keen into the missionary activities of the Jacobite Church. \"Mor Gregorios Mission\" under his leadership is an emerging movement in the current mission field of Jacobite Syrian Church. Mor Themotheos is elected as the \"Secretary\" to the Holy Episcopal Synod of the Jacobite Syrian Orthodox Church in India by the regional synod of Church held on 2018 December .", "edit_actions": [{"type": "D", "before": "a", "after": null, "start_char_pos": 26, "end_char_pos": 27, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "R", "before": "Mor Themotheos", "after": "The \"Greater Indian Outside Arch diocese\" under his jurisdiction is divided into 4 separate dioceses called Mumbai, Bangalore, Delhi and Mylapore (Chennai) which are administered by different metropolitans now. The efforts and labor of Mor Themotheos significantly helped the present growth of outside Kerala Jacobite dioceses into a worth position. Mor Themotheos", "start_char_pos": 157, "end_char_pos": 171, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "(Former Patriarchal Secretary).", "start_char_pos": 576, "end_char_pos": 576, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "a", "after": null, "start_char_pos": 610, "end_char_pos": 611, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "R", "before": "program in the Jacobite", "after": "programs in Jacobite Syrian", "start_char_pos": 635, "end_char_pos": 658, "major_intent": "fluency", "raw_intents": ["meaning-changed", "fluency", "fluency"]}, {"type": "R", "before": "it was added", "after": "on it was implemented", "start_char_pos": 716, "end_char_pos": 728, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "resident metropolitan", "after": "Resident Metropolitan", "start_char_pos": 793, "end_char_pos": 814, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and a faculty of the seminary.Later on,", "after": ".Later", "start_char_pos": 859, "end_char_pos": 898, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Mor Themotheos served as a Professor of MSOT Seminary as well. He has very well depth knowledge about the Christian church history and proved his gifted talent as a well known scholar in the Patristic studies. He is very keen into the missionary activities of the Jacobite Church. \"Mor Gregorios Mission\" under his leadership is an emerging movement in the current mission field of Jacobite Syrian Church.", "start_char_pos": 944, "end_char_pos": 944, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "presently", "after": "elected as", "start_char_pos": 963, "end_char_pos": 972, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "in India", "start_char_pos": 1056, "end_char_pos": 1056, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "of Church held", "start_char_pos": 1079, "end_char_pos": 1079, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "held at Puthencruz", "after": null, "start_char_pos": 1097, "end_char_pos": 1115, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 156, 250, 316, 666, 749, 889, 943], "domain": "wiki"}
{"doc_id": "13299", "revision_depth": 1, "before_revision": "During this period, Spain was involved in all major European wars, including the Italian Wars, the Eighty Years' War, and the Thirty Years' War. In the later 17th century Spanish power began to decline, after the death of the last Habsburg . The former Spanish Empire overseas quickly disintegrated with the Spanish American wars of independence. Only Cuba and the Philippines and some small islands were left; they revolted and the United States acquired ownership (or control, in the case of Cuba) after the SpanishAmerican War of 1898. A tenuous balance between liberal and conservative forces was struck in the establishment of a constitutional monarchy in Spain during the Borbonic restoration; this period began in 1874 and ended in 1931. The Liberal Party ( Prxedes Mateo Sagasta) and Conservative Party (Antonio Cnovas del Castillo) fought for and won short-lived control without any being sufficiently strong to bring about lasting stability. They were alternately in power. The Restoration began with Alfonso XII and the Regency of Maria Christina (18741898). Alfonso XII died aged 27 in 1885, and was succeeded by his unborn son, who became Alfonso XIII (1902-1923). Then came the dictatorship of General Primo de Rivera (1923-1930). Opposition to his regime was so great that Alfonso XIII stopped supporting him and forced him to resign in January 1930.Shlomo Ben-Ami, \"The Dictatorship of Primo de Rivera: A Political Reassessment,\" Journal of Contemporary History, Jan 1977, Vol. 12 Issue 1, pp 6584 In 1931, following a victory by the left, the Popular Front, in municipal elections, Alfonso XIII left Spain and the democratic republic was proclaimed in Spain. The Conservative Party disappeared shortly after the proclamation of the Republic in 1931.Diccionario de historia de Espaa. Alvar Ezquerra, Jaime., Contreras, Jaime. Madrid: Istmo. 2001. . OCLC 50693588. Five years later the country descended into the Spanish Civil War between the Republican and the Nationalist factions.", "after_revision": "During this period, Spain was involved in all major European wars, including the Italian Wars, the Eighty Years' War, and the Thirty Years' War. Spanish power declined in the latter part of the 17th century . In the early part of the 19th century, most of the former Spanish Empire overseas disintegrated with the Spanish American wars of independence. Only Cuba and the Philippines and some small islands were left; they revolted and the United States acquired ownership (or control, in the case of Cuba) after the SpanishAmerican War of 1898. A tenuous balance between liberal and conservative forces was struck in the establishment of a constitutional monarchy in Spain during the Borbonic restoration; this period began in 1874 and ended in 1931. The Liberal Party ( Prxedes Mateo Sagasta) and Conservative Party (Antonio Cnovas del Castillo) fought for and won short-lived control without any being sufficiently strong to bring about lasting stability. They were alternately in power. The Restoration began with Alfonso XII and the Regency of Maria Christina (18741898). Alfonso XII died aged 27 in 1885, and was succeeded by his unborn son, who became Alfonso XIII (1902-1923). Then came the dictatorship of General Primo de Rivera (1923-1930). Opposition to his regime was so great that Alfonso XIII stopped supporting him and forced him to resign in January 1930.Shlomo Ben-Ami, \"The Dictatorship of Primo de Rivera: A Political Reassessment,\" Journal of Contemporary History, Jan 1977, Vol. 12 Issue 1, pp 6584 In 1931, following a victory by the left, the Popular Front, in municipal elections, Alfonso XIII left Spain and the democratic republic was proclaimed in Spain. The Conservative Party disappeared shortly after the proclamation of the Republic in 1931.Diccionario de historia de Espaa. Alvar Ezquerra, Jaime., Contreras, Jaime. Madrid: Istmo. 2001. . OCLC 50693588. Five years later the country descended into the Spanish Civil War between the Republican and the Nationalist factions.", "edit_actions": [{"type": "R", "before": "In the later", "after": "Spanish power declined in the latter part of the", "start_char_pos": 145, "end_char_pos": 157, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "century Spanish power began to decline, after the death of the last Habsburg", "after": "century", "start_char_pos": 163, "end_char_pos": 239, "major_intent": "coherence", "raw_intents": ["coherence", "style", "coherence"]}, {"type": "R", "before": "The", "after": "In the early part of the 19th century, most of the", "start_char_pos": 242, "end_char_pos": 245, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "quickly", "after": null, "start_char_pos": 277, "end_char_pos": 284, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 144, 241, 346, 410, 538, 699, 744, 951, 983, 1069, 1177, 1244, 1365, 1675, 1766, 1799, 1841, 1856, 1879], "domain": "wiki"}
{"doc_id": "13588294", "revision_depth": 1, "before_revision": "alt=Building in Texas where recovery groups have meeting|Clubhouse in Texas where recovery groups have meetings. Drug addiction recovery groups are voluntary associations of people who share a common desire to overcome their drug addiction. Different groups use different methods, ranging from completely secular to explicitly spiritual. Some programs may advocate a reduction in the use of illegal drugs rather than outright abstention, although this is  typically not a sustainable treatment plan in the long term. One survey of members  found active involvement in any addiction recovery group correlates with higher chances of maintaining sobriety. The survey found group participation increased when the individual members' beliefs matched those of their primary support group (often people will be members of multiple addiction recovery groups). Analysis of the survey results found a significant positive correlation between the religiosity of members and their participation in twelve-step addiction recovery groups and SMART Recovery, although the correlation factor was three times smaller for SMART Recovery than for the twelve-step addiction recovery groups. Religiosity was inversely related to participation in Secular Organizations for Sobriety. A survey of a cross-sectional sample of clinicians working in outpatient facilities (selected from the SAMHSA On-line Treatment Facility Locator) found that clinicians only referring clients to twelve-step groups for treatment were more likely than those referring their clients to twelve-step groups and \"twelve-step alternatives\" to believe less strongly in the effectiveness of Cognitive Behavioral and psychodynamic-oriented therapy, and were likely to be unfamiliar with twelve-step alternatives. A logistic regression of clinician's knowledge and awareness of Cognitive Behavioral Therapy effectiveness and preference for the twelve-step model was correlated with referring exclusively to twelve-step groups.", "after_revision": "alt=Building in Texas where recovery groups have meeting|Clubhouse in Texas where recovery groups have meetings. Drug addiction recovery groups are voluntary associations of people who share a common desire to overcome their drug addiction. Different groups use different methods, ranging from completely secular to explicitly spiritual. Some programs may advocate a reduction in the use of illegal drugs rather than outright abstention, although this is for other programs typically not a sustainable treatment plan in the long term. One survey of members who found active involvement in any addiction recovery group correlates with higher chances of maintaining sobriety. The survey found group participation increased when the individual members' beliefs matched those of their primary support group (often people will be members of multiple addiction recovery groups). Analysis of the survey results found a significant positive correlation between the religiosity of members and their participation in twelve-step programs (self named \"spiritual addiction recovery programs\" and to a lesser level in non religious SMART Recovery groups, the correlation factor being three times smaller for SMART Recovery than for the twelve-step addiction recovery groups. Religiosity was inversely related to participation in Secular Organizations for Sobriety. A survey of a cross-sectional sample of clinicians working in outpatient facilities (selected from the SAMHSA On-line Treatment Facility Locator) found that clinicians only referring clients to twelve-step groups were more likely than those referring their clients to twelve-step groups and \"twelve-step alternatives\" to believe less strongly in the effectiveness of Cognitive Behavioral and psychodynamic-oriented therapy, and were likely to be unfamiliar with twelve-step alternatives. A logistic regression of clinician's knowledge and awareness of Cognitive Behavioral Therapy effectiveness and preference for the twelve-step model was correlated with referring exclusively to twelve-step groups.", "edit_actions": [{"type": "A", "before": null, "after": "for other programs", "start_char_pos": 455, "end_char_pos": 455, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "who", "start_char_pos": 539, "end_char_pos": 539, "major_intent": "others", "raw_intents": ["others", "others", "coherence"]}, {"type": "R", "before": "addiction recovery groups and SMART Recovery, although", "after": "programs (self named \"spiritual addiction recovery programs\" and to a lesser level in non religious SMART Recovery groups,", "start_char_pos": 998, "end_char_pos": 1052, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "was", "after": "being", "start_char_pos": 1076, "end_char_pos": 1079, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "D", "before": "for treatment", "after": null, "start_char_pos": 1474, "end_char_pos": 1487, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 112, 240, 337, 516, 652, 851, 1170, 1260, 1406, 1762], "domain": "wiki"}
{"doc_id": "14920509", "revision_depth": 1, "before_revision": "Markov's principle, named after Andrey Markov Jr, is a statement in computability theory that  is logical validity classically  not in intuitionistic constructive mathematics. Many particular instances of the principle are nevertheless provable there as well. There are many equivalent formulations of Markov's principle, as listed below. In computability theory In the language of computability theory, Markov's principle is a formal expression of the claim that if it is impossible that an algorithm does not terminate, then it does terminate. This is equivalent to the claim that if a set and its complement are both computably enumerable, then the set is decidable.   In constructive analysis It is equivalent, in the language of real analysis, to the following principles: Realizability If constructive arithmetic is translated using realizability into a classical meta-theory that proves the -consistency of the relevant classical theory (for example, Peano Arithmetic if we are studying Heyting arithmetic), then Markov's principle is justified: a realizer is the constant function that takes a realization that is not everywhere false to the unbounded search that successively checks if is true. If is not everywhere false, then by -consistency there must be a term for which holds, and each term will be checked by the search eventually. If however does not hold anywhere, then the domain of the constant function must be empty, so although the search does not halt it still holds vacuously that the function is a realizer. By the Law of the Excluded Middle (in our classical metatheory), must either hold nowhere or not hold nowhere, therefore this constant function is a realizer. If instead the realizability interpretation is used in a constructive meta-theory, then it is not justified. Indeed, for first-order arithmetic, Markov's principle exactly captures the difference between a constructive and classical meta-theory. Specifically, a statement is provable in Heyting arithmetic with Extended Church's thesis if and only if there is a number that provably realizes it in Heyting arithmetic; and it is provable in Heyting arithmetic with Extended Church's thesis and Markov's principle if and only if there is a number that provably realizes it in Peano arithmetic .", "after_revision": "Markov's principle, named after Andrey Markov Jr, is a conditional existence statement for which there are many equivalent formulations, as discussed below. The principle is logical validity classically , but not in intuitionistic constructive mathematics. However, many particular instances of it are nevertheless provable in a constructive context as well. In computability theory In the language of computability theory, Markov's principle is a formal expression of the claim that if it is impossible that an algorithm does not terminate, then it does terminate. This is equivalent to the claim that if a set and its complement are both computably enumerable, then the set is decidable. Realizability If constructive arithmetic is translated using realizability into a classical meta-theory that proves the -consistency of the relevant classical theory (for example, Peano Arithmetic if we are studying Heyting arithmetic), then Markov's principle is justified: a realizer is the constant function that takes a realization that is not everywhere false to the unbounded search that successively checks if is true. If is not everywhere false, then by -consistency there must be a term for which holds, and each term will be checked by the search eventually. If however does not hold anywhere, then the domain of the constant function must be empty, so although the search does not halt it still holds vacuously that the function is a realizer. By the Law of the Excluded Middle (in our classical metatheory), must either hold nowhere or not hold nowhere, therefore this constant function is a realizer. If instead the realizability interpretation is used in a constructive meta-theory, then it is not justified. Indeed, for first-order arithmetic, Markov's principle exactly captures the difference between a constructive and classical meta-theory. Specifically, a statement is provable in Heyting arithmetic with Extended Church's thesis if and only if there is a number that provably realizes it in Heyting arithmetic; and it is provable in Heyting arithmetic with Extended Church's thesis and Markov's principle if and only if there is a number that provably realizes it in Peano arithmetic. In constructive analysis It is equivalent, in the language of real analysis, to the following principles: History The principle was first studied and adopted by the Russian school of constructivism, together with | choice principles and often together with a realizability perspective on the notion of mathematical function .", "edit_actions": [{"type": "R", "before": "statement in computability theory that", "after": "conditional existence statement for which there are many equivalent formulations, as discussed below.", "start_char_pos": 55, "end_char_pos": 93, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "The principle", "start_char_pos": 94, "end_char_pos": 94, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": ", but", "start_char_pos": 127, "end_char_pos": 127, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}, {"type": "R", "before": "Many", "after": "However, many", "start_char_pos": 176, "end_char_pos": 180, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}, {"type": "R", "before": "the principle", "after": "it", "start_char_pos": 205, "end_char_pos": 218, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "there", "after": "in a constructive context", "start_char_pos": 245, "end_char_pos": 250, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "There are many equivalent formulations of Markov's principle, as listed below.", "after": null, "start_char_pos": 260, "end_char_pos": 338, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "Realizability If constructive arithmetic is translated using realizability into a classical meta-theory that proves the -consistency of the relevant classical theory (for example, Peano Arithmetic if we are studying Heyting arithmetic), then Markov's principle is justified: a realizer is the constant function that takes a realization that is not everywhere false to the unbounded search that successively checks if is true. If is not everywhere false, then by -consistency there must be a term for which holds, and each term will be checked by the search eventually. If however does not hold anywhere, then the domain of the constant function must be empty, so although the search does not halt it still holds vacuously that the function is a realizer. By the Law of the Excluded Middle (in our classical metatheory), must either hold nowhere or not hold nowhere, therefore this constant function is a realizer.", "start_char_pos": 670, "end_char_pos": 670, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "A", "before": null, "after": "If instead the realizability interpretation is used in a constructive meta-theory, then it is not justified. Indeed, for first-order arithmetic, Markov's principle exactly captures the difference between a constructive and classical meta-theory. Specifically, a statement is provable in Heyting arithmetic with Extended Church's thesis if and only if there is a number that provably realizes it in Heyting arithmetic; and it is provable in Heyting arithmetic with Extended Church's thesis and Markov's principle if and only if there is a number that provably realizes it in Peano arithmetic.", "start_char_pos": 671, "end_char_pos": 671, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "Realizability If constructive arithmetic is translated using realizability into a classical meta-theory that proves the -consistency of the relevant classical theory (for example, Peano Arithmetic if we are studying Heyting arithmetic), then Markov's principle is justified: a realizer is the constant function that takes a realization that is not everywhere false to the unbounded search that successively checks if is true. If is not everywhere false, then by -consistency there must be a term for which holds, and each term will be checked by the search eventually. If however does not hold anywhere, then the domain of the constant function must be empty, so although the search does not halt it still holds vacuously that the function is a realizer. By the Law of the Excluded Middle (in our classical metatheory), must either hold nowhere or not hold nowhere, therefore this constant function is a realizer.", "after": null, "start_char_pos": 778, "end_char_pos": 1691, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "If instead the realizability interpretation is used in a constructive meta-theory, then it is not justified. Indeed, for first-order arithmetic, Markov's principle exactly captures the difference between a constructive and classical meta-theory. Specifically, a statement is provable in Heyting arithmetic with Extended Church's thesis if and only if there is a number that provably realizes it in Heyting arithmetic; and it is provable in Heyting arithmetic with Extended Church's thesis and Markov's principle if and only if there is a number that provably realizes it in Peano arithmetic", "after": "History The principle was first studied and adopted by the Russian school of constructivism, together with | choice principles and often together with a realizability perspective on the notion of mathematical function", "start_char_pos": 1692, "end_char_pos": 2282, "major_intent": "coherence", "raw_intents": ["coherence", "others", "clarity"]}], "sents_char_pos": [0, 175, 259, 338, 545, 669, 899, 1203, 1346, 1532, 1691, 1800, 1937, 2109], "domain": "wiki"}
{"doc_id": "14920509", "revision_depth": 2, "before_revision": " In computability theory In the language of computability theory, Markov's principle is a formal expression of the claim that if it is impossible that an algorithm does not terminate, then it does terminate. This is equivalent to the claim that if a set and its complement are both computably enumerable, then the set is decidable. In intuitionistic logic A predicate P over some domain is called decidable if for every x in the domain, either P(x) is true, or P(x) is not true , which is not always the case constructively. In predicate logic, if P is a predicate  over the natural numbers, it is expressed as : That is, if P is decidable, and it cannot be false for every natural number n, then it is true for somen . This form can be justified by Brouwer's continuity principles, whereas the stronger form contradicts them. Thus it can be derived from intuitionistic, realizability, and classical reasoning, in each case for different reasons, but this principle is not valid in the general constructive sense of Bishop.Ulrich Kohlenbach, \"On weak Markov's principle\". Mathematical Logic Quarterly (2002), vol 48, issue S1, pp. 5965. History The principle was first studied and adopted by the Russian school of constructivism, together with choice principles and often together with a realizability perspective on the notion of mathematical function.", "after_revision": "History The principle was first studied and adopted by the Russian school of constructivism, together with choice principles and often with a realizability perspective on the notion of mathematical function. In computability theory In the language of computability theory, Markov's principle is a formal expression of the claim that if it is impossible that an algorithm does not terminate, then it does terminate. This is equivalent to the claim that if a set and its complement are both computably enumerable, then the set is decidable. In Intuitionistic logic In predicate logic, a predicate P over some domain is called decidable if for every x in the domain, either P(x) is true, or P(x) is not true . This is not trivially true constructively. For a decidable predicate P over the natural numbers, Markov's principle then reads : That is, if P cannot be false for all natural numbers n, then it is true for some n . This form can be justified by Brouwer's continuity principles, whereas the stronger form contradicts them. Thus it can be derived from intuitionistic, realizability, and classical reasoning, in each case for different reasons, but this principle is not valid in the general constructive sense of Bishop.Ulrich Kohlenbach, \"On weak Markov's principle\". Mathematical Logic Quarterly (2002), vol 48, issue S1, pp. 5965. ", "edit_actions": [{"type": "A", "before": null, "after": "History The principle was first studied and adopted by the Russian school of constructivism, together with choice principles and often with a realizability perspective on the notion of mathematical function.", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "intuitionistic logic A predicate", "after": "Intuitionistic logic In predicate logic, a predicate", "start_char_pos": 335, "end_char_pos": 367, "major_intent": "fluency", "raw_intents": ["fluency", "others", "style"]}, {"type": "R", "before": ", which is not always the case constructively. In predicate logic, if P is a predicate", "after": ". This is not trivially true constructively.", "start_char_pos": 478, "end_char_pos": 564, "major_intent": "clarity", "raw_intents": ["others", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "For a decidable predicate P", "start_char_pos": 565, "end_char_pos": 565, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "it is expressed as", "after": "Markov's principle then reads", "start_char_pos": 592, "end_char_pos": 610, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "is decidable, and it", "after": null, "start_char_pos": 627, "end_char_pos": 647, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "every natural number", "after": "all natural numbers", "start_char_pos": 668, "end_char_pos": 688, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "somen", "after": "some n", "start_char_pos": 712, "end_char_pos": 717, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "D", "before": "History The principle was first studied and adopted by the Russian school of constructivism, together with choice principles and often together with a realizability perspective on the notion of mathematical function.", "after": null, "start_char_pos": 1137, "end_char_pos": 1353, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 207, 331, 524, 719, 826, 1023, 1071, 1136], "domain": "wiki"}
{"doc_id": "1556660", "revision_depth": 2, "before_revision": "Gray death is a slang term used for a potent mixture of synthetic opioids. Samples have been found to contain the designer drug U-47700, heroin, fentanyl, and carfentanil . It is taken by injection, smoking, snorting, or oral ingestion. It was first discovered in the United States and was thought to be a unique chemical compound before being identified as a mixture of drugs. In addition to containing powerful opioids, what makes gray death an extremely dangerous drug is the fact that the constituent drugs' identities, their overall ratios in the product, and their uneven distribution within the productare completely unknown to the user. This means that the user cannot predict what dose to safely take; even two identical doses by mass may contain more or less of an unknown substance due to uneven physical distribution, as seen with heroin and fentanyl mixes. Opioid addicts tend to be more susceptible to the drug . They may think they are buying heroin, unaware that it is actually gray death. Reversing a gray death overdose frequently requires multiple doses of naloxone , whereas a typical overdose of heroin (which is much less potent) typically needs only one dose. This is consistent with the difficulty of overdose reversal seen with high-affinity opioids in the fentanyl chemical family or with buprenorphine. The greater affinity of these substances for the -opioid receptor causes less naloxoneto bind to receptors, increasing the dosage necessary to counteract the resulting respiratory depression. The difficulty in reversing an overdose of gray death is likely due to the presence of a fentanyl analogue, which reporting seems to corroborate as of December 2018.", "after_revision": "Gray death is a slang term which refers to a potent mixture of synthetic opioids. Samples have been found to contain heroin, fentanyl, carfentanil, and the designer drug U-47700 . It is taken by injection, smoking, snorting, or oral ingestion. The substance first appeared in America and was thought to be a unique chemical compound before being identified as a mixture of drugs. As with other illicit narcotics, gray death carries a higher risk of serious adverse effects than prescribed opioids due to the unknown and inconsistent composition of the product . As such, even experienced opioid users risk serious injury or death when taking this drug mixture. Reversing a gray death overdose may require multiple doses of naloxone . By contrast, an overdose from morphine or from high-purity heroin would ordinarily need only one dose. This difficulty is regularly encountered when treating overdoses of high-affinity opioids in the fentanyl chemical family or with buprenorphine. The greater affinity of these substances for the -opioid receptor impedes the activity of naloxone, which is an antagonist at the receptor. Increasing the dosage of naloxone or its frequency of administration may be required to counteract respiratory depression. ", "edit_actions": [{"type": "R", "before": "used for", "after": "which refers to", "start_char_pos": 27, "end_char_pos": 35, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "the designer drug U-47700,", "after": null, "start_char_pos": 110, "end_char_pos": 136, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "and carfentanil", "after": "carfentanil, and the designer drug U-47700", "start_char_pos": 155, "end_char_pos": 170, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "It was first discovered in the United States", "after": "The substance first appeared in America", "start_char_pos": 237, "end_char_pos": 281, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "In addition to containing powerful opioids, what makes gray death an extremely dangerous drug is the fact that the constituent drugs' identities, their overall ratios in the product, and their uneven distribution within the productare completely unknown to the user. This means that the user cannot predict what dose to safely take; even two identical doses by mass may contain more or less of an unknown substance due to uneven physical distribution, as seen with heroin and fentanyl mixes. Opioid addicts tend to be more susceptible to the drug", "after": "As with other illicit narcotics, gray death carries a higher risk of serious adverse effects than prescribed opioids due to the unknown and inconsistent composition of the product", "start_char_pos": 378, "end_char_pos": 924, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "They may think they are buying heroin, unaware that it is actually gray death.", "after": "As such, even experienced opioid users risk serious injury or death when taking this drug mixture.", "start_char_pos": 927, "end_char_pos": 1005, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "frequently requires", "after": "may require", "start_char_pos": 1038, "end_char_pos": 1057, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": ", whereas a typical overdose of heroin (which is much less potent) typically needs", "after": ". By contrast, an overdose from morphine or from high-purity heroin would ordinarily need", "start_char_pos": 1085, "end_char_pos": 1167, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "is consistent with the difficulty of overdose reversal seen with", "after": "difficulty is regularly encountered when treating overdoses of", "start_char_pos": 1188, "end_char_pos": 1252, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "causes less naloxoneto bind to receptors, increasing the dosage necessary to counteract the resulting", "after": "impedes the activity of naloxone, which is an antagonist at the receptor. Increasing the dosage of naloxone or its frequency of administration may be required to counteract", "start_char_pos": 1396, "end_char_pos": 1497, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "The difficulty in reversing an overdose of gray death is likely due to the presence of a fentanyl analogue, which reporting seems to corroborate as of December 2018.", "after": null, "start_char_pos": 1522, "end_char_pos": 1687, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 74, 172, 236, 377, 644, 710, 869, 926, 1005, 1182, 1329, 1521], "domain": "wiki"}
{"doc_id": "1560029", "revision_depth": 1, "before_revision": "The Kingdom of Tungning (), or the Kingdom of Formosa, was a dynastic  state that ruled part of southwestern Formosa (Taiwan)  between 1661 and 1683. It  was founded by Koxinga (Zheng Chenggong) as part of the loyalist movement to restore the Ming dynasty on the Chinese mainland after its rump state in  southern China was overthrown by the Manchu-led Qing dynasty. Koxinga hoped to recapture the mainland China from the Qing, using the island as a base of operations. Until its annexation by the Qing dynasty in 1683, the Kingdom was ruled by Koxinga's heirs, the House of Koxinga  . Names In reference to its reigning dynasty, the Kingdom of Tungning is sometimes known as the Zheng dynasty (), Zheng Family Kingdom () or Kingdom of Yanping () . Taiwan was referred to by Koxinga as Tungtu (). In the West , it was known as the Kingdom of Taiwan, and the period of rule is sometimes referred to as the Koxinga dynasty. \"Historical and Legal Aspects of the International Status of Taiwan (Formosa)\" WUFI", "after_revision": "The Kingdom of Tungning (), also known as Tywan by the British at the time, pp. 347348. was a dynastic maritime state that ruled part of southwestern Formosa (Taiwan) and Penghu islands between 1661 and 1683. At its peak the kingdom's maritime power dominated varying extents of coastal regions of southeastern China, and its vast trade network stretching from Japan to Southeast Asia. The kingdom was founded by Koxinga (Zheng Chenggong) after seizing control of Taiwan from the Dutch rule. Zheng hoped to restore the Ming dynastic rule on the Chinese mainland , when the Ming remnants' rump state in the southern China was progressively conquered by the Manchu-led Qing dynasty. Zheng dynasts used the newly-owned island as part of the loyalist movement aiming to reclaim the mainland China from the Qing, mainly as a base of military operations but also deepening the process of Sinicization on Taiwan, in an effort to consolidate the last stronghold of Han Chinese resistance against the invading Manchus.\"Historical and Legal Aspects of the International Status of Taiwan (Formosa)\" by Ng Yuzin Chiautong, published on August 28, 1971, WUFI Until its annexation by the Qing dynasty in 1683, the kingdom was ruled by Koxinga's heirs, the House of Koxinga , and the period of rule is sometimes referred to as the Koxinga dynasty . Names In reference to its reigning dynasty, the Kingdom of Tungning is sometimes known as the Zheng dynasty (), Zheng clan Kingdom () or Yanping Kingdom (), named after Koxinga's hereditary title of \"Prince of Yanping\" () that bestowed by the Yongli emperor of the South Ming . Taiwan was referred to by Koxinga as Tungtu (). In Britain , it was known as Tywan (Taiwan), named after the King's residence at the city of \"Tywan\" in present-day Tainan. The period of rule is sometimes referred to as the Koxinga dynasty. ", "edit_actions": [{"type": "R", "before": "or the Kingdom of Formosa,", "after": "also known as Tywan by the British at the time, pp. 347348.", "start_char_pos": 28, "end_char_pos": 54, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "maritime", "start_char_pos": 70, "end_char_pos": 70, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "and Penghu islands", "start_char_pos": 126, "end_char_pos": 126, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "It", "after": "At its peak the kingdom's maritime power dominated varying extents of coastal regions of southeastern China, and its vast trade network stretching from Japan to Southeast Asia.", "start_char_pos": 150, "end_char_pos": 152, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "The kingdom", "start_char_pos": 153, "end_char_pos": 153, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "as part of the loyalist movement", "after": "after seizing control of Taiwan from the Dutch rule. Zheng hoped", "start_char_pos": 195, "end_char_pos": 227, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "dynasty", "after": "dynastic rule", "start_char_pos": 248, "end_char_pos": 255, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "after its", "after": ", when the Ming remnants'", "start_char_pos": 280, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 304, "end_char_pos": 304, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "overthrown", "after": "progressively conquered", "start_char_pos": 324, "end_char_pos": 334, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "Koxinga hoped to recapture", "after": "Zheng dynasts used the newly-owned island as part of the loyalist movement aiming to reclaim", "start_char_pos": 367, "end_char_pos": 393, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "using the island", "after": "mainly", "start_char_pos": 428, "end_char_pos": 444, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "operations.", "after": "military operations but also deepening the process of Sinicization on Taiwan, in an effort to consolidate the last stronghold of Han Chinese resistance against the invading Manchus.\"Historical and Legal Aspects of the International Status of Taiwan (Formosa)\" by Ng Yuzin Chiautong, published on August 28, 1971, WUFI", "start_char_pos": 458, "end_char_pos": 469, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "Kingdom", "after": "kingdom", "start_char_pos": 524, "end_char_pos": 531, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ", and the period of rule is sometimes referred to as the Koxinga dynasty", "start_char_pos": 583, "end_char_pos": 583, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "Family", "after": "clan", "start_char_pos": 704, "end_char_pos": 710, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Kingdom of Yanping ()", "after": "Yanping Kingdom (), named after Koxinga's hereditary title of \"Prince of Yanping\" () that bestowed by the Yongli emperor of the South Ming", "start_char_pos": 725, "end_char_pos": 746, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "the West", "after": "Britain", "start_char_pos": 800, "end_char_pos": 808, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "the Kingdom of Taiwan, and the", "after": "Tywan (Taiwan), named after the King's residence at the city of \"Tywan\" in present-day Tainan. The", "start_char_pos": 827, "end_char_pos": 857, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "\"Historical and Legal Aspects of the International Status of Taiwan (Formosa)\" WUFI", "after": null, "start_char_pos": 922, "end_char_pos": 1005, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 149, 366, 469, 585, 796, 921], "domain": "wiki"}
{"doc_id": "1602860", "revision_depth": 1, "before_revision": "In the news In 2005, internet hunting became a major news story when a man in Texas set up a webcam and remotely controlled gun to allow hunters to shoot from their computers. According to the Humane Society, most internet hunts involve game ranches where animals are kept penned, making them essentially canned hunts.Humane Society Wildlife Abuse Campaign, Another less well-known incident occurred  two years prior to the Dick Cheney hunting incident when the vice president participated in a canned hunt at the Rolling Rock Club in Ligonier Township, Pennsylvania. Cheney and nine companions killed 417 out of 500 ringneck pheasants, of which the Vice President killed 70, and an unknown number of mallard ducks. In South Africa, the Minister of Environmental Affairs, Mr. Marthinus van Schalkwyk, in 2006, announced new laws to stop the practise of \"canned hunting\" in his country. South Africa environment minister announced long-awaited restrictions on lion hunting, declaring he was sickened by wealthy tourists shooting tame lions from the back of a truck and felling rhinos with a bow and arrow. This comes in response to the imbroglio created over the potential canned hunt of the African rhinoceros 'Baixinha.' Dismissing threats of legal action by the hunting industry, Marthinus Van Schalkwyk said the new law would ban \"canned\" hunting of big predators and rhinos in small enclosures that offer them no means of escape. In addition, lions bred in captivity would have to be released into the open for at least two years before they could be hunted. Van Schalkwyk said a previously proposed six-month delay would not give lions enough time to develop self-defence instincts. \"Hunting should be about fair chase ... testing the wits of a hunter against that of the animal,\" he told a press conference. \"Over the years that got eroded and now we are trying to re-establish that principle.\" This measure was later overturned. In May 2007 a much-reported hunting trip involved the killing of 1,051pound pet pig in an alleged canned hunt. The pig was named \"Monster Pig\" by the media and it was believed that the pig was a feral hog . It was soon discovered that the pig, previously named \"Fred , \"  had been someone's pet and was then sold to a hunting facility only a brief time before he was killed. On May 3, paying customers Mike Stone and his 11-year-old son , Jamison , hunted him in a fenced enclosure. Jamison shot Fred a total of eight times over a period of three hours. A June 2007 story on CNN detailed canned hunting in South Africa and includes a video of a canned lion hunt where the animal is shot against a fence. On November 30, 2014, CBS  60 Minutes broadcast a story (\"The Lion Whisperer\") about one man's sanctuary in South Africa for 26 lions, raised in captivity, who he rescued from the fate of canned lion hunting. In November 2015, the Professional Hunters' Association of South Africa (PHASA) voted to disassociate itself from the practice of canned lion hunting in South Africa. Unless the conservation value of canned lion hunting is demonstrated to PHASA, URLanization will remain against the practice and, under penalty of expulsion, no PHASA members will be permitted to participate canned lion hunts. In November 2017  PHASA came under fire over the body's reversal of its 2015 policy against the hunting of captive-bred lions.", "after_revision": "In the news In 2005, internet hunting became a major news story when a man in Texas set up a webcam and remotely controlled gun to allow hunters to shoot from their computers. According to the Humane Society, most internet hunts involve game ranches where animals are kept penned, making them essentially canned hunts.Humane Society Wildlife Abuse Campaign, Another less well-known incident occurred in late 2003, two years prior to the infamous hunting accident, when then-Vice President Dick Cheney participated in a canned hunt at the Rolling Rock Club in Ligonier Township, Pennsylvania. Cheney and nine companions killed 417 out of 500 ringneck pheasants, of which the Vice President killed 70, and an unknown number of mallard ducks. In 2006, the South African Minister of Environmental Affairs, Marthinus van Schalkwyk, announced new laws to stop the practise of \"canned hunting\" in his country. The South African Environment Minister announced long-awaited restrictions on lion hunting, declaring he was sickened by wealthy tourists shooting tamed lions from the back of a truck and felling rhinos with a bow and arrow. This comes in response to the imbroglio created over the potential canned hunt of the African rhinoceros 'Baixinha.' Dismissing threats of legal action by the hunting industry, Marthinus Van Schalkwyk said the new law would ban \"canned\" hunting of big predators and rhinos in small enclosures that offer them no means of escape. In addition, lions bred in captivity would have to be released into the open for at least two years before they could be hunted. Van Schalkwyk said a previously proposed six-month delay would not give lions enough time to develop self-defence instincts. \"Hunting should be about fair chase ... testing the wits of a hunter against that of the animal,\" he told a press conference. \"Over the years that got eroded and now we are trying to re-establish that principle.\" This measure was later overturned. In May 2007 a much-reported hunting trip involved the killing of a pet pig in an alleged canned hunt. The pig was named \"Monster Pig\" by the media and was believed as a feral hog , but it was soon discovered that the pig, previously named \"Fred \" , had been someone's pet and was then sold to a hunting facility only a brief time before he was killed. On May 3, paying customers Mike Stone and his 11-year-old son Jamison hunted the pig in a fenced enclosure. Jamison shot Fred a total of eight times over a period of three hours. A June 2007 story on CNN detailed canned hunting in South Africa and includes a video of a canned lion hunt where the animal is shot against a fence. On November 30, 2014, CBS 's 60 Minutes broadcast a story (\"The Lion Whisperer\") about one man's sanctuary in South Africa for 26 lions, raised in captivity, who he rescued from the fate of canned lion hunting. In November 2015, the Professional Hunters' Association of South Africa (PHASA) voted to disassociate itself from the practice of canned lion hunting in South Africa. Unless the conservation value of canned lion hunting is demonstrated to PHASA, URLanization will remain against the practice and, under penalty of expulsion, no PHASA members will be permitted to participate canned lion hunts. In November 2017 , PHASA came under fire over the body's reversal of its 2015 policy against the hunting of captive-bred lions.", "edit_actions": [{"type": "A", "before": null, "after": "in late 2003,", "start_char_pos": 400, "end_char_pos": 400, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Dick Cheney hunting incident when the vice president", "after": "infamous hunting accident, when then-Vice President Dick Cheney", "start_char_pos": 424, "end_char_pos": 476, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "South Africa, the", "after": "2006, the South African", "start_char_pos": 719, "end_char_pos": 736, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "Mr.", "after": null, "start_char_pos": 772, "end_char_pos": 775, "major_intent": "style", "raw_intents": ["style", "clarity", "others"]}, {"type": "D", "before": "in 2006,", "after": null, "start_char_pos": 801, "end_char_pos": 809, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "South Africa environment minister", "after": "The South African Environment Minister", "start_char_pos": 886, "end_char_pos": 919, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "tame", "after": "tamed", "start_char_pos": 1028, "end_char_pos": 1032, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "1,051pound", "after": "a", "start_char_pos": 2001, "end_char_pos": 2011, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "it was believed that the pig was", "after": "was believed as", "start_char_pos": 2096, "end_char_pos": 2128, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": ". It", "after": ", but it", "start_char_pos": 2141, "end_char_pos": 2145, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "style"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 2203, "end_char_pos": 2204, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 2207, "end_char_pos": 2207, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", Jamison , hunted him", "after": "Jamison hunted the pig", "start_char_pos": 2373, "end_char_pos": 2395, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "'s", "start_char_pos": 2666, "end_char_pos": 2666, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 3260, "end_char_pos": 3260, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 175, 318, 567, 715, 885, 1104, 1221, 1433, 1562, 1687, 1813, 1900, 1935, 2046, 2142, 2310, 2418, 2489, 2639, 2848, 3015, 3242], "domain": "wiki"}
{"doc_id": "1602860", "revision_depth": 3, "before_revision": "Criticism A number of groups object to the practice of canned hunting for reasons such as cruelty to animals or that it takes away what is known as \"fair chase.\" The Humane Society of the United States (HSUS) criticizes canned hunting. In a statement, the HSUS called canned hunts \"cruel and brutal activities , \"  in which the hunted animal has \"absolutely no chance of escape . \"  It went on to say that animals have been \"psychologically conditioned to behave as a target by life in captivity , \"  among other objections. In 2014, 62 cities across the globe participated in the first-ever global march against canned hunting. The march was organized by Campaign Against Canned Hunting. People from many parts of the world took part: Africa, Asia, Australia, Canada, Europe, Middle East, New Zealand, South America and the United States  .Your African Safari, \"Campaign against canned hunting in South Africa: interview with Chris Mercer\"", "after_revision": "Criticism A number of groups object to the practice of canned hunting for reasons such as cruelty to animals or that it takes away the hunting ethics known as \"fair chase.\" The Humane Society of the United States (HSUS) criticizes canned hunting. In a statement, the HSUS called canned hunts \"cruel and brutal activities \" , in which the hunted animal has \"absolutely no chance of escape \" . It went on to say that animals have been \"psychologically conditioned to behave as a target by life in captivity \" , among other objections. In 2014, 62 cities across the globe participated in the first-ever global march against canned hunting. The march was organized by Campaign Against Canned Hunting. People from many parts of the world took part: Africa, Asia, Australia, Europe, Middle East, New Zealand, North America (Canada and the United States ) and South America .Your African Safari, \"Campaign against canned hunting in South Africa: interview with Chris Mercer\"", "edit_actions": [{"type": "R", "before": "what is", "after": "the hunting ethics", "start_char_pos": 131, "end_char_pos": 138, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 310, "end_char_pos": 311, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 314, "end_char_pos": 314, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 378, "end_char_pos": 379, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 382, "end_char_pos": 382, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 496, "end_char_pos": 497, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 500, "end_char_pos": 500, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "D", "before": "Canada,", "after": null, "start_char_pos": 761, "end_char_pos": 768, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "South America", "after": "North America (Canada", "start_char_pos": 803, "end_char_pos": 816, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ") and South America", "start_char_pos": 839, "end_char_pos": 839, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 161, 235, 524, 628, 688, 841], "domain": "wiki"}
{"doc_id": "1611040", "revision_depth": 1, "before_revision": "Raffique Shah  is a Trinidad and Tobago trade union leader and political commentator. He is also a former Member of Parliament and mutineer, having led a mutiny of Trinidad and Tobago Regiment in 1970. He was born the son of a sugar cane worker and housewife in March 1946. His early education was at Presentation College, Chaguanas, where he gained a Grade I Cambridge School Certificate. He later won a cadetship to the prestigious British Royal Military Academy Sandhurst between 1964 and 1966. As a Lieutenant in the Trinidad and Tobago Regiment, he led an army mutiny in 1970. In the midst of Black Power riots, the People's National Movement government led by Eric Williams proclaimed a State of Emergency April, 1970 and arrested several Black Power leaders and Trade Unionists. In a move to prevent Williams from using the military against the masses, a portion of the Regiment stationed at Teteron (on the Chaguaramas Peninsula), led by Lieutenants Shah and Rex Lassalle mutinied. In response, the Trinidad and Tobago Coast Guard fired on the mutineers who returned to Teteron Barracks, abandoning their foray into Port of Spain. After ten days the mutineers surrendered. The leaders were court-martialed and Shah and Lasalle were jailed. They both went on to win their appeals at the local Court of Appeal, and judges at the Privy Council in London, England, ruled against the State's appeal in July , 1972. Shah, Lassalle and all the other mutineers were freed within 24 hours of that ruling. After being released from prison Shah joined the Trade Union movement leading the ICFTU, a body of cane farmers who sought to displace the government URLanisation, TICFA . He , URLe Weekes and Basdeo Panday went on to found the United Labour Front  in 1975 as a labor confederation and transformed it into a political party in 1976. During general election later that year, the ULF secured 10 of the 36 seats in Parliament and became the official Opposition. Basdeo Panday was appointed Leader of the Opposition, but following a bitter row in the party, Shah was able to displace him, and served as Leader of the Opposition between August 9, 1977 and March 31, 1978 when Panday regained control of the party. Shah went on to lead the cane farmers and won an election to head the Trinidad Islandwide Cane Farmers' Association (TICFA), in 1998. He has been chairman of URLanisation to date. Shah also served as editor of the Trinidad and Tobago Mirror newspaper for 17 years and today serves as an op-ed columnist in the Trinidad  Express newspaper. He is also known as the man who founded the prestigious annual Clico Trinidad & Tobago International Marathon, an event that spawned road running throughout the Caribbean. He remained chairman of URLanising committee until he stepped down in 2005. He served, too, as secretary of the local National Amateur Athletics Association and still holds a strong interest in sports in general and athletics in particular.", "after_revision": "Raffique Shah (born 1946)Raffique Shah, \"Slaves to technology\", Trinidad and Tobago News Blog, 11 April 2018. is a Trinidad and Tobago trade union leader and political commentator. He is also a former Member of Parliament and mutineer, having led a mutiny of Trinidad and Tobago Regiment in 1970. Biography Shah was born the son of a sugarcane worker and housewife in March 1946. His early education was at Presentation College, Chaguanas, where he gained a Grade I Cambridge School Certificate. He later won a cadetship to the prestigious British Royal Military Academy Sandhurst between 1964 and 1966. As a Lieutenant in the Trinidad and Tobago Regiment, he led an army mutiny in 1970. In the midst of Black Power riots, the People's National Movement government led by Eric Williams proclaimed a State of Emergency in April 1970 and arrested several Black Power leaders and Trade Unionists. In a move to prevent Williams from using the military against the masses, a portion of the Regiment stationed at Teteron (on the Chaguaramas Peninsula), led by Lieutenants Shah and Rex Lassalle mutinied. In response, the Trinidad and Tobago Coast Guard fired on the mutineers who returned to Teteron Barracks, abandoning their foray into Port of Spain. After 10 days, the mutineers surrendered. The leaders were court-martialed and Shah and Lasalle were jailed. They both went on to win their appeals at the local Court of Appeal, and judges at the Privy Council in London, England, ruled against the State's appeal in July 1972. Shah, Lassalle and all the other mutineers were freed within 24 hours of that ruling. After being released from prison Shah joined the Trade Union movement leading the ICFTU, a body of cane farmers who sought to displace the URLanisation, TICFA (Trinidad Islandwide Cane Farmers' Association). Shah , URLe Weekes and Basdeo Panday went on to found the United Labour Front (ULF) in 1975 as a labour confederation and transformed it into a political party in 1976. During general election later that year, the ULF secured 10 of the 36 seats in Parliament and became the official Opposition. Basdeo Panday was appointed Leader of the Opposition, but following a bitter row in the party, Shah was able to displace him, and served as Leader of the Opposition between August 9, 1977 and March 31, 1978 when Panday regained control of the party. Shah went on to lead the cane farmers and won an election to head the Trinidad Islandwide Cane Farmers' Association (TICFA), in 1998. He has been chairman of URLanisation to date. Shah also served as editor of the Trinidad and Tobago Mirror newspaper for 17 years and went on to serve as an op-ed columnist in the Trinidad and Tobago Express newspaper. He is also known as the man who founded the prestigious annual Clico Trinidad & Tobago International Marathon, an event that spawned road-running throughout the Caribbean. He remained chairman of URLanising committee until he stepped down in 2005. He served, too, as secretary of the local National Amateur Athletics Association and still holds a strong interest in sports in general and athletics in particular.", "edit_actions": [{"type": "A", "before": null, "after": "(born 1946)Raffique Shah, \"Slaves to technology\", Trinidad and Tobago News Blog, 11 April 2018.", "start_char_pos": 14, "end_char_pos": 14, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "He", "after": "Biography Shah", "start_char_pos": 202, "end_char_pos": 204, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "sugar cane", "after": "sugarcane", "start_char_pos": 227, "end_char_pos": 237, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "April,", "after": "in April", "start_char_pos": 712, "end_char_pos": 718, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "ten days", "after": "10 days,", "start_char_pos": 1145, "end_char_pos": 1153, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1410, "end_char_pos": 1411, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "government", "after": null, "start_char_pos": 1643, "end_char_pos": 1653, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ". He", "after": "(Trinidad Islandwide Cane Farmers' Association). Shah", "start_char_pos": 1674, "end_char_pos": 1678, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "(ULF)", "start_char_pos": 1752, "end_char_pos": 1752, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "labor", "after": "labour", "start_char_pos": 1766, "end_char_pos": 1771, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "today serves", "after": "went on to serve", "start_char_pos": 2481, "end_char_pos": 2493, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "and Tobago", "start_char_pos": 2532, "end_char_pos": 2532, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "road running", "after": "road-running", "start_char_pos": 2685, "end_char_pos": 2697, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 85, 201, 273, 389, 497, 581, 785, 989, 1138, 1180, 1247, 1417, 1503, 1836, 1962, 2212, 2346, 2392, 2551, 2723, 2799]}
{"doc_id": "166931", "revision_depth": 1, "before_revision": "Inland deltas Okavango DeltaIn  rare cases the river delta is located inside a large valley and is called an inverted river delta. Sometimes a river divides into multiple branches in an inland area, only to rejoin and continue to the sea. Such an area is called an inland delta, and often occurs on former lake beds. The  Inner Niger Delta, PeaceAthabasca Delta, and SacramentoSan Joaquin River Delta are notable examples. The Amazon also has an inland delta before the island of Maraj , and the  Danube has one in the valley on the Slovak-Hungarian border between Bratislava and Ia . In the Philippines, you would find Pampanga River as it enters the Central Luzon Plain in Gabaldon, Nueva Ecija as another example of an inland delta, as well as Liguasan Marsh of the Mindanao River .", "after_revision": "Inland deltas Okavango Delta In rare cases the river delta is located inside a large valley and is called an inverted river delta. Sometimes a river divides into multiple branches in an inland area, only to rejoin and continue to the sea. Such an area is called an inland delta, and often occurs on former lake beds. The term was first coined by Alexander von Humboldt for the middle reaches of the Orinoco River, which he visited in 1800. Other prominent examples include the Inner Niger Delta, PeaceAthabasca Delta, , the SacramentoSan Joaquin River Delta , and the Sistan delta of Iran. The Danube has one in the valley on the Slovak-Hungarian border between Bratislava and Ia .", "edit_actions": [{"type": "R", "before": "DeltaIn", "after": "Delta", "start_char_pos": 23, "end_char_pos": 30, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "In", "start_char_pos": 31, "end_char_pos": 31, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "A", "before": null, "after": "term was first coined by Alexander von Humboldt for the middle reaches of the Orinoco River, which he visited in 1800. Other prominent examples include the", "start_char_pos": 321, "end_char_pos": 321, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "and", "after": ", the", "start_char_pos": 363, "end_char_pos": 366, "major_intent": "fluency", "raw_intents": ["fluency", "others", "coherence"]}, {"type": "D", "before": "are notable examples. The Amazon also has an inland delta before the island of Maraj", "after": null, "start_char_pos": 401, "end_char_pos": 485, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "Sistan delta of Iran. The", "start_char_pos": 496, "end_char_pos": 496, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ". In the Philippines, you would find Pampanga River as it enters the Central Luzon Plain in Gabaldon, Nueva Ecija as another example of an inland delta, as well as Liguasan Marsh of the Mindanao River", "after": null, "start_char_pos": 583, "end_char_pos": 783, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 130, 238, 316, 422, 584], "domain": "wiki"}
{"doc_id": "16903", "revision_depth": 1, "before_revision": "The Kuomintang (KMT), also known as the Chinese Nationalist Party or the Nationalist Party of China, is a major political party in Taiwan and currently the largest opposition party in the Legislative Yuan. In modern-day Taiwanese politics, the KMT primarily competes with rival Democratic Progressive Party, opposing de jure Taiwan independence and Chinese unification under \"One Country, Two Systems,\" instead favoring cordial cross-strait relations with China by endorsing the 1992 Consensus and maintaining Taiwan's status quo under the Constitution of the Republic of China. The current chairman is Johnny Chiang. The Kuomintang is one of two major political parties in contemporary in China, the other being the Chinese Communist Party  . The Kuomintang traces its origins to the Revive China Society, which was founded by Sun Yat-sen in 1894 to promote the republican revolution in China and overthrow the crumbling Qing dynasty. The party was formally established in 1919 during the Beiyang regime. From 1926 to 1928, the KMT under Chiang Kai-shek successfully led the Northern Expedition against regional warlords and unified the fragmented nation. From 1937 to 1945, the Nationalists led China through the Second Sino-Japanese War against the Empire of Japan. By 1949, the KMT was decisively defeated by the Chinese Communist Party in the Chinese Civil War and withdrew the Nationalist government to Taiwan, which was restored to the ROC at the end of World War II. From 1949 to 1987, the Kuomintang ruled Taiwan under martial law hostile to the Communist mainland  , before Chiang Ching-kuo  lifted authoritarian restrictions and initiated democratic reforms in the late 1980s. During this period, the KMT also oversaw Taiwan's economic development but experienced diplomatic setbacks, including the Republic of China losing its United Nations seat and the United States switching diplomatic recognition to the People's Republic of China in the 1970s. In the 1990s, President Lee Teng-hui pursued constitutional reforms and was re-elected in the 1996 Taiwanese presidential election, the first direct presidential election under the ROC Constitution. However, the 2000 presidential election put an end to more than half a century of KMT rule on Taiwan . The KMT reclaimed power with the landslide victory of Ma Ying-jeou in the 2008 presidential election, whose presidency from 2008 to 2016 substantially improved cross-strait relations. With the 2016 presidential and legislative elections, the KMT lost both the presidency and its legislative majority, returning to the opposition.", "after_revision": "The Kuomintang (KMT), also known as the Chinese Nationalist Party or the Nationalist Party of China, is a major political party in the Republic of China (ROC), commonly known as \"Taiwan\" and currently the largest opposition party in the Legislative Yuan. In modern-day Taiwanese politics, the KMT primarily competes with rival Democratic Progressive Party, opposing de jure Taiwan independence and Chinese unification under \"One Country, Two Systems,\" instead favoring cordial cross-strait relations with the People's Republic of China (PRC) by endorsing the 1992 Consensus and maintaining Taiwan's status quo under the Constitution of the Republic of China. The current chairman is Johnny Chiang. The Kuomintang is among the two major historical contemporary parties in China, the other being the Chinese Communist Party (CCP) . The Kuomintang traces its origins to the Revive China Society, which was founded by Sun Yat-sen in 1894 to promote the republican revolution in China and overthrow the crumbling Qing dynasty. The party was formally established in 1919 during the Beiyang regime. From 1926 to 1928, the KMT under Chiang Kai-shek successfully led the Northern Expedition against regional warlords and unified the fragmented nation. From 1937 to 1945, the Nationalists led China through the Second Sino-Japanese War against Japan. By 1949, the KMT was decisively defeated by the Chinese Communist Party in the Chinese Civil War and withdrew the Nationalist government to Taiwan, a former Qing prefecture-turned-Japanese colony that ruled from 1895 to 1945. From 1949 to 1987, the Kuomintang ruled Taiwan under martial law hostile to the Communist mainland for nearly 40 years , before Chiang Ching-kuo , Chiang Kai-shek's son, lifted authoritarian restrictions and initiated democratic reforms in the late 1980s. During this period, the KMT also oversaw Taiwan's economic development but experienced diplomatic setbacks, including the Republic of China losing its United Nations seat and the United States switching diplomatic recognition to the People's Republic of China in the 1970s. In the 1990s, President Lee Teng-hui pursued constitutional reforms and was re-elected in the 1996 Taiwanese presidential election, the first direct presidential election under the ROC Constitution. However, the 2000 presidential election put an end to more than half a century of KMT rule that dominated Chinese politics for over 72 years . The KMT reclaimed power with the landslide victory of Ma Ying-jeou in the 2008 presidential election, whose presidency from 2008 to 2016 substantially improved cross-strait relations. With the 2016 presidential and legislative elections, the KMT lost both the presidency and its legislative majority, returning to the opposition.", "edit_actions": [{"type": "R", "before": "Taiwan", "after": "the Republic of China (ROC), commonly known as \"Taiwan\"", "start_char_pos": 131, "end_char_pos": 137, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "China", "after": "the People's Republic of China (PRC)", "start_char_pos": 456, "end_char_pos": 461, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "one of two major political parties in contemporary in", "after": "among the two major historical contemporary parties in", "start_char_pos": 636, "end_char_pos": 689, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "(CCP)", "start_char_pos": 741, "end_char_pos": 741, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "D", "before": "the Empire of", "after": null, "start_char_pos": 1248, "end_char_pos": 1261, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "which was restored to the ROC at the end of World War II.", "after": "a former Qing prefecture-turned-Japanese colony that ruled from 1895 to 1945.", "start_char_pos": 1417, "end_char_pos": 1474, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "for nearly 40 years", "start_char_pos": 1574, "end_char_pos": 1574, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": ", Chiang Kai-shek's son,", "start_char_pos": 1601, "end_char_pos": 1601, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "on Taiwan", "after": "that dominated Chinese politics for over 72 years", "start_char_pos": 2252, "end_char_pos": 2261, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 205, 578, 617, 743, 935, 1005, 1156, 1268, 1474, 1687, 1961, 2160, 2263, 2447], "domain": "wiki"}
{"doc_id": "172474", "revision_depth": 1, "before_revision": "Exercise and lactate During power exercises such as sprinting, when the rate of demand for energy is high, glucose is broken down and oxidized to pyruvate , and lactate is then produced from the pyruvate faster than the body can process it, causing lactate concentrations to rise.  The production of lactate is beneficial for NAD+ regeneration (pyruvate is reduced to lactate while NADH is oxidized to NAD+), which is used up in oxidation of glyceraldehyde 3-phosphate during production of pyruvate from glucose, and this ensures that energy production is maintained and exercise can continue. During intense exercise, the respiratory chain cannot keep up with the amount of hydrogen ions that join to form NADH, and cannot regenerate NAD+ quickly enough. However, lactate is continually formed at rest and during all exercise intensities. With the use of stable isotopic tracers URLe Brooks and colleagues, at the University of California, Berkeley, observed that lactate served as a metabolic fuel being produced and oxidatively disposed in resting and exercising muscle. Some causes of this are metabolism in red blood cells that lack mitochondria, and limitations resulting from the enzyme activity that occurs in muscle fibers having high glycolytic capacity. In 2004, Robergs et al. maintained that lactic acidosis during exercise is a \"construct\" or myth, pointing out that part of the H+ comes from ATP hydrolysis (ATP4 + H2O ADP3 + + H+), and that reducing pyruvate to lactate (pyruvate + NADH + H+ lactate + NAD+) actually consumes H+. Lindinger et al. countered that they had ignored the causative factors of the increase in [H+]. After all, the production of lactate from a neutral molecule must increase [H+] to maintain electroneutrality. The point of Robergs's paper, however, was that lactate is produced from pyruvate, which has the same charge. It is pyruvate production from neutral glucose that generates H+: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O Subsequent lactate production absorbs these protons: 2 + 2H+ + 2NADH 2 + 2NAD+ Overall: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O 2 + 2NAD+ + 2ATP4 + 2H2O Although the reaction glucose 2lactate + 2H+ releases two H+ when viewed on its own, the H+ are absorbed in the production of ATP. On the other hand, the absorbed acidity is released during subsequent hydrolysis of ATP: ATP4 + H2O ADP3 + + H+. So once the use of ATP is included, the overall reaction is C6H12O6 2 + 2H+", "after_revision": "Exercise and lactate During power exercises such as sprinting, when the rate of demand for energy is high, glucose is broken down and oxidized to pyruvate . Lactate is then produced from the pyruvate faster than the body can process it, causing lactate concentrations to rise. Lactate is the end product of glycolysis and increased circulating levels of lactate are associated with increased muscle glycogen breakdown with exercise. Circulating lactate levels have been found higher in heat-sensitive patients with variants in genes encoding the skeletal muscle calcium release channel, as compared to asymptomatic carriers of the same variants. The production of lactate is beneficial for NAD+ regeneration (pyruvate is reduced to lactate while NADH is oxidized to NAD+), which is used up in oxidation of glyceraldehyde 3-phosphate during production of pyruvate from glucose, and this ensures that energy production is maintained and exercise can continue. During intense exercise, the respiratory chain cannot keep up with the amount of hydrogen ions that join to form NADH, and cannot regenerate NAD+ quickly enough. Lactate served as a metabolic fuel being produced and oxidatively disposed in resting and exercising muscle. In 2004, Robergs et al. maintained that lactic acidosis during exercise is a \"construct\" or myth, pointing out that part of the H+ comes from ATP hydrolysis (ATP4 + H2O ADP3 + + H+), and that reducing pyruvate to lactate (pyruvate + NADH + H+ lactate + NAD+) actually consumes H+. Lindinger et al. countered that they had ignored the causative factors of the increase in [H+]. After all, the production of lactate from a neutral molecule must increase [H+] to maintain electroneutrality. The point of Robergs's paper, however, was that lactate is produced from pyruvate, which has the same charge. It is pyruvate production from neutral glucose that generates H+: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O Subsequent lactate production absorbs these protons: 2 + 2H+ + 2NADH 2 + 2NAD+ Overall: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O 2 + 2NAD+ + 2ATP4 + 2H2O Although the reaction glucose 2lactate + 2H+ releases two H+ when viewed on its own, the H+ are absorbed in the production of ATP. On the other hand, the absorbed acidity is released during subsequent hydrolysis of ATP: ATP4 + H2O ADP3 + + H+. So once the use of ATP is included, the overall reaction is C6H12O6 2 + 2H+", "edit_actions": [{"type": "R", "before": ", and lactate", "after": ". Lactate", "start_char_pos": 155, "end_char_pos": 168, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "Lactate is the end product of glycolysis and increased circulating levels of lactate are associated with increased muscle glycogen breakdown with exercise. Circulating lactate levels have been found higher in heat-sensitive patients with variants in genes encoding the skeletal muscle calcium release channel, as compared to asymptomatic carriers of the same variants.", "start_char_pos": 281, "end_char_pos": 281, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "However, lactate is continually formed at rest and during all exercise intensities. With the use of stable isotopic tracers URLe Brooks and colleagues, at the University of California, Berkeley, observed that lactate", "after": "Lactate", "start_char_pos": 756, "end_char_pos": 972, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "D", "before": "Some causes of this are metabolism in red blood cells that lack mitochondria, and limitations resulting from the enzyme activity that occurs in muscle fibers having high glycolytic capacity.", "after": null, "start_char_pos": 1074, "end_char_pos": 1264, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 280, 593, 755, 839, 1073, 1264, 1545, 1641, 1752, 1862, 2290, 2403], "domain": "wiki"}
{"doc_id": "172474", "revision_depth": 2, "before_revision": "Exercise and lactate During power exercises such as sprinting, when the rate of demand for energy is high, glucose is broken down and oxidized to pyruvate . Lactate is then produced from the pyruvate faster than the body can process it, causing lactate concentrations to rise. Lactate is the end product of glycolysis and increased circulating levels of lactate are associated with increased muscle glycogen breakdown with exercise. Circulating lactate levels have been found higher in heat-sensitive patients with variants in genes encoding the skeletal muscle calcium release channel, as compared to asymptomatic carriers of the same variants. The production of lactate is beneficial for NAD+ regeneration (pyruvate is reduced to lactate while NADH is oxidized to NAD+), which is used up in oxidation of glyceraldehyde 3-phosphate during production of pyruvate from glucose, and this ensures that energy production is maintained and exercise can continue. During intense exercise, the respiratory chain cannot keep up with the amount of hydrogen ions that join to form NADH, and cannot regenerate NAD+ quickly enough. Lactate served as a metabolic fuel being produced and oxidatively disposed in resting and exercising muscle.  In 2004, Robergs et al. maintained that lactic acidosis during exercise is a \"construct\" or myth, pointing out that part of the H+ comes from ATP hydrolysis (ATP4 + H2O ADP3 + + H+), and that reducing pyruvate to lactate (pyruvate + NADH + H+ lactate + NAD+) actually consumes H+. Lindinger et al. countered that they had ignored the causative factors of the increase in [H+]. After all, the production of lactate from a neutral molecule must increase [H+] to maintain electroneutrality. The point of Robergs's paper, however, was that lactate is produced from pyruvate, which has the same charge. It is pyruvate production from neutral glucose that generates H+: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O Subsequent lactate production absorbs these protons: 2 + 2H+ + 2NADH 2 + 2NAD+ Overall: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O 2 + 2NAD+ + 2ATP4 + 2H2O Although the reaction glucose 2lactate + 2H+ releases two H+ when viewed on its own, the H+ are absorbed in the production of ATP. On the other hand, the absorbed acidity is released during subsequent hydrolysis of ATP: ATP4 + H2O ADP3 + + H+. So once the use of ATP is included, the overall reaction is C6H12O6 2 + 2H+", "after_revision": "Exercise and lactate During power exercises such as sprinting, when the rate of demand for energy is high, glucose is broken down and oxidized to pyruvate , and lactate is then produced from the pyruvate faster than the body can process it, causing lactate concentrations to rise. The production of lactate is beneficial for NAD+ regeneration (pyruvate is reduced to lactate while NADH is oxidized to NAD+), which is used up in oxidation of glyceraldehyde 3-phosphate during production of pyruvate from glucose, and this ensures that energy production is maintained and exercise can continue. During intense exercise, the respiratory chain cannot keep up with the amount of hydrogen ions that join to form NADH, and cannot regenerate NAD+ quickly enough. However, lactate is continually formed at rest and during all exercise intensities. With the use of stable isotopic tracers URLe Brooks and colleagues, at the University of California, Berkeley, observed that lactate served as a metabolic fuel being produced and oxidatively disposed in resting and exercising muscle. Some causes of this are metabolism in red blood cells that lack mitochondria, and limitations resulting from the enzyme activity that occurs in muscle fibers having high glycolytic capacity. In 2004, Robergs et al. maintained that lactic acidosis during exercise is a \"construct\" or myth, pointing out that part of the H+ comes from ATP hydrolysis (ATP4 + H2O ADP3 + + H+), and that reducing pyruvate to lactate (pyruvate + NADH + H+ lactate + NAD+) actually consumes H+. Lindinger et al. countered that they had ignored the causative factors of the increase in [H+]. After all, the production of lactate from a neutral molecule must increase [H+] to maintain electroneutrality. The point of Robergs's paper, however, was that lactate is produced from pyruvate, which has the same charge. It is pyruvate production from neutral glucose that generates H+: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O Subsequent lactate production absorbs these protons: 2 + 2H+ + 2NADH 2 + 2NAD+ Overall: C6H12O6 + 2NAD+ + 2ADP3 + 2 2 + 2H+ + 2NADH + 2ATP4 + 2H2O 2 + 2NAD+ + 2ATP4 + 2H2O Although the reaction glucose 2lactate + 2H+ releases two H+ when viewed on its own, the H+ are absorbed in the production of ATP. On the other hand, the absorbed acidity is released during subsequent hydrolysis of ATP: ATP4 + H2O ADP3 + + H+. So once the use of ATP is included, the overall reaction is C6H12O6 2 + 2H+", "edit_actions": [{"type": "R", "before": ". Lactate", "after": ", and lactate", "start_char_pos": 155, "end_char_pos": 164, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "D", "before": "Lactate is the end product of glycolysis and increased circulating levels of lactate are associated with increased muscle glycogen breakdown with exercise. Circulating lactate levels have been found higher in heat-sensitive patients with variants in genes encoding the skeletal muscle calcium release channel, as compared to asymptomatic carriers of the same variants.", "after": null, "start_char_pos": 277, "end_char_pos": 645, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "Lactate", "after": "However, lactate is continually formed at rest and during all exercise intensities. With the use of stable isotopic tracers URLe Brooks and colleagues, at the University of California, Berkeley, observed that lactate", "start_char_pos": 1120, "end_char_pos": 1127, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Some causes of this are metabolism in red blood cells that lack mitochondria, and limitations resulting from the enzyme activity that occurs in muscle fibers having high glycolytic capacity.", "start_char_pos": 1229, "end_char_pos": 1229, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}], "sents_char_pos": [0, 276, 432, 645, 957, 1119, 1228, 1510, 1606, 1717, 1827, 2255, 2368], "domain": "wiki"}
{"doc_id": "17250911", "revision_depth": 2, "before_revision": "Early life Benjamin Thomas Atkins was born on August 26, 1968 in Detroit, the younger of two sons. His family lived in a poor neighborhood, and both of his parents were drug and alcohol abusers. Shortly after his birth, Tony's father left the family. In 1970, Benjamin ran away from home, and without parental supervision, wound up in an orphanage, where he spent his childhood and youth. While living there, he was physically assaulted by other children, and at the age of 10, was raped by one of the employees. Over the next five years, he was continuously subjected to sexual harassment by other boys, until he eventually escaped and reunited with his mother. For sometime, he lived together with her and his older brother, but one day, Tony realised that his mother worked as a prostitute. Due to this, both Atkins and his brother saw her have sex with clients at the house on several occasions. Disgusted with his mother's actions, he left the house again in the late 1980s, living on the streets and doing drugs, and eventually developed a drug addiction. Because he lacked formal education, Atkins was forced to work in low-skilled labor jobs for low wages, and spent the nights at homeless shelters. In his free time, he frequented places inhabited by pimps and prostitutes, but was never arrested for any serious crimes. Most of his acquaintances claimed to be very fond of him, but at the same time noted that when drunk or on drugs, he showed signs of an antisocial personality and displayed misogynistic behavior.", "after_revision": "Early life Benjamin Thomas Atkins was born on August 26, 1968 in Detroit, the younger of two sons. His family lived in a poor neighborhood, and both of his parents were drug and alcohol abusers. Shortly after his birth, Tony's father left the family. In 1970, Benjamin ran away from home, and without parental supervision, wound up in an orphanage, where he spent his childhood and youth. While living there, he was physically assaulted by other children, and at the age of 10, was raped by one of the employees. Over the next five years, he was continuously subjected to sexual harassment by other boys, until he eventually escaped and reunited with his mother. For sometime, he lived together with her and his older brother, but realised that his mother worked as a prostitute. Both Atkins and his brother saw her have sex with clients at the house on several occasions. Atkins left the house again in the late 1980s, living on the streets and doing drugs, and eventually developed a drug addiction. Because he lacked formal education, Atkins was forced to work in low-skilled labor jobs for low wages, and spent the nights at homeless shelters. Acquaintances claimed to be very fond of him, but at the same time noted that when drunk or on drugs, he showed signs of an antisocial personality and displayed misogynistic behavior.", "edit_actions": [{"type": "D", "before": "one day, Tony", "after": null, "start_char_pos": 731, "end_char_pos": 744, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "Due to this, both", "after": "Both", "start_char_pos": 794, "end_char_pos": 811, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "Disgusted with his mother's actions, he", "after": "Atkins", "start_char_pos": 900, "end_char_pos": 939, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "In his free time, he frequented places inhabited by pimps and prostitutes, but was never arrested for any serious crimes. Most of his acquaintances", "after": "Acquaintances", "start_char_pos": 1208, "end_char_pos": 1355, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}], "sents_char_pos": [0, 98, 194, 250, 388, 512, 662, 793, 899, 1061, 1207, 1329]}
{"doc_id": "17250911", "revision_depth": 3, "before_revision": "Early life Benjamin Thomas Atkins was born on August 26, 1968 in Detroit, the younger of two sons. His family lived in a poor neighborhood, and both of his parents were drug and alcohol abusers. Shortly after his birth, Tony's father left the family. In 1970, Benjamin ran away from home, and without parental supervision, wound up in an orphanage, where he spent his childhood and youth. While living there, he was physically assaulted by other children, and at the age of 10, was raped by one of the employees. Over the next five years, he was continuously subjected to sexual harassment by other boys, until he eventually escaped and reunited with his mother. For sometime, he lived together with her and his older brother, but  realised that his mother worked as a prostitute. Both Atkins and his brother saw her have sex with clients at the house on several occasions. Atkins left the house again in the late 1980s, living on the streets and doing drugs, and eventually developed a drug addiction. Because he lacked formal education, Atkins was forced to work in low-skilled labor jobs for low wages, and spent the nights at homeless shelters. Acquaintances claimed to be very fond of him, but at the same time noted that when drunk or on drugs, he showed signs of an antisocial personality and displayed misogynistic behavior. Murders As victims, Atkins chose vulnerable women, often prostitutes or drug addicts. He would lure them to abandoned buildings and houses, where he sexually assaulted  them. After strangling his victims, he would leave the bodies at the crime scenes, with some of them discovered months after their deaths. The first victim to be discovered was 30-year-old Debbie Ann Friday, found on December 14, 1991, after she had gone missing on December 8th. A few days later, on December 30th, the body of 26-year-old Bertha Jean Mason was found. She had gone missing on December 11th, and was last seen leaving her home and entering a store, after which she was never seen alive again. On January 3, 1992, while demolishing an abandoned house, workmen discovered the body of 36-year-old prostitute and drug addict Patricia Cannon URLe, who had been put on a wanted list in early December 1991 following a drug den bust within Woodward Corridor. On January 25th, the body of 39-year-old Vickie Truelove was located: like the other victims, she had been sodomized, raped and strangled. At the end of January, Atkins was arrested at an abandoned building and taken to the police station for interrogation. Due to lack of evidence to prove his guilt in the murders, he was released. On February 17th, the corpses of three women were found in three separate rooms in the former Monterey Hotel in Highland Park: they were 34-year-old Valerie Chalk, 23-year-old Juanita Hardy and a Jane Doe whose identity remains unknown. All of them had been raped and strangled . Relatives of Chalk told the police that after she was put on a wanted list in early November 1991, Valerie went missing. On April 9th, the body of 28-year-old Brenda Mitchell was found in an abandoned house, after she had gone missing four days earlier . Despite the fact Mitchell was found almost completely naked with a scarf wrapped around her neck , her death was originally classified as an overdose. %DIFDELCMD < %DIFDELCMD < %%% A few days later, on the 15th, the partially decomposed corpse of 27-year-old Vicki Beasley-Brown, who was last seen alive on March 25th, was discovered. On June 15th, the body of 45-year-old Joanne O'Rourke was found. Arrest Atkins was arrested on rape charges on August 21, 1992, after he was identified on a Detroit street by 34-year-old Darlene Saunders, who had been sodomized and raped by him in October 1991. %DIFDELCMD < %DIFDELCMD < %%% He was questioned again, but categorically denied any involvement, claiming that he had no interest in women and was a homosexual. After further interrogations, the police officers familiarized him with the psychological portrait they had compiled of the killer, which, after 12 hours, caused Atkins to admit to the murders of 11 women. He described in detail the appearance and clothing of the victims, and even indicated the whereabouts of the bodies of the 10th and 11th victims, 21-year-old Ocinena Waymer and 29-year-old LaTanya Showanda Smith. Their disappearances were not connected to the murders until Atkins' confession, and the bodies were found on the indicated place that same day. During the interrogation, Atkins said that the motive for the murders was his misogynistic views against girls and women engaged in prostitution.  Contrary to the official version of the investigation, Atkins revealed that the first victim had actually been Patricia URLe, whom he killed in the fall of 1991. According to his testimony, he didn't plan on killing the woman, but simply planned to smoke crack cocaine with her. When she decided to leave him, he flew into a rage and killed her. In several cases, he was unable to explain why he committed the murders: this was the case with Hardy, with whom Benjamin had an intimate relationship, but after doing drugs together, he said he strangled her for no apparent reason.", "after_revision": "Early life Benjamin Thomas Atkins was born on August 26, 1968 in Detroit, the younger of two sons. His family lived in a poor neighborhood, and both of his parents were drug and alcohol abusers. Shortly after his birth, Tony's father left the family. In 1970, Benjamin ran away from home, and without parental supervision, wound up in an orphanage, where he spent his childhood and youth. While living there, he was physically assaulted by other children, and at the age of 10, was raped by one of the employees. Over the next five years, he was continuously subjected to sexual harassment by other boys, until he eventually escaped and reunited with his mother. For sometime, he lived together with her and his older brother, but one day, Tony realised that his mother worked as a prostitute. Due to this, both Atkins and his brother saw her have sex with clients at the house on several occasions. Disgusted with his mother's actions, he left the house again in the late 1980s, living on the streets and doing drugs, and eventually developed a drug addiction. Because he lacked formal education, Atkins was forced to work in low-skilled labor jobs for low wages, and spent the nights at homeless shelters. In his free time, he frequented places inhabited by pimps and prostitutes, but was never arrested for any serious crimes. Most of his acquaintances claimed to be very fond of him, but at the same time noted that when drunk or on drugs, he showed signs of an antisocial personality and displayed misogynistic behavior. Murders As victims, Atkins chose young and middle-aged destitute women, often prostitutes or drug addicts. He would lure them to abandoned buildings and houses, where he sexually assaulted and sodomized them. After strangling his victims, he would leave the bodies at the crime scenes, with some of them discovered months after their deaths. The first victim to be discovered was 30-year-old Debbie Ann Friday, found on December 14, 1991, after she had gone missing on December 8th. A few days later, on December 30th, the body of 26-year-old Bertha Jean Mason was found. She had gone missing on December 11th, and was last seen leaving her home and entering a store, after which she was never seen alive again. On January 3, 1992, while demolishing an abandoned house, workmen discovered the body of 36-year-old prostitute and drug addict Patricia Cannon URLe, who had been put on a wanted list in early December 1991 following a drug den bust within Woodward Corridor. On January 25th, the body of 39-year-old Vickie Truelove was located: like the other victims, she had been sodomized, raped and strangled. At the end of January, Atkins was arrested at an abandoned building and taken to the police station for interrogation. Due to lack of evidence to prove his guilt in the murders, he was released. On February 17th, the corpses of three girls were found in three separate rooms in the former Monterey Hotel in Highland Park: they were 34-year-old Valerie Chalk, 23-year-old Juanita Hardy and a Jane Doe whose identity remains unknown. All of them had been sodomized and raped prior to their strangulations . Relatives of Chalk told the police that after she was put on a wanted list in early November 1991, Valerie went missing. On April 9th, the body of 28-year-old Brenda Mitchell was found in an abandoned house, after she had gone missing four days earlier with her two kids to go to the store. Mitchell was found almost completely naked , except for a scarf wrapped around her neck . A forensic examination determined that she had taken a lethal dose of drugs prior to her death, leading to her death being classified as an overdose. %DIFDELCMD < %DIFDELCMD < %%% A few days later, on the 15th, the partially decomposed corpse of 27-year-old Vicki Beasley-Brown, who was last seen alive on March 25th, was discovered. On June 15th, the body of 45-year-old Joanne O'Rourke was found. Arrest Atkins was arrested on rape charges on August 21, 1992, after he was identified on a Detroit street by 34-year-old Darlene Saunders, who had been sodomized and raped by him in October 1991. %DIFDELCMD < %DIFDELCMD < %%% He was questioned again, but categorically denied any involvement, claiming that he had no interest in women and was a homosexual. After further interrogations, the police officers familiarized him with the psychological portrait they had compiled of the killer, which, after 12 hours, caused Tony admit to the murders of 11 women. He described in detail the appearance and clothing of the victims, and even indicated the whereabouts of the 10th and 11th victims, 21-year-old Ocinena Waymer and 29-year-old LaTanya Showanda Smith. Their disappearances were not connected to the murders until Atkins' confession, and the bodies were found on the indicated place that same day. During the interrogation, Atkins said that the motive for the murders was his misogynistic views against girls and women engaged in prostitution. He stated that he lured his victims into abandoned houses by offering them drugs and alcohol, in addition to paying for their sexual services. Contrary to the official version of the investigation, Tony revealed that the first victim had actually been Patricia URLe, whom he killed in the fall of 1991. According to his testimony, he didn't plan on killing the woman, but simply planned to smoke crack cocaine with her. When she decided to leave him, he flew into a rage and killed her. In several cases, he was unable to explain why he committed the murders: this was the case with Hardy, with whom Benjamin had an intimate relationship, but after doing drugs together, he sodomized and strangled her for no apparent reason.", "edit_actions": [{"type": "A", "before": null, "after": "one day, Tony", "start_char_pos": 731, "end_char_pos": 731, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Both", "after": "Due to this, both", "start_char_pos": 781, "end_char_pos": 785, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "Atkins", "after": "Disgusted with his mother's actions, he", "start_char_pos": 874, "end_char_pos": 880, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "Acquaintances", "after": "In his free time, he frequented places inhabited by pimps and prostitutes, but was never arrested for any serious crimes. Most of his acquaintances", "start_char_pos": 1149, "end_char_pos": 1162, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "vulnerable", "after": "young and middle-aged destitute", "start_char_pos": 1366, "end_char_pos": 1376, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "and sodomized", "start_char_pos": 1501, "end_char_pos": 1501, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "women", "after": "girls", "start_char_pos": 2643, "end_char_pos": 2648, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "raped and strangled", "after": "sodomized and raped prior to their strangulations", "start_char_pos": 2862, "end_char_pos": 2881, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". Despite the fact", "after": "with her two kids to go to the store.", "start_char_pos": 3137, "end_char_pos": 3155, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "R", "before": "with", "after": ", except for", "start_char_pos": 3199, "end_char_pos": 3203, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ", her death was originally", "after": ". A forensic examination determined that she had taken a lethal dose of drugs prior to her death, leading to her death being", "start_char_pos": 3236, "end_char_pos": 3262, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Atkins to", "after": "Tony", "start_char_pos": 4059, "end_char_pos": 4068, "major_intent": "style", "raw_intents": ["style", "others", "style"]}, {"type": "D", "before": "bodies of the", "after": null, "start_char_pos": 4212, "end_char_pos": 4225, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "He stated that he lured his victims into abandoned houses by offering them drugs and alcohol, in addition to paying for their sexual services.", "start_char_pos": 4607, "end_char_pos": 4607, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Atkins", "after": "Tony", "start_char_pos": 4663, "end_char_pos": 4669, "major_intent": "style", "raw_intents": ["style", "others", "style"]}, {"type": "R", "before": "said he", "after": "sodomized and", "start_char_pos": 5141, "end_char_pos": 5148, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 98, 194, 250, 388, 512, 662, 780, 873, 1002, 1148, 1332, 1418, 1507, 1640, 1781, 1870, 2010, 2269, 2408, 2527, 2603, 2840, 3004, 3473, 3538, 3896, 4102, 4315, 4460, 4606, 4769, 4886, 4953], "domain": "wiki"}
{"doc_id": "17468235", "revision_depth": 1, "before_revision": "Finger bowl from the 1880s A finger bowl is a bowl of water that dinner guests use for rinsing their fingers. In a formal meal served la russe , the finger bowl is brought to the table at the time of the last course of the meal (the dessert course or the fruit course)  . In less formal service, the finger bowl may be presented after any course that involves finger food and may even be presented after more than one such course in a single meal. Formal service The finger bowl is typically delivered with the dessert plate. alt=Finger bowl|The finger bowl is moved to the upper left before dessert is served. After dessert, the guests dip their fingers into the bowl and dry them with the napkin.In formal service, the finger bowl is most commonly brought to the table with the dessert plate; there is a linen doily under the bowl, and the dessert fork and spoon are placed on either side of the bowl. The arrangement of plates and flatware are set before each guest, who then move the flatware to the sides of the dessert plate and the finger bowl with the doily to the upper left of the plate. \"This is the only time during a formal meal that a guest takes part in placing the appointments for a course\". If a separate fruit course follows dessert, the finger bowl is  brought in with the fruit plate and fruit fork and knife, arranged as they would be on a dessert plate. Alternatively, a full array of dessert and fruit dishes can be brought out at once, with the finger bowl on top of the dessert plate with its fork and spoon, and the dessert plate on top of the fruit plate, with doilies between each dish; in this type of service, the fruit fork and knife are brought out after the dessert plates and silver are cleared. After dessert (or after the fruit, if it is offered as a separate course), guests lightly dip their fingertips into the water, one hand at a time, and then wipe them on the napkin in their lap.", "after_revision": "Finger bowl from the 1880s A finger bowl is a bowl of water that dinner guests use for rinsing their fingers. In a formal meal , the finger bowl is brought to the table at the time of the last course of the meal (the dessert course or the fruit course) , and guests set it aside for use after the last course, just before leaving the table . In less formal service, the finger bowl may be presented after any course that involves finger food and may even be presented after more than one such course in a single meal. Formal service The finger bowl is typically delivered with the dessert plate. alt=Finger bowl|The finger bowl is moved to the upper left before dessert is served. After dessert, the guests dip their fingers into the bowl and dry them with the napkin.In formal service, the finger bowl is most commonly brought to the table with the dessert plate; there is a linen doily under the bowl, and the dessert fork and spoon are placed on either side of the bowl. The arrangement of plates and flatware are set before each guest, who then move the flatware to the sides of the dessert plate , and move the finger bowl with the doily to the upper left of the plate. \"This is the only time during a formal meal that a guest takes part in placing the appointments for a course\". If a separate fruit course follows dessert, the finger bowl is instead brought in with the fruit plate and fruit fork and knife, arranged as they would be on a dessert plate. Alternatively, a full array of dessert and fruit dishes can be brought out at once, with the finger bowl on top of the dessert plate with its fork and spoon, and the dessert plate on top of the fruit plate, with doilies between each dish; in this type of service, the fruit fork and knife are brought out after the dessert plates and silver are cleared. After dessert (or after the fruit, if that forms the last course), guests lightly dip their fingertips into the water, one hand at a time, and then wipe them on the napkin in their lap.", "edit_actions": [{"type": "D", "before": "served la russe", "after": null, "start_char_pos": 127, "end_char_pos": 142, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": ", and guests set it aside for use after the last course, just before leaving the table", "start_char_pos": 269, "end_char_pos": 269, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "and", "after": ", and move", "start_char_pos": 1031, "end_char_pos": 1034, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "instead", "start_char_pos": 1272, "end_char_pos": 1272, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "it is offered as a separate", "after": "that forms the last", "start_char_pos": 1769, "end_char_pos": 1796, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 109, 271, 447, 610, 698, 794, 903, 1097, 1208, 1376, 1615, 1730], "domain": "wiki"}
{"doc_id": "17468235", "revision_depth": 2, "before_revision": "In American restaurants in the mid-20th century, finger bowls were often presented at the end of the meal in high-end restaurants . In chop houses and lobster palaces of the time , finger bowls were often presented after any course with finger foods or drawn butter   .", "after_revision": "In high-end American restaurants of the early to mid-20th century, finger bowls were presented at the end of the meal . In chop houses and lobster palaces of the period , finger bowls were often presented after any course with finger foods or drawn butter . In some styles of informal and restaurant service, the dessert fork and spoon are set along the top edge of the plate when all the flatware is laid at the start of the meal. Near the end of the meal, the finger bowl is presented with the dessert plate , similar to formal service; and the guests set the bowl aside and move the fork and spoon to the sides of the plate for the dessert course. After dessert, the guests use the finger bowl before leaving the table .", "edit_actions": [{"type": "R", "before": "American restaurants in the", "after": "high-end American restaurants of the early to", "start_char_pos": 3, "end_char_pos": 30, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "D", "before": "often", "after": null, "start_char_pos": 67, "end_char_pos": 72, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "in high-end restaurants", "after": null, "start_char_pos": 106, "end_char_pos": 129, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "time", "after": "period", "start_char_pos": 174, "end_char_pos": 178, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 266, "end_char_pos": 266, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "In some styles of informal and restaurant service, the dessert fork and spoon are set along the top edge of the plate when all the flatware is laid at the start of the meal. Near the end of the meal, the finger bowl is presented with the dessert plate , similar to formal service; and the guests set the bowl aside and move the fork and spoon to the sides of the plate for the dessert course. After dessert, the guests use the finger bowl before leaving the table", "start_char_pos": 267, "end_char_pos": 267, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 131], "domain": "wiki"}
{"doc_id": "17608461", "revision_depth": 1, "before_revision": "Overview Wireless identity theft is a relatively new technique of gathering an individual' s personal information from RF-enabled cards carried on a person in their access control, credit, debit, or government issued identification cards.. Each of these cards carry a Radio frequency identification chip which responds to certain radio frequencies. When these \"tags\" come into contact with radio waves, they respond with a slightly altered signal. The response can contain encoded personal identifying information, including the card holder's name, address, Social Security Number, phone number, and pertinent account or employee information.", "after_revision": "Overview Wireless identity theft is a relatively new technique for gathering individuals' personal information from RF-enabled cards carried on a person in their access control, credit, debit, or government issued identification cards.. Each of these cards carry a radio frequency identification chip which responds to certain radio frequencies. When these \"tags\" come into contact with radio waves, they respond with a slightly altered signal. The response can contain encoded personally identifying information, including the card holder's name, address, Social Security Number, phone number, and pertinent account or employee information.", "edit_actions": [{"type": "R", "before": "of gathering an individual' s", "after": "for gathering individuals'", "start_char_pos": 63, "end_char_pos": 92, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Radio", "after": "radio", "start_char_pos": 268, "end_char_pos": 273, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "personal", "after": "personally", "start_char_pos": 481, "end_char_pos": 489, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}], "sents_char_pos": [0, 239, 348, 447], "domain": "wiki"}
{"doc_id": "1809568", "revision_depth": 1, "before_revision": "Early life and career Syed Kamal was born in Meerut, Uttar Pradesh, British India on 27 April 1937.Syed Kamal 1937 2009 The News International (newspaper), Published 2 October 2009, Retrieved 24 February 2019 Syed Kamal was a popular film star in the 1960s and the 1970s. Kamal, whose film Tauba (1963) became a success at the box office, had a striking resemblance to the Indian filmstar Raj Kapoor, and he was not evasive about this issue. He once said, \"People used to mistake me for Raj Kapoor while I was in Bombay and I enjoyed it\". In fact, according to Pakistani script writer Ali Sufyan Afaqi, Kamal used to practice Raj Kapoor's scenes in front of a mirror. Kamal's first appearance in his Pakistani movie by producer Shabab Keranvi Thandi Sarak (1957) was meant to resemble Raj Kapoor.  Syed Kamal won a Nigar Award (Special Award) for film Behan Bhai (1968)1968 Nigar Award for Syed Kamal for film Behan Bhai Retrieved 23 February 2019 and also a Nigar Award ('Lifetime Achievement Award') in 2000.  Achievements and awards Among his other activities, he founded the Karachi Film Guild and Pakistan Film and TV Academy. Special Nigar Award (Lifetime Achievement Award) in 2000Syed Kamal's Special Nigar Award (Lifetime Achievement Award) on Motion Pictures Archive of Pakistan URL) website Retrieved 15 February 2020 The following films were directed by Syed Kamal: Shehnai (1968) starring (Syed Kamal-Deeba) Syed Kamal in film Shehnai (1968) on Complete Index To World Film (C.I.T.W.F.) website Retrieved 23 February 2019 Insaan Aur Gadha (1973) starring (Syed Kamal-Nisho-Rozina) Yahan Say Wahan Tak starring (Syed Kamal-Mumtaz) Honeymoon (1970) starring (Syed Kamal-Husna-Rozina) Syed Kamal in film Honeymoon (1970) on C.I.T.W.F. website Retrieved 23 February 2019 Roop Behroop starring (producer-director-actor Syed Kamal opposite Nisho) Doosri Maan starring (Syed Kamal-Shamim Ara) Meray Bacchay Meri Aankhein starring (Syed Kamal-Deeba) Dard-E-Dil starring (Syed Kamal and Zeba) Aaliya starring (Syed Kamal and Shamim Ara) Jatt Kurrian Taun Darda (1976) A Punjabi film, starring Syed Kamal, Neelo and Aslam PervaizSyed Kamal in film Jatt Kurrian Taun Darda (1976) on Complete Index To World Film (C.I.T.W.F.) website Retrieved 23 February 2019 Ajj Dian Kurrian (1977) A Punjabi film, starring Syed Kamal and Neelo Kal De Munday (1978) Jat Kamala Gaya Dubai (1984) Siyasat (1986)", "after_revision": "Early life and career Syed Kamal was born in Meerut, Uttar Pradesh, British India on 27 April 1937.Syed Kamal 1937 2009 The News International (newspaper), Published 2 October 2009, Retrieved 3 July 2021 Syed Kamal was a popular film star in the 1960s and the 1970s. Kamal, whose film Tauba (1963) became a success at the box office, had a striking resemblance to the Indian filmstar Raj Kapoor, and he was not evasive about this issue. He once said, \"People used to mistake me for Raj Kapoor while I was in Bombay and I enjoyed it\". In fact, according to Pakistani script writer Ali Sufyan Afaqi, Kamal used to practice Raj Kapoor's scenes in front of a mirror. Kamal's first appearance in his Pakistani movie by producer Shabab Keranvi Thandi Sarak (1957) was meant to resemble Raj Kapoor. Achievements and awards Syed Kamal won a Nigar Award (Special Award) for film Behan Bhai (1968)1968 Nigar Award for Syed Kamal for film Behan Bhai Retrieved 23 February 2019 and also a Nigar Award ('Lifetime Achievement Award') in 2000. Special Nigar Award (Lifetime Achievement Award) in 2000. Among his other activities, he founded the Karachi Film Guild and Pakistan Film and TV Academy. The following films were directed by Syed Kamal: Shehnai (1968) starring (Syed Kamal-Deeba) Insaan Aur Gadha (1973) starring (Syed Kamal-Nisho-Rozina) Yahan Say Wahan Tak starring (Syed Kamal-Mumtaz) Honeymoon (1970) starring (Syed Kamal-Husna-Rozina) Roop Behroop starring (producer-director-actor Syed Kamal opposite Nisho) Doosri Maan starring (Syed Kamal-Shamim Ara) Meray Bacchay Meri Aankhein starring (Syed Kamal-Deeba) Dard-E-Dil starring (Syed Kamal and Zeba) Aaliya starring (Syed Kamal and Shamim Ara) Jatt Kurrian Taun Darda (1976) A Punjabi film, starring Syed Kamal, Neelo and Aslam Pervaiz Ajj Dian Kurrian (1977) A Punjabi film, starring Syed Kamal and Neelo Kal De Munday (1978) Jat Kamala Gaya Dubai (1984) Siyasat (1986)", "edit_actions": [{"type": "R", "before": "24 February 2019", "after": "3 July 2021", "start_char_pos": 192, "end_char_pos": 208, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Achievements and awards", "start_char_pos": 797, "end_char_pos": 797, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "A", "before": null, "after": "Special Nigar Award (Lifetime Achievement Award) in 2000.", "start_char_pos": 1011, "end_char_pos": 1011, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "Achievements and awards", "after": null, "start_char_pos": 1012, "end_char_pos": 1035, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "Special Nigar Award (Lifetime Achievement Award) in 2000Syed Kamal's Special Nigar Award (Lifetime Achievement Award) on Motion Pictures Archive of Pakistan URL) website Retrieved 15 February 2020", "after": null, "start_char_pos": 1132, "end_char_pos": 1328, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "D", "before": "Syed Kamal in film Shehnai (1968) on Complete Index To World Film (C.I.T.W.F.) website Retrieved 23 February 2019", "after": null, "start_char_pos": 1421, "end_char_pos": 1534, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "Syed Kamal in film Honeymoon (1970) on C.I.T.W.F. website Retrieved 23 February 2019", "after": null, "start_char_pos": 1695, "end_char_pos": 1779, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "PervaizSyed Kamal in film Jatt Kurrian Taun Darda (1976) on Complete Index To World Film (C.I.T.W.F.) website Retrieved 23 February 2019", "after": "Pervaiz", "start_char_pos": 2125, "end_char_pos": 2261, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 99, 271, 441, 538, 667, 796, 1010, 1131, 1499, 1558, 1593, 1642, 1659, 1694, 1730, 1853, 2226], "domain": "wiki"}
{"doc_id": "1809568", "revision_depth": 2, "before_revision": "Death and legacy Syed Kamal died on 1 October 2009 at Karachi after a lingering illness at age 72. He had been a heart patient for many years and had fallen accidentally at his home about a month ago. He was bedridden since his fall. His survivors include his wife, his son Ghalib Kamal and three daughters. Retrieved 23 February 2019Film star Syed Kamal passes away Dawn (newspaper), Published 2 Oct 2009, Retrieved 23 February 2019 Filmography Syed Kamal's notable films as an actor : Thandi Sarak (1957)Actor Kamal dies The Nation (newspaper), Published 2 October 2009, Retrieved 24 February 2019 Sawera (1959) Apna Paraya (1959) Banjaran (1962) Tauba (1963) Aashiana (1964) Aisa Bhi Hota Hai (1965) Behan Bhai (1968) Nai Laila Naya Majnoon (1969) Road To Swat (1970) Insan Aur Gadha (1973) The following films were directed by Syed Kamal: Shehnai (1968) starring (Syed Kamal-Deeba) Insaan Aur Gadha (1973) starring (Syed Kamal-Nisho-Rozina) Yahan Say Wahan Tak starring (Syed Kamal-Mumtaz) Honeymoon (1970) starring (Syed Kamal-Husna-Rozina) Roop Behroop starring (producer-director-actor Syed Kamal opposite Nisho) Doosri Maan starring (Syed Kamal-Shamim Ara) Meray Bacchay Meri Aankhein starring (Syed Kamal-Deeba) Dard-E-Dil starring (Syed Kamal and Zeba) Aaliya starring (Syed Kamal and Shamim Ara) Jatt Kurrian Taun Darda (1976) A Punjabi film, starring Syed Kamal, Neelo and Aslam Pervaiz Ajj Dian Kurrian (1977) A Punjabi film, starring Syed Kamal and Neelo Kal De Munday (1978) Jat Kamala Gaya Dubai (1984) Siyasat (1986)", "after_revision": "Death and legacy Syed Kamal died on 1 October 2009 at Karachi after a lingering illness at age 72. He had been a heart patient for many years and had fallen accidentally at his home about a month ago. He was bedridden since his fall. His survivors include his wife, his son Ghalib Kamal and three daughters. Retrieved 23 February 2019Film star Syed Kamal passes away Dawn (newspaper), Published 2 Oct 2009, Retrieved 3 July 2021 Filmography As an actor Thandi Sarak (1957)Actor Kamal dies The Nation (newspaper), Published 2 October 2009, Retrieved 24 February 2019 Sawera (1959) Apna Paraya (1959) Banjaran (1962) Tauba (1963) Aashiana (1964) Aisa Bhi Hota Hai (1965) Behan Bhai (1968) Nai Laila Naya Majnoon (1969) Road To Swat (1970) Insan Aur Gadha (1973) As a director Shehnai (1968) starring (Syed Kamal-Deeba) Insaan Aur Gadha (1973) starring (Syed Kamal-Nisho-Rozina) Yahan Say Wahan Tak starring (Syed Kamal-Mumtaz) Honeymoon (1970) starring (Syed Kamal-Husna-Rozina) Roop Behroop starring (producer-director-actor Syed Kamal opposite Nisho) Doosri Maan starring (Syed Kamal-Shamim Ara) Meray Bacchay Meri Aankhein starring (Syed Kamal-Deeba) Dard-E-Dil starring (Syed Kamal and Zeba) Aaliya starring (Syed Kamal and Shamim Ara) Jatt Kurrian Taun Darda (1976) A Punjabi film, starring Syed Kamal, Neelo and Aslam Pervaiz Ajj Dian Kurrian (1977) A Punjabi film, starring Syed Kamal and Neelo Kal De Munday (1978) Jat Kamala Gaya Dubai (1984) Siyasat (1986)", "edit_actions": [{"type": "R", "before": "23 February 2019", "after": "3 July 2021", "start_char_pos": 417, "end_char_pos": 433, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Syed Kamal's notable films as an actor :", "after": "As an actor", "start_char_pos": 446, "end_char_pos": 486, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "The following films were directed by Syed Kamal:", "after": "As a director", "start_char_pos": 794, "end_char_pos": 842, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 98, 200, 233, 307, 486, 842, 857, 885, 909, 944, 993, 1010, 1045], "domain": "wiki"}
{"doc_id": "18134326", "revision_depth": 1, "before_revision": "Integrated Business Planning (IBP) is a process for translating desired business outcomes into financial and operational resource requirements, with the overarching objective of maximizing profit and / or cash flow, while minimizing risk. The business outcomes, on which IBP processes focus, can be expressed in terms of the achievement of the following types of targets: History The term Integrated Business Planning is an outgrowth of Sales and Operations Planning, or S&OP , a term referring to processes that balance demand with manufacturing resources. There has been a lot of focus on Integrated Business Planning in the context of Sales and Operations Planning . Gartner refers to a 5-stage S&OP Maturity model wherein IBP is referred to as the Phased 4 & 5. Integrated Business Planning however is broader than S&OP. It is an approach that combines Enterprise Performance Management (EPM) and S&OP to provide incremental capabilities that neither provides individually. In so doing, IBP platforms address long-standing challenges that financial and operational professionals have struggled to overcome. The result: opportunities for step change improvements to how manufacturers plan, manage and govern their business. Here, the focus is on strengthening the financial integration and reconciliation of plans, as well as increasing the responsiveness of the supply chain using ad-hoc reports and what-if scenario analyses. To better predict customer demand, machine-learning technology helps to identify correlation patterns and automate the detection of demand changes.", "after_revision": "Integrated business planning (IBP) is a process for translating desired business outcomes into financial and operational resource requirements, with the overarching objective of maximizing profit and / or cash flow, while minimizing risk. The business outcomes, on which IBP processes focus, can be expressed in terms of the achievement of the following types of targets: History The term integrated business planning is an outgrowth of sales and operations planning ( S&OP ) a term referring to processes that balance demand with manufacturing resources. There has been a lot of focus on integrated business planning in the context of sales and operations planning . Gartner refers to a 5-stage S&OP maturity model wherein IBP is referred to as the Phased 4 & 5. Integrated Business Planning however is broader than S&OP. It is an approach that combines Enterprise Performance Management (EPM) and S&OP to provide incremental capabilities that neither provides individually. In so doing, IBP platforms address long-standing challenges that financial and operational professionals have struggled to overcome. The result: opportunities for step change improvements to how manufacturers plan, manage and govern their business. Here, the focus is on strengthening the financial integration and reconciliation of plans, as well as increasing the responsiveness of the supply chain using ad-hoc reports and what-if scenario analyses. To better predict customer demand, machine-learning technology helps to identify correlation patterns and automate the detection of demand changes.", "edit_actions": [{"type": "R", "before": "Business Planning", "after": "business planning", "start_char_pos": 11, "end_char_pos": 28, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "Integrated Business Planning", "after": "integrated business planning", "start_char_pos": 389, "end_char_pos": 417, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Sales and Operations Planning, or", "after": "sales and operations planning (", "start_char_pos": 437, "end_char_pos": 470, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": ",", "after": ")", "start_char_pos": 476, "end_char_pos": 477, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Integrated Business Planning", "after": "integrated business planning", "start_char_pos": 591, "end_char_pos": 619, "major_intent": "others", "raw_intents": ["fluency", "others", "others"]}, {"type": "R", "before": "Sales and Operations Planning", "after": "sales and operations planning", "start_char_pos": 638, "end_char_pos": 667, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "R", "before": "Maturity", "after": "maturity", "start_char_pos": 703, "end_char_pos": 711, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 238, 557, 824, 977, 1110, 1226, 1430], "domain": "wiki"}
{"doc_id": "18309949", "revision_depth": 1, "before_revision": "A discordant system or pattern does not correlate to the topography and geology of the area. Discordant drainage patterns are classified into two main types: antecedent and superimposed, while anteposition drainage patterns combine the two. In antecedent drainage, a river's vertical incision ability matches that of land uplift due to tectonic forces. Superimposed drainage develops differently: initially, a drainage system develops on a surface composed of 'younger' rocks, but due to denudative activities this surface of younger rocks is removed and the river continues to flow over a seemingly new surface, but one in fact made up of rocks of old geological formation. Dendritic drainage patterns Dendritic drainage pattern Dendritic drainage systems (from Greek , dendrites, \"of or like a tree\") are not straight and are the most common form of  drainage system. In this, there are many sub-tributaries (analogous to the twigs of a tree), which merge into tributaries of the main river (the branches and the trunk of the tree, respectively). They are seen to feed a river channel that matches and is strongly accordant to the overriding gradient of the land. Truly dendritic systems form in V-shaped valleys; as a result, the rock types must be quite  impervious and non-porous. Trellis drainage pattern The geometry of a trellis drainage system is similar to that of a common garden trellis. Along a strike valley, smaller tributaries feed into from the steep slopes of mountain sides . These tributaries enter the main river about perpendicular, causing a trellis-like appearance of the system. They form where hard and soft formations exist on both banks of the main river, and are reflective of height, accentuated by erosion. Trellis drainage is characteristic of folded mountains, such as the Appalachian Mountains in North America and in the north part of Trinidad.Ritter, Michael E., The Physical Environment: an Introduction to Physical Geography. 2006 Rectangular drainage pattern Rectangular drainage develops on rocks that are of approximately uniform resistance to erosion, but which have two directions of jointing at approximately right angles or 90 degrees. The joints are usually less resistant to erosion than the bulk rock so erosion tends to preferentially open the joints and streams eventually develop along the joints. The result is a stream system in which streams consist mainly of straight line segments with right angle bends and tributaries join larger streams at right angles. This pattern can be found with the Arun River in Nepal.", "after_revision": "A discordant system or pattern does not correlate to the topography and geology of the area. Discordant drainage patterns are classified into two main types: antecedent and superimposed, while ante position drainage patterns combine the two. In antecedent drainage, a river's vertical incision ability matches that of land uplift due to tectonic forces. Superimposed drainage develops differently: initially, a drainage system develops on a surface composed of 'younger' rocks, but due to denudation activities this surface of younger rocks is removed and the river continues to flow over a seemingly new surface, but one in fact made up of rocks of old geological formation. Dendritic drainage patterns Dendritic drainage pattern Dendritic drainage systems (from Greek , dendrites, \"of or like a tree\") are not straight and are the most common form of the drainage system. In this, there are many sub-tributaries (analogous to the twigs of a tree), which merge into tributaries of the main river (the branches and the trunk of the tree, respectively). They are seen to feed a river channel that matches and is strongly accordant to the overriding gradient of the land. Truly dendritic systems form in V-shaped valleys; as a result, the rock types must be quite obviously impervious and non-porous. Trellis drainage pattern The geometry of a trellis drainage system is similar to that of a common garden trellis. Along a strike valley, smaller tributaries feed into the steep slopes of mountainsides . These tributaries enter the main river about perpendicular, causing a trellis-like appearance of the system. They form where hard and soft formations exist on both banks of the main river, and are reflective of height, accentuated by erosion. Trellis drainage is characteristic of folded mountains, such as the Appalachian Mountains in North America and in the north part of Trinidad.Ritter, Michael E., The Physical Environment: an Introduction to Physical Geography. 2006 Rectangular drainage pattern Rectangular drainage develops on rocks that are of approximately uniform resistance to erosion, but which have two directions of jointing at approximately right angles or 90 degrees. The joints are usually less resistant to erosion than the bulk rock so erosion tends to preferentially open the joints and streams eventually develop along the joints. The result is a stream system in which streams consist mainly of straight line segments with right-angle bends and tributaries join larger streams at right angles. This pattern can be found with the Arun River in Nepal.", "edit_actions": [{"type": "R", "before": "anteposition", "after": "ante position", "start_char_pos": 193, "end_char_pos": 205, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "denudative", "after": "denudation", "start_char_pos": 488, "end_char_pos": 498, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 852, "end_char_pos": 852, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "A", "before": null, "after": "obviously", "start_char_pos": 1258, "end_char_pos": 1258, "major_intent": "style", "raw_intents": ["style", "meaning-changed", "style"]}, {"type": "D", "before": "from", "after": null, "start_char_pos": 1453, "end_char_pos": 1457, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "mountain sides", "after": "mountainsides", "start_char_pos": 1478, "end_char_pos": 1492, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "right angle", "after": "right-angle", "start_char_pos": 2442, "end_char_pos": 2453, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 92, 240, 352, 674, 869, 1048, 1165, 1215, 1285, 1399, 1494, 1603, 1737, 1879, 1963, 2180, 2348, 2512], "domain": "wiki"}
{"doc_id": "18309949", "revision_depth": 2, "before_revision": "An example of an integrated drainage is the watershed of the Rio Grande River. The sedimentary basins forming the modern Rio Grande Valley were not integrated into a single river system draining into the Gulf of Mexico until relatively recent geologic time. Instead, the basins formed by the opening of the Rio Grande rift were initially bolsons, with no external drainage and a central playa. An axial river existed in the Espanola Basin as early as 13 million years ago, reaching the Santo Domingo Basin by 6.9 million years ago. However, at this time, the river drained into a playa in the southern Albuquerque Basin where it deposited the Popotosa Formation. The upper reach of this river corresponded to the modern Rio Chama, but by 5 million years ago, an ancestral Rio Grande draining the eastern San Juan Mountains had joined the ancestral Rio Chama. The ancestral Rio Grande progressively integrated basins to the south, reaching the Palomas basin by 4.5 million years  , the Mesilla basin by 3.1 million years, to Texas by 2.06 million years, and finally joining the Pecos River at 800,000 years to drain into the Gulf of Mexico. Volcanism in the Taos Plateau reduced drainage from the San Luis Basin until a spillover event 440,000 years ago that drained Lake Alamosa and fully reintegrated the San Luis Basin into the Rio Grande watershed .", "after_revision": "An example of an integrated drainage is the area drained by the Rio Grande River. The sedimentary basins forming the modern Rio Grande Valley were not integrated into a single river system draining into the Gulf of Mexico until relatively recent geologic time. Instead, the basins formed by the opening of the Rio Grande rift were initially bolsons, with no external drainage and a central playa. An axial river existed in the Espanola Basin as early as 13 million years ago, reaching the Santo Domingo Basin by 6.9 million years ago. However, at this time, the river drained into a playa in the southern Albuquerque Basin where it deposited the Popotosa Formation. The upper reach of this river corresponded to the modern Rio Chama, but by 5 million years ago, an ancestral Rio Grande draining the eastern San Juan Mountains had joined the ancestral Rio Chama. The ancestral Rio Grande progressively integrated basins to the south, reaching the Palomas basin by 4.5 million years ago , the Mesilla basin by 3.1 million years, to Texas by 2.06 million years, and finally joining the Pecos River at 800,000 years to drain into the Gulf of Mexico. Volcanism in the Taos Plateau reduced drainage from the San Luis basin until a spillover event 440,000 years ago that drained Lake Alamosa and fully reintegrated the San Luis basin into the Rio Grande basin .", "edit_actions": [{"type": "R", "before": "watershed of", "after": "area drained by", "start_char_pos": 44, "end_char_pos": 56, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "ago", "start_char_pos": 978, "end_char_pos": 978, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "R", "before": "Basin", "after": "basin", "start_char_pos": 1205, "end_char_pos": 1210, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Basin", "after": "basin", "start_char_pos": 1315, "end_char_pos": 1320, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "watershed", "after": "basin", "start_char_pos": 1341, "end_char_pos": 1350, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 78, 257, 393, 531, 662, 858, 1139], "domain": "wiki"}
{"doc_id": "1838496", "revision_depth": 1, "before_revision": "Derived characters In the science fiction themed Lost Universe, Kanzaka's another media franchise, the traits and personas of Lina and Gourry were mixed together to create its male protagonist, Kain Blueriver. Several different versions of Lina populate the Slayers world itself, due to various plots. In one story from the prequels, Lina and Naga are exposed to a mirror that creates an exact opposite of whoever looks in it. The wizard using it expects this to produce dark and violent version of the two that would side with him and destroy the originals, but the mirror copies turn out sweet, kind, generous, and extremely concerned about the feelings of each other and everyone around them. In their attempts to break the spell, Lina and Naga actually wind up creating dozens of such doppelgngers of themselves, which go on to found an entire village full of philanthropic Linas and Nagas. A slightly adapted version of this story, which resulted in the mirror being destroyed without creating more than the initial copies, served as the basis of the Slayers Special OAV episode \"Mirror Mirror\". In the Slayers TV series, Gourry and Lina are both the subject of an attempt at cloning, which spawns a number of miniature, super deformed versions of them both; the tiny clones fight to humorous effect, and after the mini Linas win, they all disintegrate. Even more, in this case mostly lifelike, clones of Lina were created by Xellos' master Zellas Metallium in the video game Slayers in which one of them, who believes herself to be the real Lina, is a playable character and the game's initial protagonist. In the film Slayers Great , a super deformed, giant, toylike fighting golem named \"Piko-Piko Lina-chan\" is constructed in  a distorted image of Lina's that emphasises her kawaii (cute) attributes, much to Lina's dismay and anger as she believed she had been chosen to model for its creation because of her (self-imagined) great beauty . It is also featured in the video game Slayers Royal 2.  The manga miniseries Slayers Light Magic features a young boy named Light Inverse who dreams to become a mage in a scientifically advanced futuristic world.", "after_revision": "Derived characters Several different versions of Lina populate the world of Slayers due to various plots. In one story from the prequels, Lina and Naga are exposed to a mirror that creates an exact opposite of whoever looks in it. The wizard using it expects this to produce dark and violent version of the two that would side with him and destroy the originals, but the mirror copies turn out sweet, kind, generous, and extremely concerned about the feelings of each other and everyone around them. In their attempts to break the spell, Lina and Naga actually wind up creating dozens of such doppelgngers of themselves, which go on to found an entire village full of philanthropic Linas and Nagas. A slightly adapted version of this story, which resulted in the mirror being destroyed without creating more than the initial copies, served as the basis of the Slayers Special OAV episode \"Mirror Mirror\". In the Slayers TV series, Gourry and Lina are both the subject of an attempt at cloning, which spawns a number of miniature, super deformed versions of them both; the tiny clones fight to humorous effect, and after the mini Linas win, they all disintegrate. Even more, in this case mostly lifelike, clones of Lina were created by Xellos' master Zellas Metallium in the video game Slayers in which one of them, who believes herself to be the real Lina, is a playable character and the game's initial protagonist. Furthermore , a super deformed, giant, toylike fighting golem named \"Piko-Piko Lina-chan\" is constructed in the film Slayers Great in a distorted image of Lina's that emphasises her kawaii (cute) attributes, much to Lina's dismay and anger as she believed she had been chosen to model for its creation because of her (self-imagined) great beauty ; it is also featured in the video game Slayers Royal 2. In the science fiction themed Lost Universe, Kanzaka's another media franchise, the traits and personas of Lina and Gourry were mixed together to create its male protagonist, Kain Blueriver. The manga miniseries Slayers Light Magic features a young boy named Light Inverse who dreams to become a mage in a scientifically advanced futuristic world.", "edit_actions": [{"type": "D", "before": "In the science fiction themed Lost Universe, Kanzaka's another media franchise, the traits and personas of Lina and Gourry were mixed together to create its male protagonist, Kain Blueriver.", "after": null, "start_char_pos": 19, "end_char_pos": 209, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "Slayers world itself,", "after": "world of Slayers", "start_char_pos": 258, "end_char_pos": 279, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "In the film Slayers Great", "after": "Furthermore", "start_char_pos": 1613, "end_char_pos": 1638, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "the film Slayers Great in", "start_char_pos": 1735, "end_char_pos": 1735, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". It", "after": "; it", "start_char_pos": 1948, "end_char_pos": 1952, "major_intent": "coherence", "raw_intents": ["fluency", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "In the science fiction themed Lost Universe, Kanzaka's another media franchise, the traits and personas of Lina and Gourry were mixed together to create its male protagonist, Kain Blueriver.", "start_char_pos": 2005, "end_char_pos": 2005, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 209, 301, 426, 695, 894, 1100, 1263, 1358, 1612, 1949, 2004], "domain": "wiki"}
{"doc_id": "18599045", "revision_depth": 1, "before_revision": " History Among the tribal names given in \" Ashkharatsuyts \" there is an ethnonym ashtigor, which is considered to be the name of the Digors . This fact, and other linguistic considerations, have led experts to believe that Digor dialect became separated from Proto-Ossetian during the Mongol conquests. In \"Ashkharatsuyts\" Ashtigors are given separately from Alans - the ancestors of present-day Ossetians. Also starting from the 18th century  the ethnonym digor is widely used by travelers and in Russian official documents. Based on these facts, Ossetian Soviet linguist-Iranist Vasily Abaev suppose that the Digors could be an iranized Circassian tribe and the first part of the name - dig- can have the same roots with the Circassian endonym Adyge and -or can be just a prefix for plurality as in many contemporary Caucasian languages. Digors make the majority of the Ossetians in Digoria - the western part of the North OssetiaAlania (Digorsky and Irafsky districts) and in Kabardino-Balkaria. In the beginning of the 19th century some families from Digoria resettled in Mozdoksky District and there are 2 large settlements of them. Digoria was annexed to the Russian Empire quite late compared to the rest of Ossetia. The Digors were converted to Sunni Islam in the 17-18th centuries  under the influence of the neighboring Kabarday people who introduced Islam to them. In the second half of the 19th century  large numbers of Muslim Digors emigrated to the Ottoman Empire. (see: Ossetians in Turkey) During World War II, North OssetiaAlania was occupied by German armies in 1942. While under Nazi occupation the Ossetians remained unmoved, after the Germans were forced out of the region  the Muslim Digors, like other Muslim peoples, were accused of Collaboration with the Germans and deported to Central Asia. Estimates say 50\\% of the Digors died during deportation. Their reputation was rehabilitated in the mid 1950s , and they were allowed to return to their homelands   .", "after_revision": "Name Scholars generally link the root dig- with the Circassian endonym A-dyg-e, where the suffix -or could be a mark of plurality as found in many contemporary Caucasian languages. Ossetian Soviet linguist Vasily Abaev conjectured that the Digors may be an Iranized Circassian tribe. History The early medieval Ashkharatsuyts makes mention of the \" nation of the Ash-Tigor Alans \" (azg Alanac' A-Tigor), or simply the \"Dikor nation\" (Dik'ori-n), which is generally regarded as an early reference to the Digor . This fact, and other linguistic considerations, have led scholars to believe that Digor dialect became separated from Proto-Ossetian during the Mongol conquests. Starting from the 18th century , the ethnonym digor became widely used by travelers and in Russian official documents. Digoria was annexed to the Russian Empire quite late compared to the rest of Ossetia. The Digors were converted to Sunni Islam in the 17-18th centuries , under the influence of the neighboring Kabarday people who introduced Islam to them. In the second half of the 19th century , large numbers of Muslim Digors emigrated to the Ottoman Empire. During World War II, North OssetiaAlania was occupied by German armies in 1942. After the Germans were forced out of the region , the Muslim Digors, like other Muslim peoples, were accused of Collaboration with the Germans and deported to Central Asia. Estimates say 50\\% of the Digors died during deportation. Their reputation was rehabilitated in the mid-1950s , and they were allowed to return to their homelands . Demographics Digors make the majority of the Ossetians in Digoria, the western part of the North OssetiaAlania (Digorsky and Irafsky districts), and in Kabardino-Balkaria. In the beginning of the 19th century some families from Digoria resettled in Mozdoksky District, where there are 2 large settlements of them .", "edit_actions": [{"type": "A", "before": null, "after": "Name Scholars generally link the root dig- with the Circassian endonym A-dyg-e, where the suffix -or could be a mark of plurality as found in many contemporary Caucasian languages. Ossetian Soviet linguist Vasily Abaev conjectured that the Digors may be an Iranized Circassian tribe.", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Among the tribal names given in", "after": "The early medieval Ashkharatsuyts makes mention of the", "start_char_pos": 9, "end_char_pos": 40, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Ashkharatsuyts", "after": "nation of the Ash-Tigor Alans", "start_char_pos": 43, "end_char_pos": 57, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "there is an ethnonym ashtigor, which is considered to be the name of the Digors", "after": "(azg Alanac' A-Tigor), or simply the \"Dikor nation\" (Dik'ori-n), which is generally regarded as an early reference to the Digor", "start_char_pos": 60, "end_char_pos": 139, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "others", "clarity"]}, {"type": "R", "before": "experts", "after": "scholars", "start_char_pos": 199, "end_char_pos": 206, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "In \"Ashkharatsuyts\" Ashtigors are given separately from Alans - the ancestors of present-day Ossetians. Also starting from the", "after": "Starting from the", "start_char_pos": 303, "end_char_pos": 429, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 443, "end_char_pos": 443, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "is", "after": "became", "start_char_pos": 463, "end_char_pos": 465, "major_intent": "clarity", "raw_intents": ["clarity", "others", "fluency"]}, {"type": "D", "before": "Based on these facts, Ossetian Soviet linguist-Iranist Vasily Abaev suppose that the Digors could be an iranized Circassian tribe and the first part of the name - dig- can have the same roots with the Circassian endonym Adyge and -or can be just a prefix for plurality as in many contemporary Caucasian languages.", "after": null, "start_char_pos": 526, "end_char_pos": 839, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "Digors make the majority of the Ossetians in Digoria - the western part of the North OssetiaAlania (Digorsky and Irafsky districts) and in Kabardino-Balkaria. In the beginning of the 19th century some families from Digoria resettled in Mozdoksky District and there are 2 large settlements of them.", "after": null, "start_char_pos": 840, "end_char_pos": 1137, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1290, "end_char_pos": 1290, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1415, "end_char_pos": 1415, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "(see: Ossetians in Turkey)", "after": null, "start_char_pos": 1480, "end_char_pos": 1506, "major_intent": "coherence", "raw_intents": ["style", "coherence", "coherence"]}, {"type": "R", "before": "While under Nazi occupation the Ossetians remained unmoved, after the", "after": "After the", "start_char_pos": 1587, "end_char_pos": 1656, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1695, "end_char_pos": 1695, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "mid 1950s", "after": "mid-1950s", "start_char_pos": 1919, "end_char_pos": 1928, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 1982, "end_char_pos": 1982, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "Demographics Digors make the majority of the Ossetians in Digoria, the western part of the North OssetiaAlania (Digorsky and Irafsky districts), and in Kabardino-Balkaria. In the beginning of the 19th century some families from Digoria resettled in Mozdoksky District, where there are 2 large settlements of them", "start_char_pos": 1983, "end_char_pos": 1983, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 141, 302, 406, 525, 839, 998, 1137, 1223, 1375, 1479, 1586, 1818, 1876], "domain": "wiki"}
{"doc_id": "18842359", "revision_depth": 4, "before_revision": "Ocean temperatures depends on the amount of solar radiation falling on its surface. In the tropics, the temperature of the surface layers can rise to over while near the poles the temperature in equilibrium with the sea ice is about . Deep seawater has a temperature between and in all parts of the globe. There is a continuous circulation of water in the oceans . An ocean current is a continuous, directed movement of seawater generated by a number of forces acting upon the water, including wind, the Coriolis effect, temperature and salinity differences. Ocean currents have different origins, such as tidesfor tidal currents, or wind and waves for surface currents . The global pattern of winds or atmospheric circulation creates a global pattern of ocean currents driven by the wind and the effect the circulation of the earth or the coriolis force. Theses major ocean currents include the Gulf Stream, Kuroshio current, Agulhas current and Antarctic Circumpolar Current. Collectively, currents move enormous amounts of water and heat around the globe. This circulation has important impacts on global climate and the uptake and redistribution of pollutants such as carbon dioxide by moving these contaminants from the surface into the deep ocean.", "after_revision": "Ocean temperatures depend on the amount of solar radiation reaching the ocean surface. In the tropics, surface temperatures can rise to over . Near the poles, where sea ice forms, the temperature in equilibrium is about . Deep seawater temperature is between and in all parts of the ocean. Water continuously circulates in the oceans creating ocean currents which are directed movements of seawater. Currents are generated by a number of forces acting upon the water, including wind, the Coriolis effect, temperature and salinity differences. Tidal currents originate from tides, while surface currents are caused by wind and waves . The global pattern of winds and atmospheric circulation creates a global pattern of ocean currents driven by the wind and the effect the circulation of the earth or the coriolis force. These major ocean currents include the Gulf Stream, Kuroshio current, Agulhas current and Antarctic Circumpolar Current. Collectively, currents move enormous amounts of water and heat around the globe. This circulation has important impacts on global climate and the uptake and redistribution of pollutants such as carbon dioxide by moving these contaminants from the surface into the deep ocean.", "edit_actions": [{"type": "R", "before": "depends", "after": "depend", "start_char_pos": 19, "end_char_pos": 26, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "falling on its", "after": "reaching the ocean", "start_char_pos": 60, "end_char_pos": 74, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the temperature of the surface layers", "after": "surface temperatures", "start_char_pos": 100, "end_char_pos": 137, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "while near the poles", "after": ". Near the poles, where sea ice forms,", "start_char_pos": 155, "end_char_pos": 175, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "D", "before": "with the sea ice", "after": null, "start_char_pos": 207, "end_char_pos": 223, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "has a temperature", "after": "temperature is", "start_char_pos": 249, "end_char_pos": 266, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "globe. There is a continuous circulation of water", "after": "ocean. Water continuously circulates", "start_char_pos": 299, "end_char_pos": 348, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": ". An ocean current is a continuous, directed movement of seawater", "after": "creating ocean currents which are directed movements of seawater. Currents are", "start_char_pos": 363, "end_char_pos": 428, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "Ocean currents have different origins, such as tidesfor tidal currents, or", "after": "Tidal currents originate from tides, while surface currents are caused by", "start_char_pos": 559, "end_char_pos": 633, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "for surface currents", "after": null, "start_char_pos": 649, "end_char_pos": 669, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "or", "after": "and", "start_char_pos": 700, "end_char_pos": 702, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "Theses", "after": "These", "start_char_pos": 856, "end_char_pos": 862, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}], "sents_char_pos": [0, 83, 234, 305, 364, 558, 671, 855, 977, 1058], "domain": "wiki"}
{"doc_id": "18967", "revision_depth": 1, "before_revision": "The flowering plants, also known as Angiospermae (), or Magnoliophyta (), are the most diverse group of land plants, with 64 orders, 416 families, approximately 13,000 known genera and 300,000 known species. Like gymnosperms, angiosperms are seed-producing plants . They are distinguished from gymnosperms by characteristics including flowers, endosperm within their seeds , and the production of fruits that contain the seeds. Etymologically, \"angiosperm\" literally means a plant that produces seeds within an enclosure; in other words, a fruiting plant. The term comes from the Greek words ('case') and ('seed') . The ancestors of flowering plants diverged from the common ancestor of all living gymnosperms during the Carboniferous, over 300 million years ago, with the earliest record of angiosperm pollen appearing around 134 million years ago. The first remains of flowering plants are known from 125 million years ago. They diversified extensively during the Early Cretaceous, became widespread by 120 million years ago, and replaced conifers as the dominant trees from 60 to 100  million years ago.", "after_revision": "The flowering plants, also known as Angiospermae (), or Magnoliophyta (), are the most diverse group of land plants, with 64 orders, 416 families, approximately 13,000 known genera and 300,000 known species. Like gymnosperms, angiosperms are seed-producing plants , but regarding seeds, they differ from gymnosperms in their production of three critical structures: flowers; endosperm within their seeds ; and fruits that encase their seeds. Etymologically, \"angiosperm\" comes from the Greek words ('case') and ('seed') ; the term means a plant that produces seeds encased within an enclosure, say a 'fruit'. In other words, a fruiting plant. The ancestors of flowering plants diverged from the common ancestor of all living gymnosperms during the Carboniferous, over 300 million years ago, with the earliest record of angiosperm pollen appearing around 134 million years ago. The first remains of flowering plants are known from 125 million years ago. They diversified extensively during the Early Cretaceous, became widespread by 120 million years ago, and replaced conifers as the dominant trees from 100 to 60 million years ago.", "edit_actions": [{"type": "R", "before": ". They are distinguished from gymnosperms by characteristics including flowers,", "after": ", but regarding seeds, they differ from gymnosperms in their production of three critical structures: flowers;", "start_char_pos": 264, "end_char_pos": 343, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": ", and the production of fruits that contain the", "after": "; and fruits that encase their", "start_char_pos": 373, "end_char_pos": 420, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "literally means a plant that produces seeds within an enclosure; in other words, a fruiting plant. The term", "after": null, "start_char_pos": 457, "end_char_pos": 564, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ".", "after": "; the term means a plant that produces seeds encased within an enclosure, say a 'fruit'. In other words, a fruiting plant.", "start_char_pos": 614, "end_char_pos": 615, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "60 to", "after": null, "start_char_pos": 1077, "end_char_pos": 1082, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "to 60", "start_char_pos": 1087, "end_char_pos": 1087, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 207, 265, 427, 521, 555, 615, 849, 925], "domain": "wiki"}
{"doc_id": "18967", "revision_depth": 2, "before_revision": "The flowering plants, also known as Angiospermae (), or Magnoliophyta (), are the most diverse group of land plants, with 64 orders, 416 families, approximately 13,000 known genera and 300,000 known species. Like gymnosperms, angiosperms are seed-producing plants , but regarding seeds, they differ from gymnosperms in their production of three critical structures: flowers; endosperm within their seeds ; and fruits that encase their seeds. Etymologically, \"angiosperm\"  comes from the Greek words ('case') and ('seed') ; the term means a plant that produces seeds encased within an enclosure, say a 'fruit'. In other words, a fruiting plant. The ancestors of flowering plants diverged from the common ancestor of all living gymnosperms during the Carboniferous, over 300 million years ago, with the earliest record of angiosperm pollen appearing around 134 million years ago. The first remains of flowering plants are known from 125 million years ago. They diversified extensively during the Early Cretaceous, became widespread by 120 million years ago, and replaced conifers as the dominant trees from 100 to 60  million years ago.", "after_revision": "The flowering plants, also known as Angiospermae (), or Magnoliophyta (), are the most diverse group of land plants, with 64 orders, 416 families, approximately 13,000 known genera and 300,000 known species. Like gymnosperms, angiosperms are seed-producing plants . They are distinguished from gymnosperms by characteristics including flowers, endosperm within their seeds , and the production of fruits that contain the seeds. Etymologically, \"angiosperm\" literally means a plant that produces seeds within an enclosure; in other words, a fruiting plant. The term comes from the Greek words ('case') and ('seed') . The ancestors of flowering plants diverged from the common ancestor of all living gymnosperms during the Carboniferous, over 300 million years ago, with the earliest record of angiosperm pollen appearing around 134 million years ago. The first remains of flowering plants are known from 125 million years ago. They diversified extensively during the Early Cretaceous, became widespread by 120 million years ago, and replaced conifers as the dominant trees from 60 to 100 million years ago.", "edit_actions": [{"type": "R", "before": ", but regarding seeds, they differ from gymnosperms in their production of three critical structures: flowers;", "after": ". They are distinguished from gymnosperms by characteristics including flowers,", "start_char_pos": 264, "end_char_pos": 374, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "; and fruits that encase their", "after": ", and the production of fruits that contain the", "start_char_pos": 404, "end_char_pos": 434, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": "literally means a plant that produces seeds within an enclosure; in other words, a fruiting plant. The term", "start_char_pos": 471, "end_char_pos": 471, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "; the term means a plant that produces seeds encased within an enclosure, say a 'fruit'. In other words, a fruiting plant.", "after": ".", "start_char_pos": 521, "end_char_pos": 643, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "100 to", "after": null, "start_char_pos": 1105, "end_char_pos": 1111, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "to 100", "start_char_pos": 1115, "end_char_pos": 1115, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 207, 374, 405, 441, 522, 609, 643, 877, 953], "domain": "wiki"}
{"doc_id": "20254389", "revision_depth": 1, "before_revision": "Smiths and miners within medieval society Metallurgists throughout medieval Europe were  free to move within different regions. German metallurgists in search of rich precious metal ores, for instance, took the leading part in mining and affected the course of metal production, not only in East and South Germany but in almost all Central Europe and the Eastern Alps. As mining gradually became a task for specialized craftsmen, miners moved in large groups and they formed settlements with their own customs close to mines. They were always welcomed by the regional authorities  since the latter were interested in increasing the revenue and the exploitation of the mineral-rich subsurface was quite profitable. The authorities claimed a part of the output, and smiths and miners were provided with land for cottages, mills, URLes, farming, and pasture, and they were allowed to use streams and lumber (Nef 1987, 706-715). In medieval societies, liberal and mechanical arts were considered as totally different from each other . Metallurgists, as all craftsmen and artisans, lacked the methodical intellectual background but they were the pioneers of causal thinking , based on empirical observation and experimentation (Zilsel 2000).", "after_revision": "Smiths and miners within medieval society Metallurgists throughout medieval Europe were generally free to move within different regions. German metallurgists in search of rich precious metal ores, for instance, took the leading part in mining and affected the course of metal production, not only in East and South Germany but in almost all Central Europe and the Eastern Alps. As mining gradually became a task for specialized craftsmen, miners moved in large groups and formed settlements with their own customs close to mines. They were always welcomed by the regional authorities , since the latter were interested in increasing the revenue and the exploitation of the mineral-rich subsurface was quite profitable. These authorities claimed a part of the output, and smiths and miners were provided with land for cottages, mills, URLes, farming, and pasture, and they were allowed to use streams and lumber (Nef 1987, 706-715). In medieval societies, liberal and mechanical arts were considered to be totally different disciplines . Metallurgists, as all craftsmen and artisans, almost always lacked the formal education that would inform a methodical intellectual background . However, they were the pioneers of causal thinking based on empirical observation and experimentation (Zilsel 2000).", "edit_actions": [{"type": "A", "before": null, "after": "generally", "start_char_pos": 88, "end_char_pos": 88, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "D", "before": "they", "after": null, "start_char_pos": 463, "end_char_pos": 467, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 580, "end_char_pos": 580, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "The", "after": "These", "start_char_pos": 714, "end_char_pos": 717, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "as totally different from each other", "after": "to be totally different disciplines", "start_char_pos": 992, "end_char_pos": 1028, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "lacked the", "after": "almost always lacked the formal education that would inform a", "start_char_pos": 1077, "end_char_pos": 1087, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "but", "after": ". However,", "start_char_pos": 1123, "end_char_pos": 1126, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1169, "end_char_pos": 1170, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}], "sents_char_pos": [0, 127, 368, 525, 713, 924, 1030], "domain": "wiki"}
{"doc_id": "20254389", "revision_depth": 2, "before_revision": "Early Middle Ages, 500-1000 AD The social, political, economic stagnation and decline that followed the Roman World affected Europe throughout the early medieval period, and had critical impact upon technological progress, trade and URLanization. Technological developments that affected the course of metal production were only feasible within a stable political environment, and this was not the case until the 9th century (Martinon-Torres & Rehren in press, a). During the first medieval centuries, the output of metal was in a steady decline and constraint in small scale activities. Miners adopted methods much less efficient than those of the Roman times. Ores were extracted only from shallow depths or from remnants of former abandoned mines. The vicinity of the mine to villages or towns was also a determinant factor when deciding about working on site , because of the high cost of material transportation (Martinon-Torres & Rehren in press, b). Apparently only the output of iron diminished less in relation to the other base and precious metals until the 8th century. This fact, correlated with the dramatic decrease in copper production, may indicate a possible displacement of copper and bronze artifacts by iron ones (Forbes 1957, 64; Bayley et al. 2008, 50). By the end of the 9th century, economic  and social conditions , that dictated the increased need of metal for agriculture, arms, stirrups and decoration, started to favor metallurgy and a slow but steady general progress was noted . Starting from the reign of the emperor Otto I in the 960s, smelting sites were multiplied. New mines were discovered and exploited, like the well-known Mines of Rammelsberg, close to the town of Goslar in the Harz Mountains. Open-cast mining and metallurgical activities were mostly concentrated in the Eastern Alps, Saxony, Bohemia, Tuscany, the Rhineland, Gaul  and Spain (Nef 1987). French, Flemish, but mainly German miners and metallurgists were the generators of metal production  .", "after_revision": "Early Middle Ages, 500-1000 AD The social, political, economic stagnation and decline that followed the Roman World affected Europe throughout the early medieval period, leaving a critical impact upon technological progress, trade and URLanization. Technological developments that affected the course of metal production were only feasible within a stable political environment, and this was not the case until the 9th century (Martinon-Torres & Rehren in press, a). During the first medieval centuries, the output of metal was in a steady decline and constraint in small scale activities. Miners adopted methods much less efficient than those of the Roman times. Ores were extracted only from shallow depths or from remnants of former abandoned mines. The vicinity of the mine to villages or towns was also a determinant factor when deciding about working on site due to the high cost of material transportation (Martinon-Torres & Rehren in press, b). Only the output of iron diminished less in relation to the other base and precious metals until the 8th century. This fact, correlated with the dramatic decrease in copper production, may indicate a possible displacement of copper and bronze artifacts by iron ones (Forbes 1957, 64; Bayley et al. 2008, 50). By the end of the 9th century, economic , and social conditions dictated a greater need of metal for agriculture, arms, stirrups , and decoration. Consequently, conditions began to favor metallurgy and a slow but steady general progress developed . Starting from the reign of the emperor Otto I in the 960s, smelting sites were multiplied. New mines were discovered and exploited, like the well-known Mines of Rammelsberg, close to the town of Goslar in the Harz Mountains. Open-cast mining and metallurgical activities were mostly concentrated in the Eastern Alps, Saxony, Bohemia, Tuscany, the Rhineland, Gaul , and Spain (Nef 1987). It was mostly mainly German miners and metallurgists were the generators of metal production , but the French and Flemish made contributions to the developments .", "edit_actions": [{"type": "R", "before": "and had", "after": "leaving a", "start_char_pos": 170, "end_char_pos": 177, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "style"]}, {"type": "R", "before": ", because of", "after": "due to", "start_char_pos": 863, "end_char_pos": 875, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "Apparently only", "after": "Only", "start_char_pos": 957, "end_char_pos": 972, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1316, "end_char_pos": 1316, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": ", that dictated the increased", "after": "dictated a greater", "start_char_pos": 1339, "end_char_pos": 1368, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "and decoration, started", "after": ", and decoration. Consequently, conditions began", "start_char_pos": 1415, "end_char_pos": 1438, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "was noted", "after": "developed", "start_char_pos": 1498, "end_char_pos": 1507, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1873, "end_char_pos": 1873, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "French, Flemish, but", "after": "It was mostly", "start_char_pos": 1896, "end_char_pos": 1916, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": ", but the French and Flemish made contributions to the developments", "start_char_pos": 1996, "end_char_pos": 1996, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}], "sents_char_pos": [0, 246, 464, 587, 661, 750, 956, 1080, 1250, 1275, 1600, 1734, 1895], "domain": "wiki"}
{"doc_id": "202898", "revision_depth": 1, "before_revision": "The exosphere is the outermost layer of Earth's atmosphere (i.e. the upper limit of the atmosphere). It extends from the exobase, which is located at the top of the thermosphere at an altitude of about 700 km above sea level, to about 10,000 km (6,200 mi; 33,000,000 ft)  where it merges into the solar wind. The exosphere is located too far above Earth for any meteorological phenomena to be possible. However,  the aurora borealis  and aurora australis  sometimes occur in the lower part of the exosphere, where they overlap into the thermosphere. The exosphere contains many of the satellites orbiting Earth.", "after_revision": "The exosphere is the outermost layer of Earth's atmosphere (i.e. the upper limit of the atmosphere). It extends from the thermopause, at the top of the thermosphere at an altitude of about 700 km above sea level, to about 10,000 km (6,200 mi; 33,000,000 ft) , where it merges into the solar wind. The exosphere is too far above Earth for meteorological phenomena to be possible. However, Earth's auroras\u2014 the aurora borealis (northern lights) and aurora australis (southern lights)\u2014 sometimes occur in the lower part of the exosphere, where they overlap into the thermosphere. The exosphere contains many of the artificial satellites that orbit Earth.", "edit_actions": [{"type": "R", "before": "exobase, which is located", "after": "thermopause,", "start_char_pos": 121, "end_char_pos": 146, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "others"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 271, "end_char_pos": 271, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "located", "after": null, "start_char_pos": 326, "end_char_pos": 333, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "any", "after": null, "start_char_pos": 358, "end_char_pos": 361, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "Earth's auroras\u2014", "start_char_pos": 412, "end_char_pos": 412, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "(northern lights)", "start_char_pos": 433, "end_char_pos": 433, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "(southern lights)\u2014", "start_char_pos": 455, "end_char_pos": 455, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "satellites orbiting", "after": "artificial satellites that orbit", "start_char_pos": 585, "end_char_pos": 604, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 100, 255, 308, 402, 549], "domain": "wiki"}
{"doc_id": "20529614", "revision_depth": 1, "before_revision": "Post-trial commutation of death sentence  Peterson maintains his innocence. Peterson was granted post-conviction relief and re-sentenced to  120 years in prison December 12, 2004. His sentence was later commuted to 120 years in prison. He is scheduled for release on April 16th, 2051 according to the Indiana Department of Corrections .", "after_revision": "Reversal of death sentence and resentencing Peterson maintains his innocence. Peterson was granted post-conviction relief and re-sentenced to a total of 120 years in prison December 12, 2004. He is scheduled for release on April 16th, 2051 according to the Indiana Department of Correction .", "edit_actions": [{"type": "R", "before": "Post-trial commutation", "after": "Reversal", "start_char_pos": 0, "end_char_pos": 22, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "and resentencing", "start_char_pos": 41, "end_char_pos": 41, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "A", "before": null, "after": "a total of", "start_char_pos": 140, "end_char_pos": 140, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "His sentence was later commuted to 120 years in prison.", "after": null, "start_char_pos": 180, "end_char_pos": 235, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Corrections", "after": "Correction", "start_char_pos": 323, "end_char_pos": 334, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 75, 179, 235], "domain": "wiki"}
{"doc_id": "21045681", "revision_depth": 2, "before_revision": "Gesture music Sofia Gubaidulina Silence in music happens when the music stops during a performance. It is sometimes replaced by gesture music. In his Sofia Gubaidulina biography,Michael Kurtz Sofia Gubaidulina: A Biography, p. 184, english translation Indiana University Press, 2007 (written 2001) Michael Kurtz mentions the silent solo performance by the conductor included in ...Stimmenn... verstummen..., an orchestral work from 1986.  Milan Knizak From 1960, the International Fluxus Movement created a number of Events or Verbal Pieces, whose temporal structures were typically vague so as to be sometimes without beginning nor end, with or without sound, with or without music. A remarkable example is that of Czech artist Milan Knizak's 1965 Snowstorm N1 whose score states: Paper gliders are distributed to an idle and waiting audience.  Helmut Lachenmann Lachenmann composed Salut Fr Caldwell for 2 guitars in 1977. The piece includes silent moments when \"the players silent motions and gestures created a space of unheard music \". Michael KurtzMichael Kurtz Sofia Gubaidulina: A Biography, p. 192, English translation Indiana University Press, 2007 (written 2001) Takehisa Kosugi In 1963 Takehisa Kosugi composed for Fluxus 1 a musical piece called Theatre Music in the form of a rectangle of cardstock that bore the trace of a spiral of moving feet. This was paired with the instructions: \"Keep walking intently\".Charles Mereweather ed., Art Anti-Art Non-Art, Getty Research Institute, 2007, p. 21 Juan Mara Solare His work Gestenstcke (2008) is a collection of five pieces for 4 performers in which a musical structure is used to put order in non-sounding elements, concretely gestures. For instance, the first piece of the cycle is a canon of gestures. Premiere: University Bremen, Ensemble Neues Musiktheater, June 12, 2008. Another non-sounding piece is his conceptual work called Tense Atmosphere a graphic score which consists of a silence with a sforzato sign (2013). The optophonic piano Wladimir Baranoff-RossinSee Baranoff-Rossin 's official web page started building his optophonic piano in the 1915s. A set of painted glass discs are rotated via the small keyboard. Light is projected through the discs onto the wall. The player can control intensity of light and speed of rotation. 6 or 8 keys of the 3-octaves keyboard are devoted to colored discs. It is not clear what kind of sound the keyboard is able to produce. Oscillator frequencies (as stated in some articles) are rather unlikely between 1915-1920. More probably the light show was performed with piano accompaniment, maybe performed on the reduced keyboard of the Optosonic Piano, akin to a toy piano. Unfinished/aborted music At one point, the music exists in the composer's mind. In 1928, Edgard Varse started working on an opera called L'Astronome based on North American Indian legends, a project he never completed and destroyed the drafts. In 1932, he asked Antonin Artaud to write the libretto of a large scale oratorio, Il n'ya plus de firmament (There is no longer any firmament). In his book Phantasmatic Radio Allen S. Weiss translates the beginning of Artaud's text: {{cquote|Darkness. Explosions in this darkness. Harmonies suddenly broken off. Harsh sounds. Depressurized soundings. The music will give the impression of a distant cataclysm and will fill the room, falling as from vertiginous heights. Chords will originate in the sky and then deteriorate, going from one extreme to the other. Sounds will fall as if from very high, then suddenly stop and spread out in bursts, forming vaults and parasols. Tiers of sounds... in Phantasmatic Radio Allen S. Weiss, Phantasmatic Radio, Duke University Press, 1995, page 42}} The piece was never completed and Varse turned to other projects, including a radiophonic work involving various synchronised choirs located in different places of the world. He never found the technology for it. In 1948, Artaud insisted on having noise sounds included in his Pour en finir avec le judgement de Dieu (To have done with the judgment of God). This was done in the Radiodiffusion Television Franaise studios where Pierre Schaeffer was working at the time. Silent music Silent Music (1994)Bernd Schulz (editor) Silent Music - zwischen Klangkunst und Akustikdesign/Between Sound Art and Acoustic Design, book+CD, 1999 is an installation work by Canadian artist Robin Minard (b1953) with several hundred wall-mounted piezo loudspeakers and four-channel audio. The miniature loudspeakers are displayed in plant-like shapes on the walls of public spaces. They reproduce the minuscule ambient sounds of the public space where they are installed or play random synthetic sounds at barely audible volume. Minard deals with remote sounds best suited to intimate listening. In 2006 he created 'A voir en silence', a small artist book with loudspeakers and hand-written text.Susan Meinhardt Silent Music/Adelaide Festival DVD, Australia, 2007 Discursive music Marcel Proust The Vinteuil Sonata is an imaginary violin and piano sonata by fictitious composer Vinteuil recurring several times in Marcel Proust's la recherche du temps perdu (In Search of Lost Time), particularly in Un amour de Swann (1913). In the latter volume, Charles Swann associates strong emotions and memories to the melody composed by Vinteuil. The French composer Reynaldo Hahn noticed how much Marcel Proust \"vivait la musique de son temps\" (experienced contemporary music).See Adrien De Vries' article : Sources de la Sonate de Vinteuil, April 2007 (in French) For example, Proust immediately praised and enjoyed Debussy's 1902 Pellas et Mlisande opera. Critics disagree on which composer inspired the sonata. Possibly Gabriel FaurSee Margaret Hunter and Abigail Al-Doory's article: L'homme et la Musique Dans Un Amour de Swann, 1996 or Csar Franck. In Les plaisirs et les jours (1896), Proust focusses on Hans Sachs's monologue from Wagner's Die Meistersinger von Nrnberg, act 2. In Jean Santeuil (1952), a Camille Saint-Sans sonata for violin and piano (Op. 75, 1885) plays a key role and is presumably the model for the Sonate de Vinteuil. In Un amour de Swann  , the Vinteuil Sonata is played during evenings at the Verdurins' by pianist Dechambre. The main character's emotions are mirrored by Proust's musical reminiscences. Walter Marchetti A good example of imaginary music can be found in a Walter Marchetti poem where he mentions a Juan Hidalgo imaginary composition (both Hidalgo and Marchetti were members of the Spanish Zaj Group of Madrid in the late 1950s). <poem style=\"margin-left: 2em;\">Eat an iced popsicle and thus perform his free transcription, for only one performer, ofMusic For Five Dogs, an iced Popsicle and Six Male Performers by Juan Hidalgo. Walter Marchetti in A Zaj SamplerSomething Else Press, New York, 1967, reissued by Great Bear Pamphlets", "after_revision": "Gesture music Sofia Gubaidulina Silence in music happens when the music stops during a performance. It is sometimes replaced by gesture music. In his Sofia Gubaidulina biography,Michael Kurtz Sofia Gubaidulina: A Biography, p. 184, english translation Indiana University Press, 2007 (written 2001) Michael Kurtz mentions the silent solo performance by the conductor included in ...Stimmenn... verstummen..., an orchestral work from 1986. The symphony is notable for its careful and innovative use of silence. Though the eighth movement has the largest proportions of the work, the climax actually takes place in the ninth movement when the conductor motions before a silent orchestra. The motions the conductor makes are meant to make his hands move increasingly farther apart from each other according to the Fibonacci sequence. This \"conductor solo\" is repeated at the end of the work, when after the last note is sounded the conductor continues to motion for several minutes. Milan Knizak From 1960, the International Fluxus Movement created a number of Events or Verbal Pieces, whose temporal structures were typically vague so as to be sometimes without beginning nor end, with or without sound, with or without music. An example is that of Czech artist Milan Knizak's 1965 Snowstorm N1 whose score states: Paper gliders are distributed to an idle and waiting audience. What results is a snowstorm of quietly gliding paper airplanes as the audience returns them back and forth and so on. The exchange of sheets is experimentally beautiful, like the caring gesture toward the instrument as tended by Brecht.James Martin Harding and John Rouse (editors) Not the Other Avant-Garde: The Transnational Foundations of Avant-Garde Performance, University of Michigan Press, 2006, , Other examples Helmut Lachenmann composed Salut Fr Caldwell for 2 guitars in 1977. The piece includes silent moments when \"the players silent motions and gestures created a space of unheard music \". Michael Kurtz Sofia Gubaidulina: A Biography, p. 192, English translation Indiana University Press, 2007 (written 2001) In 1963 Takehisa Kosugi composed for Fluxus 1 a musical piece called Theatre Music in the form of a rectangle of cardstock that bore the trace of a spiral of moving feet. This was paired with the instructions: \"Keep walking intently\".Charles Mereweather ed., Art Anti-Art Non-Art, Getty Research Institute, 2007, p. 21 Juan Mara Solare 's work Gestenstcke (2008) is a collection of five pieces for 4 performers in which a musical structure is used to put order in non-sounding elements, concretely gestures. For instance, the first piece of the cycle is a canon of gestures. Premiere: University Bremen, Ensemble Neues Musiktheater, June 12, 2008. Another non-sounding piece is his conceptual work called Tense Atmosphere a graphic score which consists of a silence with a sforzato sign (2013). The optophonic piano Wladimir Baranoff-Rossin started building his optophonic piano in the 1915s. A set of painted glass discs are rotated via the small keyboard. Light is projected through the discs onto the wall. The player can control intensity of light and speed of rotation. 6 or 8 keys of the 3-octaves keyboard are devoted to colored discs. It is not clear what kind of sound the keyboard is able to produce. Oscillator frequencies are rather unlikely between 1915-1920. More probably the light show was performed with piano accompaniment, maybe performed on the reduced keyboard of the Optosonic Piano, akin to a toy piano. Unfinished/aborted music At one point, the music exists in the composer's mind. In 1928, Edgard Varse started working on an opera called L'Astronome based on North American Indian legends, a project he never completed and destroyed the drafts. In 1932, he asked Antonin Artaud to write the libretto of a large scale oratorio, Il n'ya plus de firmament (There is no longer any firmament). In his book Phantasmatic Radio Allen S. Weiss translates the beginning of Artaud's text: {{cquote|Darkness. Explosions in this darkness. Harmonies suddenly broken off. Harsh sounds. Depressurized soundings. The music will give the impression of a distant cataclysm and will fill the room, falling as from vertiginous heights. Chords will originate in the sky and then deteriorate, going from one extreme to the other. Sounds will fall as if from very high, then suddenly stop and spread out in bursts, forming vaults and parasols. Tiers of sounds... Allen S. Weiss, Phantasmatic Radio, Duke University Press, 1995, page 42}} The piece was never completed and Varse turned to other projects, including a radiophonic work involving various synchronised choirs located in different places of the world. He never found the technology for it. In 1948, Artaud insisted on having noise sounds included in his Pour en finir avec le judgement de Dieu (To have done with the judgment of God). This was done in the Radiodiffusion Television Franaise studios where Pierre Schaeffer was working at the time. Silent music Silent Music (1994)Bernd Schulz (editor) Silent Music - zwischen Klangkunst und Akustikdesign/Between Sound Art and Acoustic Design, book+CD, 1999 is an installation work by Canadian artist Robin Minard with several hundred wall-mounted piezo loudspeakers and four-channel audio. The miniature loudspeakers are displayed in plant-like shapes on the walls of public spaces. They reproduce the minuscule ambient sounds of the public space where they are installed or play random synthetic sounds at barely audible volume. Minard deals with remote sounds best suited to intimate listening. In 2006 he created 'A voir en silence', a small artist book with loudspeakers and hand-written text.Susan Meinhardt Silent Music/Adelaide Festival DVD, Australia, 2007 Discursive music Marcel Proust The Vinteuil Sonata is an imaginary violin and piano sonata by fictitious composer Vinteuil recurring several times in Marcel Proust's la recherche du temps perdu (In Search of Lost Time), particularly in Un amour de Swann (1913). In the latter volume, Charles Swann associates strong emotions and memories to the melody composed by Vinteuil. The French composer Reynaldo Hahn noticed how much Marcel Proust \"vivait la musique de son temps\" (experienced contemporary music).See Adrien De Vries' article : Sources de la Sonate de Vinteuil, April 2007 (in French) For example, Proust immediately praised and enjoyed Debussy's 1902 Pellas et Mlisande opera. Critics disagree on which composer inspired the sonata. Possibly Gabriel FaurSee Margaret Hunter and Abigail Al-Doory's article: L'homme et la Musique Dans Un Amour de Swann, 1996 or Csar Franck. In Les plaisirs et les jours (1896), Proust focusses on Hans Sachs's monologue from Wagner's Die Meistersinger von Nrnberg, act 2. In Jean Santeuil (1952), a Camille Saint-Sans sonata for violin and piano (Op. 75, 1885) plays a key role and is presumably the model for the Sonate de Vinteuil. In Un amour de Swann '' , the Vinteuil Sonata is played during evenings at the Verdurins' by pianist Dechambre. The main character's emotions are mirrored by Proust's musical reminiscences. Walter Marchetti An example of imaginary music can be found in a Walter Marchetti poem where he mentions a Juan Hidalgo imaginary composition (both Hidalgo and Marchetti were members of the Spanish Zaj Group of Madrid in the late 1950s). ", "edit_actions": [{"type": "A", "before": null, "after": "The symphony is notable for its careful and innovative use of silence. Though the eighth movement has the largest proportions of the work, the climax actually takes place in the ninth movement when the conductor motions before a silent orchestra. The motions the conductor makes are meant to make his hands move increasingly farther apart from each other according to the Fibonacci sequence. This \"conductor solo\" is repeated at the end of the work, when after the last note is sounded the conductor continues to motion for several minutes.", "start_char_pos": 438, "end_char_pos": 438, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "A remarkable", "after": "An", "start_char_pos": 684, "end_char_pos": 696, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "A", "before": null, "after": "What results is a snowstorm of quietly gliding paper airplanes as the audience returns them back and forth and so on. The exchange of sheets is experimentally beautiful, like the caring gesture toward the instrument as tended by Brecht.James Martin Harding and John Rouse (editors) Not the Other Avant-Garde: The Transnational Foundations of Avant-Garde Performance, University of Michigan Press, 2006, ,", "start_char_pos": 845, "end_char_pos": 845, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Helmut Lachenmann Lachenmann", "after": "Other examples Helmut Lachenmann", "start_char_pos": 846, "end_char_pos": 874, "major_intent": "fluency", "raw_intents": ["meaning-changed", "fluency", "fluency"]}, {"type": "R", "before": "Michael KurtzMichael", "after": "Michael", "start_char_pos": 1041, "end_char_pos": 1061, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "Takehisa Kosugi", "after": null, "start_char_pos": 1174, "end_char_pos": 1189, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "His", "after": "'s", "start_char_pos": 1526, "end_char_pos": 1529, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "D", "before": "Baranoff-RossinSee", "after": null, "start_char_pos": 2016, "end_char_pos": 2034, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "D", "before": "'s official web page", "after": null, "start_char_pos": 2051, "end_char_pos": 2071, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "(as stated in some articles)", "after": null, "start_char_pos": 2465, "end_char_pos": 2493, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "in Phantasmatic Radio", "after": null, "start_char_pos": 3625, "end_char_pos": 3646, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "(b1953)", "after": null, "start_char_pos": 4408, "end_char_pos": 4415, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "''", "start_char_pos": 6164, "end_char_pos": 6164, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "A good", "after": "An", "start_char_pos": 6348, "end_char_pos": 6354, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "<poem style=\"margin-left: 2em;\">Eat an iced popsicle and thus perform his free transcription, for only one performer, ofMusic For Five Dogs, an iced Popsicle and Six Male Performers by Juan Hidalgo. Walter Marchetti in A Zaj SamplerSomething Else Press, New York, 1967, reissued by Great Bear Pamphlets", "after": null, "start_char_pos": 6573, "end_char_pos": 6875, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 99, 142, 381, 437, 683, 844, 924, 1040, 1360, 1424, 1698, 1765, 1838, 1985, 2123, 2188, 2240, 2305, 2373, 2441, 2532, 2686, 2766, 2930, 3074, 3114, 3182, 3211, 3242, 3256, 3281, 3400, 3492, 3605, 3655, 3896, 3934, 4079, 4191, 4492, 4585, 4732, 4799, 4900, 5229, 5341, 5473, 5653, 5709, 5849, 5980, 6142, 6252, 6330, 6572, 6771]}
{"doc_id": "21492663", "revision_depth": 3, "before_revision": "Asbestos (pronounced: or ) is a naturally occurring fibrous silicate mineral. There are six types, all of which are composed of long and thin fibrous crystals, each fibre being composed of many microscopic \"fibrils\" that can be released into the atmosphere by abrasion and other processes. Asbestos is an excellent electrical insulator and is highly heat-resistant, so for many years it was used as a building material.Bureau of Naval Personnel, Basic Electricity. 1969: US Navy. However, it is now a notorious health and safety hazard and the use of asbestos as a building material is illegal in many countries. Inhalation of asbestos fibres can lead to various serious lung conditions, including asbestosis and cancer  . Archaeological studies have found evidence of asbestos being used as far back as the Stone Age to strengthen ceramic pots, but large-scale mining began at the end of the 19th century when manufacturers and builders began using asbestos for its desirable physical properties. Asbestos was widely used during the 20th century until the 1970s when public recognition of the health hazards of asbestos dust led to its prohibition in mainstream  construction and fireproofing in most countries. Despite this, and in part because the consequences of exposure can take decades to arise, at least 100,000 people are thought to die each year from diseases related to asbestos exposure.  Despite the severity of asbestos-related diseases, the material has been widely used all over the world, and most buildings constructed before the 1980s are thought to contain asbestos. Many developing countries still support the use of asbestos as a building material, and mining of asbestos is ongoing, with top producer Russia having  estimated production of 790,000 tonnes in 2020.", "after_revision": "Asbestos (pronounced: or ) is a naturally occurring fibrous silicate mineral. There are six types, all of which are composed of long and thin fibrous crystals, each fibre being composed of many microscopic \"fibrils\" that can be released into the atmosphere by abrasion and other processes. Inhalation of asbestos fibres can lead to various serious lung conditions, including asbestosis and cancer , so it is now notorious as a health and safety hazard . Archaeological studies have found evidence of asbestos being used as far back as the Stone Age to strengthen ceramic pots, but large-scale mining began at the end of the 19th century when manufacturers and builders began using asbestos for its desirable physical properties. Asbestos is an excellent electrical insulator and is highly heat-resistant, so for much of the 20th century it was widely used across the world as a building material until its effects on human health were recognized in the 1970s .Bureau of Naval Personnel, Basic Electricity. 1969: US Navy. Many modern buildings constructed before the 1980s are thought to contain asbestos. Inhalation of asbestos fibres can lead to various serious lung conditions, including asbestosis and cancer, so it is now notorious as a health and safety hazard. The use of asbestos for construction and fireproofing has been made illegal in many countries. Despite this, at least 100,000 people are thought to die each year from diseases related to asbestos exposure. In part, this is because many older buildings still contain asbestos; in addition, the consequences of exposure can take decades to arise. Many developing countries still support the use of asbestos as a building material, and mining of asbestos is ongoing, with top producer Russia having an estimated production of 790,000 tonnes in 2020.", "edit_actions": [{"type": "D", "before": "Asbestos is an excellent electrical insulator and is highly heat-resistant, so for many years it was used as a building material.Bureau of Naval Personnel, Basic Electricity. 1969: US Navy. However, it is now a notorious health and safety hazard and the use of asbestos as a building material is illegal in many countries.", "after": null, "start_char_pos": 290, "end_char_pos": 612, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": ", so it is now notorious as a health and safety hazard", "start_char_pos": 720, "end_char_pos": 720, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Asbestos", "after": "Asbestos is an excellent electrical insulator and is highly heat-resistant, so for much of the 20th century it", "start_char_pos": 998, "end_char_pos": 1006, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "during the 20th century until", "after": "across the world as a building material until its effects on human health were recognized in", "start_char_pos": 1023, "end_char_pos": 1052, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "when public recognition of the health hazards of asbestos dust led to its prohibition in mainstream", "after": ".Bureau of Naval Personnel, Basic Electricity. 1969: US Navy. Many modern buildings constructed before the 1980s are thought to contain asbestos.", "start_char_pos": 1063, "end_char_pos": 1162, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "Inhalation of asbestos fibres can lead to various serious lung conditions, including asbestosis and cancer, so it is now notorious as a health and safety hazard. The use of asbestos for", "start_char_pos": 1163, "end_char_pos": 1163, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in most", "after": "has been made illegal in many", "start_char_pos": 1194, "end_char_pos": 1201, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "and in part because the consequences of exposure can take decades to arise,", "after": null, "start_char_pos": 1227, "end_char_pos": 1302, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "In part, this is because many older buildings still contain asbestos; in addition, the consequences of exposure can take decades to arise.", "start_char_pos": 1400, "end_char_pos": 1400, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "D", "before": "Despite the severity of asbestos-related diseases, the material has been widely used all over the world, and most buildings constructed before the 1980s are thought to contain asbestos.", "after": null, "start_char_pos": 1401, "end_char_pos": 1586, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "an", "start_char_pos": 1738, "end_char_pos": 1738, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 77, 289, 419, 464, 479, 612, 997, 1212, 1399, 1586], "domain": "wiki"}
{"doc_id": "222168", "revision_depth": 1, "before_revision": "Sisulu was born in Ngcobo in the Union of South Africa. His mother, Alice Mase Sisulu, was a Xhosa domestic worker and his father, Albert Victor Dickinson, who was white. Dickinson worked in the Railway Department of the Cape Colony from 1903 to 1909 and was transferred to the Office of the Chief Magistrate in Umtata in 1910.David Beresford, \"Walter Sisulu\" (obituary), The Guardian, 7 May 2003. His mother was related to Evelyn Mase, Nelson Mandela's first wife. Dickinson did not play a part in his son's upbringing, and the boy and his sister, Rosabella, were raised by his mother's family, who were descended from the Thembu clan.Walter Sisulu Walter Sisulu Dickinson later went on to become the Attorney General of the Transvaal. He founded Sitha Investments in 1939. It was situated at Barclay Arcade between West Street and Commissioner Street in the business district of Johannesburg. Its objective is to help blacks and Indians to buy houses in Apartheid South Africa. During its operations, Sitha was the only black-owned  estate agency in South Africa. An adopted daughter, Beryl Rose Sisulu, served as ambassador from the Republic of South Africa to Norway.Female ambassadors luncheon. He joined the ANC in 1941. In 1943, together with Nelson Mandela and Oliver Tambo, he joined the ANC Youth League, founded by Anton Lembede, of which he was initially the treasurer. He later distanced himself from Lembede after Lembede, who died in 1947, had ridiculed his parentage (Sisulu was the son of a white foreman). Sisulu was a political networker and had a prominent planning role in the militant Umkhonto we Sizwe (\"Spear of the Nation\"). He was made secretary general of the ANC in 1949, displacing the more passive older leadership, and held that post until 1954. He also joined the South African Communist Party.", "after_revision": "Sisulu was born in Ngcobo in the Union of South Africa. His mother, Alice Mase Sisulu, was a Xhosa domestic worker and his father, Albert Victor Dickinson, who was white. Dickinson worked in the Railway Department of the Cape Colony from 1903 to 1909 and was transferred to the Office of the Chief Magistrate in Umtata in 1910.David Beresford, \"Walter Sisulu\" (obituary), The Guardian, 7 May 2003. Sisulu's mother was related to Evelyn Mase, Nelson Mandela's first wife. Dickinson did not play a part in his son's upbringing, and the boy and his sister, Rosabella, were raised by his mother's family, who were descended from the Thembu clan.Walter Sisulu Walter Sisulu Dickinson later went on to become the Attorney General of the Transvaal. He founded Sitha Investments in 1939. It was situated at Barclay Arcade between West Street and Commissioner Street in the business district of Johannesburg. Its objective was to help blacks and Indians buy houses in Apartheid South Africa. During its operations, Sitha was the only black-owned real estate agency in South Africa. An adopted daughter, Beryl Rose Sisulu, served as ambassador for the Republic of South Africa in Norway.Female ambassadors luncheon. He joined the ANC in 1941. In 1943, together with Nelson Mandela and Oliver Tambo, he joined the ANC Youth League, founded by Anton Lembede, of which he was initially the treasurer. He later distanced himself from Lembede after Lembede, who died in 1947, had ridiculed his parentage (Sisulu was the son of a white foreman). Sisulu was a political networker and had a prominent planning role in the militant Umkhonto we Sizwe (\"Spear of the Nation\"). He became secretary general of the ANC in 1949, displacing the more passive older leadership, in a post he held until 1954. He also joined the South African Communist Party.", "edit_actions": [{"type": "R", "before": "His", "after": "Sisulu's", "start_char_pos": 398, "end_char_pos": 401, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "is", "after": "was", "start_char_pos": 909, "end_char_pos": 911, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "D", "before": "to", "after": null, "start_char_pos": 939, "end_char_pos": 941, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "real", "start_char_pos": 1034, "end_char_pos": 1034, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "from", "after": "for", "start_char_pos": 1127, "end_char_pos": 1131, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "to", "after": "in", "start_char_pos": 1161, "end_char_pos": 1163, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "R", "before": "was made", "after": "became", "start_char_pos": 1653, "end_char_pos": 1661, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "and held that post", "after": "in a post he held", "start_char_pos": 1746, "end_char_pos": 1764, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}], "sents_char_pos": [0, 55, 170, 327, 397, 465, 636, 736, 774, 894, 979, 1065, 1171, 1199, 1226, 1381, 1523, 1649, 1776], "domain": "wiki"}
{"doc_id": "22444476", "revision_depth": 1, "before_revision": "Following the 1956 Loi Cadre, rudiments of internal rule were created with the Territorial Assembly of Niger elected by popular vote, but holding only advisory and administrative powers. In 1958 the French Union succeed the French Community On 25 August 1958, the Lieutenant Governor became High Commissioner of Niger, but remained Head of State of a quasi-independent state which controlled some purely internal administration. The Constitution of 25 February 1959 was created by and then ratified vote of the Constituent Assembly of Niger, a body created for this purpose from the elected Territorial Assembly of Niger  in December 1958. On 12 March 1959 the Constituent Assembly became the Legislative Assembly of Niger, with the head of government, Hamani Diori, retaining the title of President of the Council. Nominal executive powers were vested in the Assembly, with the constitution establishing elements, such as the Flag of Niger, the National anthem of Niger and the Coat of Arms of Niger, along with language on naming of political bodies, rights and powers which have been retained in subsequent texts.", "after_revision": "Following the 1956 Loi Cadre, rudiments of internal rule were created with the Territorial Assembly of Niger elected by popular vote, but holding only advisory and administrative powers. In 1958 the French Community succeed the French Union. On 25 August 1958, the Lieutenant Governor became High Commissioner of Niger, but remained Head of State of a quasi-independent state which controlled some purely internal administration. The Constitution of 25 February 1959 was ratified by the Constituent Assembly of Niger, a body created for this purpose from the Territorial Assembly of Niger elected in December 1958. On 12 March 1959 the Constituent Assembly became the Legislative Assembly of Niger, with the head of government, Hamani Diori, retaining the title of President of the Council. Nominal executive powers were vested in the Assembly, with the constitution establishing elements, such as the Flag of Niger, the National anthem of Niger and the Coat of Arms of Niger, along with language on naming of political bodies, rights and powers which have been retained in subsequent texts.", "edit_actions": [{"type": "R", "before": "Union", "after": "Community", "start_char_pos": 206, "end_char_pos": 211, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "Community", "after": "Union.", "start_char_pos": 231, "end_char_pos": 240, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "created by and then ratified vote of", "after": "ratified by", "start_char_pos": 470, "end_char_pos": 506, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "elected", "after": null, "start_char_pos": 583, "end_char_pos": 590, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "A", "before": null, "after": "elected", "start_char_pos": 621, "end_char_pos": 621, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 186, 428, 639, 815], "domain": "wiki"}
{"doc_id": "2307026", "revision_depth": 1, "before_revision": "The India Office was a British government department established in London in 1858 to oversee the administration, through a Viceroy and other officials, of the Provinces of British India. These territories comprised most of the modern-day nations of Bangladesh, Burma, India, and Pakistan, as well as Aden and other territories around the Indian Ocean. The department was headed by the Secretary of State for India, a member of the British cabinet, who was formally advised by the Council of India.Kaminsky, 1986. Upon the partition of British India in 1947 into the two new independent dominions of India and Pakistan, the India Office was closed down. Responsibility for the United Kingdom's relations with the two new countries was transferred to the Commonwealth Relations Office (formerly the Dominions Office).", "after_revision": "The India Office was a British government department established in London in 1858 to oversee the administration, through a Viceroy and other officials, of the Provinces of British India. These territories comprised most of the modern-day nations of Indian Subcontinent as well as Yemen and other territories around the Indian Ocean. The department was headed by the Secretary of State for India, a member of the British cabinet, who was formally advised by the Council of India.Kaminsky, 1986. Upon the independence of India in 1947 into the new independent dominion of India of the India Office was closed down. Responsibility for the United Kingdom's relations with the new country was transferred to the Commonwealth Relations Office (formerly the Dominions Office).", "edit_actions": [{"type": "R", "before": "Bangladesh, Burma, India, and Pakistan,", "after": "Indian Subcontinent", "start_char_pos": 250, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Aden", "after": "Yemen", "start_char_pos": 301, "end_char_pos": 305, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "partition of British", "after": "independence of", "start_char_pos": 523, "end_char_pos": 543, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "two new independent dominions of India and Pakistan,", "after": "new independent dominion of India of", "start_char_pos": 567, "end_char_pos": 619, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "two new countries", "after": "new country", "start_char_pos": 713, "end_char_pos": 730, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 187, 352, 498, 513, 653], "domain": "wiki"}
{"doc_id": "23362", "revision_depth": 1, "before_revision": "The history of Pakistan encompasses the region of the Indus Valley, which spans the northwestern expanse of the Indian subcontinent and the eastern Iranian plateau. The region served both as the fertile ground of a major civilisation and as the gateway of South Asia to Central Asia and the Near East. Quote: \"Numerous passageways through the western frontiers of the Indian subcontinent in modern Pakistan and Afghanistan served as migration routes to South Asia from the Iranian plateau and the Central Asian steppes. Prehistoric and protohistoric exchanges across the Hindu Kush, Karakoram, and Himalaya ranges demonstrate earlier precedents for routes through the high mountain passes and river valleys in later historical periods. Typological similarities between Northern Neolithic sites in Kashmir and Swat and sites in the Tibetan plateau and northern China show that 'Mountain chains have often integrated rather than isolated peoples.' Ties between the trading post of Shortughai in Badakhshan (northeastern Afghanistan) and the lower Indus valley provide evidence for long-distance commercial networks and 'polymorphous relations' across the Hindu Kush until c. 1800 B.C.' The Bactria-Margiana Archaeological Complex (BMAC) may have functioned as a 'filter' for the introduction of Indo-Iranian languages to the northwestern Indian subcontinent, although routes and chronologies remain hypothetical. (page 55)\" Quote: \"Here also, in ancient days, was the meeting-place of three great trade-routes , one, from Hindustan and Eastern India, which was to become the ' royal highway' described by Megasthenes as running from Pataliputra to the north-west of the Maurya empire; the second from Western Asia through Bactria, Kapisi and Pushkalavati and so across the Indus at Ohind to Taxila; and the third from Kashmir and Central Asia by way of the Srinagar valley and Baramula to Mansehra and so down the Haripur valley. These three trade-routes, which carried the bulk of the traffic passing by land between India and Central and Western Asia, played an all-important part in the history of Taxila. (page 1)\"", "after_revision": "The history of Pakistan for the period preceding the country's creation in 1947Pakistan was created as the Dominion of Pakistan on 14 August 1947 after the end of British rule in, and partition of British India. is shared with that of Afghanistan, India, and Iran. Spanning the western expanse of the Indian subcontinent and the eastern borderlands of the Iranian plateau, the region of present-day Pakistan served both as the fertile ground of a major civilization and as the gateway of South Asia to Central Asia and the Near East. Quote: \"Numerous passageways through the northwestern frontiers of the Indian subcontinent in modern Pakistan and Afghanistan served as migration routes to South Asia from the Iranian plateau and the Central Asian steppes. Prehistoric and protohistoric exchanges across the Hindu Kush, Karakoram, and Himalaya ranges demonstrate earlier precedents for routes through the high mountain passes and river valleys in later historical periods. Typological similarities between Northern Neolithic sites in Kashmir and Swat and sites in the Tibetan plateau and northern China show that 'Mountain chains have often integrated rather than isolated peoples.' Ties between the trading post of Shortughai in Badakhshan (northeastern Afghanistan) and the lower Indus valley provide evidence for long-distance commercial networks and 'polymorphous relations' across the Hindu Kush until c. 1800 B.C.' The Bactria-Margiana Archaeological Complex (BMAC) may have functioned as a 'filter' for the introduction of Indo-Iranian languages to the northwestern Indian subcontinent, although routes and chronologies remain hypothetical. (page 55)\" Quote: \"Here also, in ancient days, was the meeting-place of three great trade-routes , one, from Hindustan and Eastern India, which was to become the ` royal highway' described by Megasthenes as running from Pataliputra to the north-west of the Maurya empire; the second from Western Asia through Bactria, Kapisi and Pushkalavati and so across the Indus at Ohind to Taxila; and the third from Kashmir and Central Asia by way of the Srinagar valley and Baramula to Mansehra and so down the Haripur valley. These three trade-routes, which carried the bulk of the traffic passing by land between India and Central and Western Asia, played an all-important part in the history of Taxila. (page 1)\"", "edit_actions": [{"type": "R", "before": "encompasses the region of the Indus Valley, which spans the northwestern", "after": "for the period preceding the country's creation in 1947Pakistan was created as the Dominion of Pakistan on 14 August 1947 after the end of British rule in, and partition of British India. is shared with that of Afghanistan, India, and Iran. Spanning the western", "start_char_pos": 24, "end_char_pos": 96, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Iranian plateau. The region", "after": "borderlands of the Iranian plateau, the region of present-day Pakistan", "start_char_pos": 148, "end_char_pos": 175, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": "civilisation", "after": "civilization", "start_char_pos": 221, "end_char_pos": 233, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "western", "after": "northwestern", "start_char_pos": 343, "end_char_pos": 350, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "'", "after": "`", "start_char_pos": 1573, "end_char_pos": 1574, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}], "sents_char_pos": [0, 164, 301, 519, 735, 945, 1410, 1682, 1796, 1927, 2106], "domain": "wiki"}
{"doc_id": "23427012", "revision_depth": 1, "before_revision": "Another source of dispute has been the Baghdad stone i ption in a Turkish script , which some interpret saying Baba Nanak Fakir was there in 15111512 , while some state 15211522 (and that he lived in the Middle East for 11 years away from his family). Others, particularly Western scholars, argue that the stone inscription is from the 19th century and the stone is not a reliable evidence that Guru Nanak visited Baghdad in early 16th century. Moreover, beyond the stone, no evidence or mention of Guru Nanak's journey in the Middle East has been found in any other Middle Eastern textual or epigraphical records. Claims have been asserted of additional inscriptions, but no one has been able to locate and verify them.", "after_revision": "Another source of dispute has been the Baghdad stone , bearing an inscription in a Turkish script . Some interpret the inscription as saying Baba Nanak Fakir was there in 15111512 ; others read it as saying 15211522 (and that he lived in the Middle East for 11 years away from his family). Others, particularly Western scholars, argue that the stone inscription is from the 19th century and the stone is not a reliable evidence that Guru Nanak visited Baghdad in early 16th century. Moreover, beyond the stone, no evidence or mention of Guru Nanak's journey in the Middle East has been found in any other Middle Eastern textual or epigraphical records. Claims have been asserted of additional inscriptions, but no one has been able to locate and verify them.", "edit_actions": [{"type": "R", "before": "i ption", "after": ", bearing an inscription", "start_char_pos": 53, "end_char_pos": 60, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": ", which some interpret", "after": ". Some interpret the inscription as", "start_char_pos": 81, "end_char_pos": 103, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": ", while some state", "after": "; others read it as saying", "start_char_pos": 150, "end_char_pos": 168, "major_intent": "style", "raw_intents": ["style", "style", "coherence"]}], "sents_char_pos": [0, 251, 444, 614], "domain": "wiki"}
{"doc_id": "24319601", "revision_depth": 1, "before_revision": "House of Carlos Jose Gonzalez Llorente. Cry of Independence, Bogota. The Colombian Declaration of Independence refers to the events of July 20, 1810 , in Santa Fe de Bogota, in the Spanish colonial Viceroyalty of New Granada . They resulted in the establishment of a Junta de Santa Fe that day. The experience in self-government eventually led to the creation of the Republic of Colombia.(Note: The initial ambitious area, in accordance with the Viceroyalty of New Granada and Captaincy of Venezuela, included much more than current Colombia; to differentiate it , historians call this supra-nation: Republic of Gran Colombia.)  Political background And finally, Torres defended the right of the Viceroyalty of New Granada to establish a junta given the political circumstances. Although the draft expressed many of common sentiments of Criollos at the time and probably was discussed by prominent members of the capital's society, it was never adopted by the cabildo. It would be published for the first time only in 1832.", "after_revision": "House of Carlos Jose Gonzalez Llorente. Cry of Independence, Bogota. The Colombian Declaration of Independence occurred on July 20, 1810 when the Junta de Santa Fe was formed in Santa Fe de Bogota, the capital of the Spanish colonial Viceroyalty of New Granada , to govern the territory autonomously from Spain. The event inspired similar independence movements across Latin America, and triggered an almost decade-long rebellion culminating in the founding of the Republic of Gran Colombia, which spanned present-day Colombia, mainland Ecuador, Panama, and Venezuela, along with parts of northern Peru and northwestern Brazil.At the time, the state was known as Colombia; the term Gran Colombia is used historiographically to distinguish it from the current Republic of Colombia, which is also the official name of the former state. Although Gran Colombia would ultimately dissolve in 1831, it was for a time among the most powerful countries in the Western Hemisphere, and played an influential role in shaping the political development of other newly sovereign Latin American states. The modern nation-state of Colombia recognizes the event as its national independence day. Political background Torres defended the right of the Viceroyalty of New Granada to establish a junta given the political circumstances. Although the draft expressed many of common sentiments of Criollos at the time and probably was discussed by prominent members of the capital's society, it was never adopted by the cabildo. It would be published for the first time only in 1832.", "edit_actions": [{"type": "R", "before": "refers to the events of", "after": "occurred on", "start_char_pos": 111, "end_char_pos": 134, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ",", "after": "when the Junta de Santa Fe was formed", "start_char_pos": 149, "end_char_pos": 150, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in the", "after": "the capital of the", "start_char_pos": 174, "end_char_pos": 180, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ". They resulted in the establishment of a Junta de Santa Fe that day. The experience in self-government eventually led to the creation", "after": ", to govern the territory autonomously from Spain. The event inspired similar independence movements across Latin America, and triggered an almost decade-long rebellion culminating in the founding", "start_char_pos": 225, "end_char_pos": 359, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Colombia.(Note: The initial ambitious area, in accordance with the Viceroyalty of New Granada and Captaincy of Venezuela, included much more than current Colombia; to differentiate it , historians call this supra-nation: Republic of Gran Colombia.)", "after": "Gran Colombia, which spanned present-day Colombia, mainland Ecuador, Panama, and Venezuela, along with parts of northern Peru and northwestern Brazil.At the time, the state was known as Colombia; the term Gran Colombia is used historiographically to distinguish it from the current Republic of Colombia, which is also the official name of the former state.", "start_char_pos": 379, "end_char_pos": 627, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "Although Gran Colombia would ultimately dissolve in 1831, it was for a time among the most powerful countries in the Western Hemisphere, and played an influential role in shaping the political development of other newly sovereign Latin American states. The modern nation-state of Colombia recognizes the event as its national independence day.", "start_char_pos": 628, "end_char_pos": 628, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "And finally,", "after": null, "start_char_pos": 650, "end_char_pos": 662, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 39, 68, 226, 294, 388, 542, 778, 968], "domain": "wiki"}
{"doc_id": "24558035", "revision_depth": 2, "before_revision": "In the 1850s the term 'Aryan' was adopted as a racial category by French writer Arthur de Gobineau, who, through the later works of Houston Stewart Chamberlain, influenced the Nazi racial ideology. Under Nazi rule (19331945), the term applied to the inhabitants of Germany of non-Jewish origin; 'non-Aryans ' were discriminated against , then the target of systematic mass killings known as the Holocaust. The atrocities committed in the name of Aryanist supremacist ideologies have led academics to avoid the term 'Aryan', which has been replaced in most cases by 'Indo-Iranian', with only the South Asian branch still being called 'Indo-Aryan'.", "after_revision": "In the 1850s the term 'Aryan' was adopted as a racial category by French writer Arthur de Gobineau, who, through the later works of Houston Stewart Chamberlain, influenced the Nazi racial ideology. Under Nazi rule (19331945), the term applied to most inhabitants of Germany who were not Jews, Poles, Slavs, or Russians. Those classified as 'non-Aryans ,' especially Jews, were discriminated against before suffering the systematic mass killing known as the Holocaust. The atrocities committed in the name of Aryanist supremacist ideologies have led academics to avoid the term 'Aryan', which has been replaced in most cases by 'Indo-Iranian', with only the South Asian branch still being called 'Indo-Aryan'.", "edit_actions": [{"type": "R", "before": "the", "after": "most", "start_char_pos": 246, "end_char_pos": 249, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "of non-Jewish origin;", "after": "who were not Jews, Poles, Slavs, or Russians. Those classified as", "start_char_pos": 273, "end_char_pos": 294, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "'", "after": ",' especially Jews,", "start_char_pos": 307, "end_char_pos": 308, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": ", then the target of systematic mass killings", "after": "before suffering the systematic mass killing", "start_char_pos": 336, "end_char_pos": 381, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 197, 294, 405], "domain": "wiki"}
{"doc_id": "250641", "revision_depth": 1, "before_revision": "The role of the librarian has changed much over time, with the past century in particular bringing many new media and technologies into play. From the earliest libraries in the ancient world to the modern information hub, there have been keepers and disseminators of the information held in data stores. Roles and responsibilities vary widely depending on the type of library and specialist librarian, from academic libraries to public libraries, as well as the functions needed behind the scenes to maintain the collection and make it available to its users. The Great Library of Alexandria, created by Ptolemy I after the death of Alexander the Great in 323 BC, was created to house the entirety of Greek literature. It was notable for its famous librarians: Demetrius, Zenodotus, Eratosthenes, Apollonius, Aristophanes, Aristarchus, and Callimachus. These scholars contributed significantly to the collection and cataloging of the wide variety of scrolls in the library's collection. Most notably, Callimachus created what is considered to be the first subject catalogue of the library holdings, called the pinakes. The pinakes contained 120 scrolls arranged into ten subject classes; each class was then subdivided, listing authors alphabetically by titles. The librarians at Alexandria were considered the \"custodians of learning\". As the crisis escalated,  was a high demand for contact tracers, and the CDC had earlier named librarians as key public health staff to support COVID-19 case investigation and contact tracing, so many librarians and library staff volunteered to help with contact tracing. Librarians also supported their community in other ways  such as staffing non-emergency hotlines and manning shelters for the homeless, for which they were able to retain their income, while others were furloughed for a time. Additional responsibilities Experienced librarians may take administrative positions such as library or information center director or learning resource officer. Similar to the management of any URLanization, they are concerned with the long-term planning of the library, and its relationship with its URLanization (the city or county for a public library, the college/university for an academic library, or URLanization served by a special library). In smaller or specialized libraries, librarians typically perform a wide range of the different duties. upJustin Winsor, Librarian of Congress, c. 1885 Representative examples of librarian responsibilities: Researching topics of interest for their constituencies. Referring patrons to other URLanizations and government offices. Suggesting appropriate books (\"readers' advisory\") for children of different reading levels, and recommending novels for recreational reading. Reviewing books and journal databases Working with other URLanisations to establish continual, lifelong learning and further education initiatives Facilitating and promoting reading clubs. Developing programs for library users of all ages and backgrounds. Managing access to electronic information resources. Assessing library services and collections in order to best meet library users' needs. Building and maintaining collections to respond to changing community needs or demands Creating pathfinders Writing grants to gain funding for expanded program or collections Digitizing collections for online access Publishing articles in library science journals Answering incoming reference questions via telephone, postal mail, email, fax, and chat delivering arts and cultural activities to local communities Initiating and establishing creative digital activities to introduce children to coding, engineering and website building Marking promotion and advocacy of library services Assisting job seekers and local businesses Making and enforcing computer appointments on the public access Internet computers.\"The librarian's Internet survival guide: strategies for the high-tech reference desk\", Irene E. McDermott, Barbara E. Quint, p. 1-2, Information Today, Special library Special libraries can be described as libraries designed to perform some specific function to a particular set of people or URLanization  i.e. some include news, law, medical, theological, prison, corporate, museum, or any other type of library owned and operated by URLanization . They can be highly specialized, serving a discrete user group with a restricted collection area. In an increasingly global and virtual workplace, many special librarians may not even work in a library at all but instead manage and facilitate the use of electronic collections. Funding for special libraries varies widely. Librarians in some types of special libraries may be required to have additional training, such as a law degree for a librarian in an academic law library or appropriate subject degrees for subject specialties such as chemistry, engineering, etc. Many belong to the Special Libraries Association. There are also more specific associations such as the American Association of Law Libraries, Art Libraries Society of North America, the American Theological Library Association, the Medical Library Association, or the Visual Resources Association.", "after_revision": "The role of the librarian has changed much over time, with the past century in particular bringing many new media and technologies into play. From the earliest libraries in the ancient world to the modern information hub, there have been keepers and disseminators of the information held in data stores. Roles and responsibilities vary widely depending on the type of library , the specialty of the librarian, and the functions needed to maintain collections and make them available to its users. The Great Library of Alexandria, created by Ptolemy I after the death of Alexander the Great in 323 BC, was created to house the entirety of Greek literature. It was notable for its famous librarians: Demetrius, Zenodotus, Eratosthenes, Apollonius, Aristophanes, Aristarchus, and Callimachus. These scholars contributed significantly to the collection and cataloging of the wide variety of scrolls in the library's collection. Most notably, Callimachus created what is considered to be the first subject catalog of the library holdings, called the pinakes. The pinakes contained 120 scrolls arranged into ten subject classes; each class was then subdivided, listing authors alphabetically by titles. The librarians at Alexandria were considered the \"custodians of learning\". As the crisis escalated, there was a high demand for contact tracers, and the CDC had earlier named librarians as key public health staff to support COVID-19 case investigation and contact tracing, so many librarians and library staff volunteered to help with contact tracing. Librarians also supported their community in other ways , such as staffing non-emergency hotlines and manning shelters for the homeless, for which they were able to retain their income, while others were furloughed for a time. Additional responsibilities Experienced librarians may take administrative positions such as library or information center director or learning resource officer. Similar to the management of any URLanization, they are concerned with the long-term planning of the library, and its relationship with its URLanization (the city or county for a public library, the college/university for an academic library, or URLanization served by a special library). In smaller or specialized libraries, librarians typically perform a wide range of the different duties. upJustin Winsor, Librarian of Congress, c. 1885 Representative examples of librarian responsibilities: Researching topics of interest for their constituencies. Referring patrons to other URLanizations and government offices. Suggesting appropriate books (\"readers' advisory\") for children of different reading levels, and recommending novels for recreational reading. Reviewing books and journal databases Working with other URLanisations to establish continual, lifelong learning and further education initiatives Facilitating and promoting reading clubs. Developing programs for library users of all ages and backgrounds. Managing access to electronic information resources. Assessing library services and collections in order to best meet library users' needs. Building and maintaining collections to respond to changing community needs or demands Creating pathfinders Writing grants to gain funding for expanded program or collections Digitizing collections for online access Publishing articles in library science journals Answering incoming reference questions via telephone, postal mail, email, fax, and chat Delivering arts and cultural activities to local communities Initiating and establishing creative digital activities to introduce children to coding, engineering and website building Marking promotion and advocacy of library services Assisting job seekers and local businesses Making and enforcing computer appointments on the public access Internet computers.\"The librarian's Internet survival guide: strategies for the high-tech reference desk\", Irene E. McDermott, Barbara E. Quint, p. 1-2, Information Today, Special library Special libraries are libraries designed to perform some specific function for a particular set of people or URLanization , i.e. news, law, medical, theological, prison, corporate, or museum . They can be highly specialized, serving a discrete user group with a restricted collection area. In an increasingly global and virtual workplace, many special librarians may not even work in a library at all but instead manage and facilitate the use of electronic collections. Funding for special libraries varies widely. Librarians in some types of special libraries may be required to have additional training, such as a law degree for a librarian in an academic law library or appropriate subject degrees for subject specialties such as chemistry, engineering, etc. Many belong to the Special Libraries Association. There are also more specific associations such as the American Association of Law Libraries, Art Libraries Society of North America, the American Theological Library Association, the Medical Library Association, or the Visual Resources Association.", "edit_actions": [{"type": "R", "before": "and specialist librarian, from academic libraries to public libraries, as well as", "after": ", the specialty of the librarian, and", "start_char_pos": 376, "end_char_pos": 457, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "behind the scenes to maintain the collection and make it", "after": "to maintain collections and make them", "start_char_pos": 479, "end_char_pos": 535, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "catalogue", "after": "catalog", "start_char_pos": 1064, "end_char_pos": 1073, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "A", "before": null, "after": "there", "start_char_pos": 1362, "end_char_pos": 1362, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1665, "end_char_pos": 1665, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "delivering", "after": "Delivering", "start_char_pos": 3506, "end_char_pos": 3516, "major_intent": "others", "raw_intents": ["fluency", "others", "others"]}, {"type": "R", "before": "can be described as", "after": "are", "start_char_pos": 4053, "end_char_pos": 4072, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "to", "after": "for", "start_char_pos": 4126, "end_char_pos": 4128, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 4172, "end_char_pos": 4172, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "some include", "after": null, "start_char_pos": 4178, "end_char_pos": 4190, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "museum, or any other type of library owned and operated by URLanization", "after": "or museum", "start_char_pos": 4243, "end_char_pos": 4314, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 141, 303, 559, 718, 852, 986, 1118, 1187, 1261, 1336, 1608, 1834, 1996, 2285, 2549, 2614, 2757, 2946, 3013, 3066, 3153, 3471, 3866, 4316, 4413, 4593, 4638, 4885, 4935], "domain": "wiki"}
{"doc_id": "25664190", "revision_depth": 1, "before_revision": "While Aristotle considered Socrates a major philosopher, his writing didn't focus on him to the same degree as it did on other, pre-Socratic philosophers, and most of his followers didn't comment on Socrates at all. One of Aristotle pupil's unleashed an ad hominem attack on Socrates: Aristoxenus authored a book full of Socrates' scandals; it was not well-received by ancient critics. The Epicureans later weaponized Socratic irony in their polemic against Socrates. They also attacked him for superstition, given his story with the Delphi oracle. Epicurus, the founder of epicureanism, living in the 3rd and 4th  century BC, came across various currents claiming to be Socratic. They criticized Socrates for his character and various faults, and focused mostly on his irony, which was deemed inappropriate for a philosopher , anti-pedagogical . Also, his Socratic ignorance didn't resonate well with their criteria of truths.", "after_revision": "While Aristotle considered Socrates a major philosopher, his writing didn't focus on him to the same degree as it did on other, pre-Socratic philosophers, and most of his followers didn't comment on Socrates at all. One of Aristotle pupil's unleashed an ad hominem attack on Socrates: Aristoxenus authored a book full of Socrates' scandals; it was not well-received by ancient critics. The Epicureans later weaponized Socratic irony in their polemic against Socrates. They also attacked him for superstition, given his story with the Delphi oracle. Epicurus, the founder of epicureanism, living in the 4th and 3rd century BC, came across various currents claiming to be Socratic. The Epicureans criticized Socrates for his character and various faults, and focused mostly on his irony, which was deemed inappropriate for a philosopher and unsuitable for teaching . Also, his Socratic ignorance didn't resonate well with their criteria of truths.", "edit_actions": [{"type": "D", "before": "3rd and", "after": null, "start_char_pos": 602, "end_char_pos": 609, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "and 3rd", "start_char_pos": 614, "end_char_pos": 614, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "They", "after": "The Epicureans", "start_char_pos": 681, "end_char_pos": 685, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": ", anti-pedagogical", "after": "and unsuitable for teaching", "start_char_pos": 826, "end_char_pos": 844, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 215, 340, 385, 467, 548, 680, 846], "domain": "wiki"}
{"doc_id": "25664190", "revision_depth": 2, "before_revision": "Love Socrates and Alcibiades, by Christoffer Wilhelm Eckersberg There exist textual passages suggesting that Socrates had a love affair with Alcibiades and other young males  while other texts suggest that Socrates' friendship with young boys sought  to improve them and were not sexual. In URLias, Socrates claims he was a dual lover of Alcibiades and philosophy, and his flirtatiousness is evident in Protagoras, Meno (76ac) and Phaedrus (227cd). However, the exact nature of the relation is not clear since Socrates was known for his self-restraint, while Alcibiades admits in the Symposium that he had tried to seduce Socrates but failed. The Socratic theory of love is mostly deduced from Lysis, where Socrates talks about love at a wrestling school with Lysis and his friends. They start their dialogue by investigating parental love and how it manifests with respect to the freedom and boundaries they set for their child. Socrates concludes that if Lysis is utterly useless, nobody will love him, not even his parents. While most scholars consider this text to be humorous in intention, Gregory Vlastos suggests that it reveals the Socratic doctrine on love, which is an egoistic one, according to which we only love people that are useful to us in some way. Others scholars disagree with Vlastos' view; some argue that Socrates leaves room for non-egoistical love for a spouse, while others deny that Socrates is suggesting any egoistical motivation at all. A form of utility children have for parents, as Socrates claims in Symposium, is that they offer the false impression of immortality. Scholars note that for Socrates, love is rational.", "after_revision": "Love Socrates and Alcibiades, by Christoffer Wilhelm Eckersberg There exist textual passages suggesting that Socrates had a love affair with Alcibiades and other young males , while other texts suggest that Socrates' friendship with young boys sought only to improve them and were not sexual. In URLias, Socrates claims he was a dual lover of Alcibiades and philosophy, and his flirtatiousness is evident in Protagoras, Meno (76ac) and Phaedrus (227cd). However, the exact nature of his relationship with Alcibiades is not clear since Socrates was known for his self-restraint, while Alcibiades admits in the Symposium that he had tried to seduce Socrates but failed. The Socratic theory of love is mostly deduced from Lysis, where Socrates engages in a discussion about love at a wrestling school in the company of Lysis and his friends. They start their dialogue by investigating parental love and how it manifests with respect to the freedom and boundaries which parents set for their child. Socrates concludes that if Lysis is utterly useless, nobody will love him, not even his parents. While most scholars consider this text to be humorous in intention, Gregory Vlastos suggests that it reveals the Socratic doctrine on love, which is an egoistic one, according to which we only love people that are useful to us in some way. Others scholars disagree with Vlastos' view; some argue that Socrates leaves room for non-egoistical love for a spouse, while others deny that Socrates is suggesting any egoistical motivation at all. A form of utility children have for parents, as Socrates claims in Symposium, is that they offer the false impression of immortality. Scholars note that for Socrates, love is rational.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 174, "end_char_pos": 174, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "A", "before": null, "after": "only", "start_char_pos": 250, "end_char_pos": 250, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "the relation", "after": "his relationship with Alcibiades", "start_char_pos": 478, "end_char_pos": 490, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "talks", "after": "engages in a discussion", "start_char_pos": 716, "end_char_pos": 721, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "with", "after": "in the company of", "start_char_pos": 755, "end_char_pos": 759, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "they", "after": "which parents", "start_char_pos": 904, "end_char_pos": 908, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 287, 448, 642, 782, 929, 1026, 1266, 1311, 1466, 1600], "domain": "wiki"}
{"doc_id": "25893514", "revision_depth": 1, "before_revision": "When breast cancer is surgically excised, it can come back (local recurrence) in the remaining breast or on the chest wall in a small proportion of women. Adjuvant radiotherapy is necessary if breast cancer is treated by removing only the cancerous lump with a rim of surrounding normal tissue, as it reduces the chance of local recurrence significantly . When cancer does come back, it most commonly occurs in the tissues surrounding the original cancer (the tumour bed), even though there are multicentric cancers in remote areas of the breast. This suggests that it is most important to treat the tumour bed. The rationale for TARGIT is to deliver a high dose of radiation precisely to the tumour bed. Conventional radiation techniques such as external beam radiotherapy (EBRT) following surgical removal of the tumour have been time tested and proven to be effective. EBRT is usually given as a course of whole breast radiotherapy and an additional tumour bed boost  . However, it has a few drawbacks; for example, the tumour bed where the boost dose should be applied can be missed (\"geographical miss\") due to the difficulties in localization of the complex wound cavity  even when modern radiotherapy planning is used. Additionally, the usual delay (\"temporal miss\") between the surgical removal of the tumour and EBRT may allow a repopulation of the tumour cells  . These potentially harmful effects can be avoided by delivering the radiation more precisely to the targeted tissues leading to immediate sterilization of residual tumour cells. Furthermore, TARGIT inhibits the stimulating effects of wound fluid on cancer cells, suggesting for the first time, a beneficial effect of intraoperative radiotherapy (IORT) on tumour microenvironment. In these studies, women treated with  TARGIT-IORT  during lumpectomy for breast cancer combined with whole breast radiotherapy had a significantly low mortality from non-breast-cancer causes, such as cardiovascular and lung complications, as well as other cancers compared with those who received whole breast radiotherapy. Therefore, this reduction was thought to be due to the delivery of TARGIT-IORT to the fresh tumour bed during the trauma of surgery rather than avoidance of scattered radiation from whole breast radiotherapy .", "after_revision": "When breast cancer is surgically excised, it can come back (local recurrence) in the remaining breast or on the chest wall in a small proportion of women. Adjuvant radiotherapy is necessary if breast cancer is treated by removing only the cancerous lump with a rim of surrounding normal tissue, as it reduces the chance of local recurrence . When cancer does come back, it most commonly occurs in the tissues surrounding the original cancer (the tumour bed), even though there are multicentric cancers in remote areas of the breast. This suggests that it is most important to treat the tumour bed. The rationale for TARGIT is to deliver a high dose of radiation precisely to the tumour bed. Conventional radiation techniques such as external beam radiotherapy (EBRT) following surgical removal of the tumour have been time tested and proven to be effective. EBRT is usually given as a course of whole breast radiotherapy and an additional tumour bed boost , or partial breast irradiation of a smaller area . However, it has a few drawbacks; for example, the tumour bed where the boost dose should be applied can be missed due to the difficulties in localization of the complex wound cavity (\"geographical miss\"), even when modern radiotherapy planning is used. Additionally, the usual delay between the surgical removal of the tumour and EBRT may allow a repopulation of the tumour cells (\"temporal miss\") . These potentially harmful effects may be avoided by delivering the radiation more precisely to the targeted tissues leading to immediate sterilization of residual tumour cells. The use in TARGIT of a small treatment device which can be positioned in close physical proximity to the treatment site aims to avoid some of these practical issues. TARGIT irradiation has also been shown to affect the properties of wound fluid, which may be linked to cancer cell proliferation and possibly local recurrence. The TARGIT technique is the subject of clinical trials. A major phase III trial comparing TARGIT-IORT to EBRT during lumpectomy for breast cancer completed in 2012, with long-term follow-up ongoing .", "edit_actions": [{"type": "D", "before": "significantly", "after": null, "start_char_pos": 340, "end_char_pos": 353, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ", or partial breast irradiation of a smaller area", "start_char_pos": 970, "end_char_pos": 970, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "(\"geographical miss\")", "after": null, "start_char_pos": 1087, "end_char_pos": 1108, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "(\"geographical miss\"),", "start_char_pos": 1177, "end_char_pos": 1177, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "(\"temporal miss\")", "after": null, "start_char_pos": 1256, "end_char_pos": 1273, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "(\"temporal miss\")", "start_char_pos": 1371, "end_char_pos": 1371, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "can", "after": "may", "start_char_pos": 1408, "end_char_pos": 1411, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "Furthermore, TARGIT inhibits the stimulating effects of wound fluid on cancer cells, suggesting for the first time, a beneficial effect of intraoperative radiotherapy (IORT) on tumour microenvironment. In these studies, women treated with", "after": "The use in TARGIT of a small treatment device which can be positioned in close physical proximity to the treatment site aims to avoid some of these practical issues. TARGIT irradiation has also been shown to affect the properties of wound fluid, which may be linked to cancer cell proliferation and possibly local recurrence.", "start_char_pos": 1551, "end_char_pos": 1789, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "The TARGIT technique is the subject of clinical trials. A major phase III trial comparing", "start_char_pos": 1790, "end_char_pos": 1790, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "fluency", "meaning-changed"]}, {"type": "A", "before": null, "after": "to EBRT", "start_char_pos": 1803, "end_char_pos": 1803, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "combined with whole breast radiotherapy had a significantly low mortality from non-breast-cancer causes, such as cardiovascular and lung complications, as well as other cancers compared with those who received whole breast radiotherapy. Therefore, this reduction was thought to be due to the delivery of TARGIT-IORT to the fresh tumour bed during the trauma of surgery rather than avoidance of scattered radiation from whole breast radiotherapy", "after": "completed in 2012, with long-term follow-up ongoing", "start_char_pos": 1840, "end_char_pos": 2284, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 154, 355, 546, 611, 704, 871, 972, 1005, 1225, 1373, 1550, 1752, 2076], "domain": "wiki"}
{"doc_id": "25893514", "revision_depth": 3, "before_revision": "Targeted intra-operative radiotherapy, also known as targeted IORT or TARGIT (TARGeted Intraoperative radioTherapy) is a technique of giving radiotherapy to the tissues surrounding a cancer after its surgical removal, a form of intraoperative radiation therapy. The technique was designed in 1998 at the University College London. In patients having lumpectomy for breast cancer, the TARGIT-A(lone) randomized controlled trial (recruitment from 20002012) tested whether TARGIT within a risk-adapted approach is non-inferior to conventional course of external beam postoperative radiotherapy given over several weeks. TARGIT is a method where the radiation is applied during an operation and targeted to the peri-tumoural tissues. TARGIT technique was designed at University College London by Jayant S Vaidya (who coined the TARGIT acronym) and Michael Baum along with Jeffrey S Tobias in 1998. The term was first used when the technique was described, and the protocol for a randomised trial was published by The Lancet. Technique TARGeted Intraoperative radioTherapy (TARGIT) IORT for breast cancer The machine used for TARGIT is Intrabeam (Carl Zeiss , Germany) . It is a miniature and mobile X-ray source which emits low energy X-ray radiation (max. 50 kV) in isotropic distribution. Due to the higher ionization density caused by soft X-ray radiation in the tissue, the relative biological effectiveness (RBE) of low-energy X-rays on tumour cells is higher when compared to high-energy X-rays or gamma rays which are delivered by linear accelerators. The radiation which is produced by mobile radiation systems has a limited range . For this reason, conventional walls are regarded sufficient to stop the radiation scatter produced in the operating room and no extra measures for radiation protection are necessary. This makes IORT for breast cancer by the TARGIT technique available in most operating rooms. The surgical technique is relatively simple but needs to be meticulously followed .", "after_revision": "Targeted intra-operative radiotherapy, also known as targeted IORT is a technique of giving radiotherapy to the tissues surrounding a cancer after its surgical removal, a form of intraoperative radiation therapy. The technique was designed in 1998 at the University College London. In patients having lumpectomy for breast cancer, the TARGIT-A(lone) randomized controlled trial (recruitment from 20002012) tested whether TARGIT within a risk-adapted approach is non-inferior to conventional course of external beam postoperative radiotherapy given over several weeks. One IORT technique is known as TARGIT (TARGeted Intraoperative radioTherapy). TARGIT is a method where the radiation is applied during an operation and targeted to the peri-tumoural tissues. TARGIT technique was designed at University College London by Jayant S Vaidya (who coined the TARGIT acronym) and Michael Baum along with Jeffrey S Tobias in 1998. The term was first used when the technique was described, and the protocol for a randomised trial was published by The Lancet. Technique The Intrabeam system, manufactured by Carl Zeiss AG, is used for TARGIT . It is a miniature and mobile X-ray source which emits low energy X-ray radiation (max. 50 kV) in isotropic distribution. Due to the higher ionisation density caused by soft X-ray radiation in the tissue, the relative biological effectiveness (RBE) of low-energy X-rays on tumour cells is higher when compared to high-energy X-rays or gamma rays which are delivered by linear accelerators. Lower energy radiation such as this has a limited range , and therefore conventional walls may be sufficient to stop the radiation scatter produced in the operating room without extra measures for radiation protection .", "edit_actions": [{"type": "D", "before": "or TARGIT (TARGeted Intraoperative radioTherapy)", "after": null, "start_char_pos": 67, "end_char_pos": 115, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "TARGIT", "after": "One IORT technique is known as TARGIT (TARGeted Intraoperative radioTherapy). TARGIT", "start_char_pos": 617, "end_char_pos": 623, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "TARGeted Intraoperative radioTherapy (TARGIT) IORT for breast cancer The machine used for TARGIT is Intrabeam (Carl Zeiss , Germany)", "after": "The Intrabeam system, manufactured by Carl Zeiss AG, is used for TARGIT", "start_char_pos": 1031, "end_char_pos": 1163, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "ionization", "after": "ionisation", "start_char_pos": 1305, "end_char_pos": 1315, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "The radiation which is produced by mobile radiation systems", "after": "Lower energy radiation such as this", "start_char_pos": 1555, "end_char_pos": 1614, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ". For this reason, conventional walls are regarded", "after": ", and therefore conventional walls may be", "start_char_pos": 1635, "end_char_pos": 1685, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "and no", "after": "without", "start_char_pos": 1758, "end_char_pos": 1764, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "are necessary. This makes IORT for breast cancer by the TARGIT technique available in most operating rooms. The surgical technique is relatively simple but needs to be meticulously followed", "after": null, "start_char_pos": 1805, "end_char_pos": 1994, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 261, 330, 616, 729, 893, 1020, 1165, 1286, 1554, 1636, 1819, 1912], "domain": "wiki"}
{"doc_id": "28245224", "revision_depth": 1, "before_revision": "The Ngqika people are a Xhosa monarchy who lived west of the Great Kei River2017. Ngqika ka Mlawu, 3rd Paramount Chief of the amaRarabe. Geni. Accessed 13 March. in what is today the Eastern Cape of South Africa. They were first ruled by Rarabe kaPhalo who died with his son Mlawu, who was destined for chieftaincy. The clan would be named after Ngqika ka Mlawu, the son of the then late Mlawu.2017. Maqoma's final resistance. News24. Accessed 13 March. It would be years before the child would rule his people who fought in the Xhosa Wars following the invasion of the British and Dutch settlers . In colonial times, the Ngqika lands were known as British Kaffraria. Later the Apartheid government gave them a form of independence as the former \"Ciskei\" homeland. The occupation by white settlers(Dutch and British) resulted in the century of Xhosa Wars (commonly known as the Frontier Wars) from 1779 to 1879 involving Xhosa tribes (including the Ngqika), Khoikhoi and San people. Relations between the European invaders were fluid throughout this period with the varying indigenous people and Europeans working for or against one another to secure land, resources and cattle. These groups changed allies depending on who would help advance their objectives over the years. In 1793, in collaboration with other Xhosa clans, the Ngqikas fought against white settlers during the Second Frontier War. The native groups lost most of their land and were absorbed into the British empire during this century. The Xhosa eventually lost access to the fertile land between the Great Kei River and Great Fish River to British colonists . 2017. Battle of Amalinde. Buffalo City. Accessed 13 March.", "after_revision": "The Ngqika people are a Xhosa monarchy who lived west of the Great Kei River2017. Ngqika ka Mlawu, 3rd Paramount Chief of the amaRarabe. Geni. Accessed 13 March. in what is today the Eastern Cape of South Africa. They were first ruled by Rarabe kaPhalo who died with his son Mlawu, who was destined for chieftaincy. The clan would be named after Ngqika ka Mlawu, the son of the then late Mlawu.2017. Maqoma's final resistance. News24. Accessed 13 March. It would be years before the child would rule his people who fought in the Xhosa Wars , which were sparked by the encroachment of European settlers on Xhosa lands . In colonial times, the Ngqika lands were known as British Kaffraria. Later the , Apartheid government of South Africa gave them a form of independence as the former \"Ciskei\" homeland. Encroachment by European settlers, beginning in the Dutch colonial period, resulted in the century of Xhosa Wars (commonly known as the Frontier Wars) from 1779 to 1879 involving Xhosa tribes (including the Ngqika), Khoikhoi and San people. Relations between the European invaders were fluid throughout this period with the varying indigenous people and Europeans working for or against one another to secure land, resources and cattle. These groups changed allies depending on who would help advance their objectives over the years. In 1793, in collaboration with other Xhosa clans, the Ngqikas fought against white settlers during the Second Frontier War. The native groups lost most of their land and were absorbed into the Cape Colony during this century. The Xhosa eventually lost access to the fertile land between the Great Kei River and Great Fish River to encroaching European settlers . 2017. Battle of Amalinde. Buffalo City. Accessed 13 March.", "edit_actions": [{"type": "R", "before": "following the invasion of the British and Dutch settlers", "after": ", which were sparked by the encroachment of European settlers on Xhosa lands", "start_char_pos": 540, "end_char_pos": 596, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Apartheid government", "after": ", Apartheid government of South Africa", "start_char_pos": 678, "end_char_pos": 698, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "The occupation by white settlers(Dutch and British)", "after": "Encroachment by European settlers, beginning in the Dutch colonial period,", "start_char_pos": 765, "end_char_pos": 816, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "British empire", "after": "Cape Colony", "start_char_pos": 1469, "end_char_pos": 1483, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "British colonists", "after": "encroaching European settlers", "start_char_pos": 1610, "end_char_pos": 1627, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "others"]}], "sents_char_pos": [0, 81, 136, 212, 315, 399, 426, 453, 598, 667, 764, 982, 1178, 1275, 1399, 1504, 1635, 1655, 1669], "domain": "wiki"}
{"doc_id": "28583590", "revision_depth": 1, "before_revision": "Northern Rhodesia's first commemorative stamps were a set of five on 30 May 1953, marking the birth centennial of Cecil Rhodes, along with another stamp issued on the same day for the Rhodes Centenary Exhibition. As Elizabeth II of the United Kingdom was already on the throne, the stamps included her profile, even though the official coronation stamp did not come out until 2 June. This was followed in September by a definitive series using the same design as before, with the monarch's profile updated once again. Only a few months later, these stamps were withdrawn, and stamps of the Federation of Rhodesia and Nyasaland issued in their place , i. e. NR QEII stamps are relatively rare. Northern Rhodesia used the Federation's stamps from 1954 until it was dissolved in 1963.   On 10 December 1963, a new definitive series depicted the Queen and the territory's arms were issued , but  withdrawn on 23 October 1964 when Zambia became URL Retrieved 12 August 2018.", "after_revision": "Northern Rhodesia's first commemorative stamps were a set of five on 30 May 1953, marking the birth centennial of Cecil Rhodes, along with another stamp issued on the same day for the Rhodes Centenary Exhibition. As Elizabeth II of the United Kingdom was already on the throne, the stamps included her profile, even though the official coronation stamp did not come out until 2 June. This was followed in September by a definitive series using the same design as before, with the monarch's profile updated once again. Only a few months later, these stamps were withdrawn, and stamps of the Federation of Rhodesia and Nyasaland issued in their place . Northern Rhodesia used the Federation's stamps from 1954 until it was dissolved in 1963. For an interim period, the stamps of Northern Rhodesia were valid for postage in all three territories: Northern Rhodesia, Southern Rhodesia Nyasaland. On 10 December 1963, a new definitive series depicted the Queen and the territory's arms were issued . These were hastily printed by Harrison Sons Ltd and can be found with multiple errors such as missing values and double printing or missing colours. This definitive series was withdrawn on 23 October 1964 when Northern Rhodesia achieved independence and became the Republic of URL Retrieved 12 August 2018.", "edit_actions": [{"type": "R", "before": ", i. e. NR QEII stamps are relatively rare.", "after": ".", "start_char_pos": 649, "end_char_pos": 692, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "For an interim period, the stamps of Northern Rhodesia were valid for postage in all three territories: Northern Rhodesia, Southern Rhodesia", "start_char_pos": 782, "end_char_pos": 782, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Nyasaland.", "start_char_pos": 783, "end_char_pos": 783, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}, {"type": "R", "before": ", but", "after": ". These were hastily printed by Harrison", "start_char_pos": 885, "end_char_pos": 890, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "A", "before": null, "after": "Sons Ltd and can be found with multiple errors such as missing values and double printing or missing colours. This definitive series was", "start_char_pos": 891, "end_char_pos": 891, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Zambia became", "after": "Northern Rhodesia achieved independence and became the Republic of", "start_char_pos": 926, "end_char_pos": 939, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 212, 383, 517, 692, 781], "domain": "wiki"}
{"doc_id": "28623021", "revision_depth": 1, "before_revision": "Examples of drought deciduous species Drought deciduous plants are very commonly found near the borders of desert or long the coastal areas in the Northern Americas. Encelia fairnosa (commonly known as brittlebush) is a very common broadleaf drought deciduous plant species  near the desert around the Northern part of Mexico and along the West Coast of the United States. Ambrosia Dumosa (commonly known as Burroweed) is a common microphylllous drought deciduous species which is found near the mentioned areas and live within similar ecosystems as the brittlebush. These two examples although both possess the ability of undergoing abscission due to droughts or dry seasons, they are of different plant families.", "after_revision": "Examples of drought deciduous species Drought deciduous plants are very commonly found near the borders of deserts or along the coastal areas in the Northern Americas. Encelia fairnosa (commonly known as brittlebush) is a very common broadleaf drought deciduous plant species found near the desert around the northern part of Mexico and along the West Coast of the United States. Ambrosia Dumosa (commonly known as Burroweed) is a common microphyllous drought deciduous species which is found near those same areas as it lives within similar ecosystems as the brittlebush. Although both these examples are able to undergo abscission in response to droughts or dry seasons, they are from different plant families.", "edit_actions": [{"type": "R", "before": "desert or long", "after": "deserts or along", "start_char_pos": 107, "end_char_pos": 121, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "A", "before": null, "after": "found", "start_char_pos": 274, "end_char_pos": 274, "major_intent": "clarity", "raw_intents": ["clarity", "others", "coherence"]}, {"type": "R", "before": "Northern", "after": "northern", "start_char_pos": 302, "end_char_pos": 310, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "microphylllous", "after": "microphyllous", "start_char_pos": 431, "end_char_pos": 445, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the mentioned areas and live", "after": "those same areas as it lives", "start_char_pos": 492, "end_char_pos": 520, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "These two examples although both possess the ability of undergoing abscission due", "after": "Although both these examples are able to undergo abscission in response", "start_char_pos": 567, "end_char_pos": 648, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "of", "after": "from", "start_char_pos": 686, "end_char_pos": 688, "major_intent": "clarity", "raw_intents": ["clarity", "others", "fluency"]}], "sents_char_pos": [0, 165, 372, 566], "domain": "wiki"}
{"doc_id": "2882854", "revision_depth": 1, "before_revision": "Noon sights obtain the observer's Latitude. It is impossible to determine longitude with an accuracy better than 10nmi by means of a noon sight. A noon sight is called a Meridian Altitude.Basic Principles of Marine Navigation by D A Moore Published by Kandy p81 While it is very easy to determine the observer's latitude at noon without knowing the exact time, longitude cannot accurately be measured at noon. At noon the sun's change of altitude is very slow, so determining the exact time that the sun is at its zenith by direct observation is impossible, and therefore it is impossible to obtain an accurate longitude at the moment of Zenith . However, it is possible to determine the time of zenith for longitude with a useful accuracy by performing a mean time of observation when the sun is on its ascent and descent prior to and following its moment of Zenith . By taking a sextant reading within 15 to 30 minutes prior to local noon ( zenith ) and noting the time, then leaving the sextant set to the same angle and subsequently observing the moment in time at which the sun passes through the sight tube on its descent from Zenith between a half-hour and hour later, the two times can be averaged to obtain a longitude sufficiently accurate for navigation (within 2nmi).American Practical Navigator (bowditch) by Nathanial Bowditch p253", "after_revision": "Noon sights obtain the observer's Latitude. It is impossible to determine longitude with an accuracy better than 10nmi by means of a noon sight. A noon sight is called a Meridian Altitude.Basic Principles of Marine Navigation by D A Moore Published by Kandy p81 While it is very easy to determine the observer's latitude at noon without knowing the exact time, longitude cannot accurately be measured at noon. At noon the sun's change of altitude is very slow, so determining the exact time that the sun is at its highest by direct observation is impossible, and therefore it is impossible to obtain an accurate longitude at the moment of culmination . However, it is possible to determine the time of culmination for longitude with a useful accuracy by performing a mean time of observation when the sun is on its ascent and descent prior to and following its moment of culmination . By taking a sextant reading within 15 to 30 minutes prior to local noon ( culmination ) and noting the time, then leaving the sextant set to the same angle and subsequently observing the moment in time at which the sun passes through the sight tube on its descent from its highest altitude between a half-hour and hour later, the two times can be averaged to obtain a longitude sufficiently accurate for navigation (within 2nmi).American Practical Navigator (bowditch) by Nathanial Bowditch p253", "edit_actions": [{"type": "R", "before": "zenith", "after": "highest", "start_char_pos": 514, "end_char_pos": 520, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Zenith", "after": "culmination", "start_char_pos": 638, "end_char_pos": 644, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "zenith", "after": "culmination", "start_char_pos": 696, "end_char_pos": 702, "major_intent": "style", "raw_intents": ["style", "others", "clarity"]}, {"type": "R", "before": "Zenith", "after": "culmination", "start_char_pos": 860, "end_char_pos": 866, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "zenith", "after": "culmination", "start_char_pos": 943, "end_char_pos": 949, "major_intent": "clarity", "raw_intents": ["others", "clarity", "clarity"]}, {"type": "R", "before": "Zenith", "after": "its highest altitude", "start_char_pos": 1133, "end_char_pos": 1139, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 43, 144, 188, 409, 646, 868, 1279], "domain": "wiki"}
{"doc_id": "2884986", "revision_depth": 2, "before_revision": "Once the figure for the Meridian Zenith Distance is obtained the sum or the difference is obtained between it and the declination of the object to obtain the latitude . This is the latitude where the position line crosses the  DR longitude.", "after_revision": "Once the value of the Meridian Zenith Distance is obtained the algebraic sum of it with the declination of the object gives the latitude of a point where the position line crosses the meridian of DR longitude.", "edit_actions": [{"type": "R", "before": "figure for", "after": "value of", "start_char_pos": 9, "end_char_pos": 19, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "sum or the difference is obtained between it and", "after": "algebraic sum of it with", "start_char_pos": 65, "end_char_pos": 113, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "to obtain the latitude . This is the latitude", "after": "gives the latitude of a point", "start_char_pos": 144, "end_char_pos": 189, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "meridian of", "start_char_pos": 226, "end_char_pos": 226, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 168], "domain": "wiki"}
{"doc_id": "2884986", "revision_depth": 3, "before_revision": "Ex- Meridian is an astronomical navigation method of calculating an observer\u2019s position on earth . The method gives the observer a position line on which the observer is situated. It is usually used when the sun is obscured at noon  and as a result  a meridian altitude is not possible. The navigator measures the altitude of the sun as close to noon as possible and then calculates where the position line lies.", "after_revision": "Ex-meridian is a celestial navigation method of calculating an observer\u2019s position on Earth . The method gives the observer a position line on which the observer is situated. It is usually used when the Sun is obscured at noon , and as a result , a meridian altitude is not possible. The navigator measures the altitude of the Sun as close to noon as possible and then calculates where the position line lies.", "edit_actions": [{"type": "R", "before": "Ex- Meridian is an astronomical", "after": "Ex-meridian is a celestial", "start_char_pos": 0, "end_char_pos": 31, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "earth", "after": "Earth", "start_char_pos": 91, "end_char_pos": 96, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "sun", "after": "Sun", "start_char_pos": 208, "end_char_pos": 211, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 232, "end_char_pos": 232, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 249, "end_char_pos": 249, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "sun", "after": "Sun", "start_char_pos": 330, "end_char_pos": 333, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 98, 179, 286], "domain": "wiki"}
{"doc_id": "292089", "revision_depth": 1, "before_revision": "Yuma Proving Ground The 16-inch HARP gun at Yuma Proving Ground was constructed in 1966 in order to establish a functional 16-inch gun on American soil and holds the record for achieving the highest projectile launched. It was almost identical to the 16-inch gun on Barbados, being 119ft long, but was limited by a 35-mile range restriction. However, unlike the Barbados gun, its projectiles could be recovered since they were not lost in the ocean upon their journey back down. The Yuma 16-inch gun was primarily used for flight tests, such as those testing altitude control and telemetry components. In 1966, the 16-inch Yuma gun underwent three firing series using wooden slugs, Martlet 2C's, and a Low Altitude High Velocity cone. +Performance Test of Yuma 16-inch HARP gun in 1966DateRound numberAll projectiles used for this test were Martlet 2C's, except for the ones indicated with (W), i.e. wooded slugs, and (L), i.e. Low Altitude High Velocity cone.Mass (lb)Muzzle velocity (ft/s)Apogee (kilofeet)June 7001 (W)7003360No trackJune 8002 (W)8003190No trackJune 13003 (W)6604810No trackJune 130047605930415June 130057805810398June 140067806060400June 140078006270DamagedJune 150087605630375June 150097805850410Oct. 2501010955250310Oct. 2601112255950410Oct. 260129206800540Oct. 27013 (L)9007100No trackOct. 2701412755900415Oct. 270159206780535Oct. 270169507040DamagedNov. 1601712905900396Nov. 1601812925900395Nov. 1601912965850415Nov.1702012965950415Nov. 170211290NADamagedNov. 1802212635900400Nov. 1802312635850410Nov. 180249226650510Nov. 180258806400490Nov. 190269106650530Nov. 1902712705850400Nov. 190289607000590Nov. 190291270NADamagedNov. 190309606350480Nov. 1903112705650367Nov. 1903212705650370Nov. 190338806750550 There were several models of test projectiles  fired or designed during Project HARP: These projectiles were fired on the island of Barbados and some were fired by the US Army's Ballistic Research Lab. The slender design of the tube which contained the rocket's payload  was very narrow and long, limiting what objects could be inserted into the tube. This limitation on size was extremely inconvenient when considering the future proposed payloads of Martlet rockets which included satellites and space probes. The cannon-like design also eliminated the capacity for crewed space travel as well as the launching of satellites carrying extremely sensitive scientific instruments and payloads due to the extreme acceleration placed on the projectile during firing. Martlet 2A, 2B, 2C family The Martlet 2A, 2B, and 2C represented the earliest of the Martlet 2 16-inch (406mm) test projectiles. Marlet 2A was designed simultaneously with the Martlet 1 with a range of interest being 70 to 200 kilometers. Most carried multi-type research payloads studying the upper atmosphere and near-space conditions. Due to their low cost per missile launch, they were used to test out single payloads. Despite similarities in missile air frame , the Martlet 2A, 2B, and 2C featured differences in their structural materials and mechanical details. For the Martlet 2A, the liquid payload was loaded into an aluminum, tapered liner inside the missile body. But by the development of the Martlet 2C series, the aluminum insert was abandoned altogether to allow the liquid payload to be housed in contact with the steel body, increasing the quantity of the liquid payload that could be carried.", "after_revision": "Yuma Proving Ground The 16-inch HARP gun at Yuma Proving Ground was constructed in 1966 in order to establish a functional 16-inch gun on American soil and holds the record for achieving the highest projectile launched. It was almost identical to the 16-inch gun on Barbados, being 119ft long, but was limited by a 35-mile range restriction. However, unlike the Barbados gun, its projectiles could be recovered since they were not lost in the ocean upon their journey back down. The Yuma 16-inch gun was primarily used for flight tests, such as those testing altitude control and telemetry components. In 1966, the 16-inch Yuma gun underwent three firing series using wooden slugs, Martlet 2C's, and a Low Altitude High-Velocity cone. +Performance Test of Yuma 16-inch HARP gun in 1966DateRound numberAll projectiles used for this test were Martlet 2C's, except for the ones indicated with (W), i.e. wooded slugs, and (L), i.e. Low Altitude High Velocity cone.Mass (lb)Muzzle velocity (ft/s)Apogee (kilofeet)June 7001 (W)7003360No trackJune 8002 (W)8003190No trackJune 13003 (W)6604810No trackJune 130047605930415June 130057805810398June 140067806060400June 140078006270DamagedJune 150087605630375June 150097805850410Oct. 2501010955250310Oct. 2601112255950410Oct. 260129206800540Oct. 27013 (L)9007100No trackOct. 2701412755900415Oct. 270159206780535Oct. 270169507040DamagedNov. 1601712905900396Nov. 1601812925900395Nov. 1601912965850415Nov.1702012965950415Nov. 170211290NADamagedNov. 1802212635900400Nov. 1802312635850410Nov. 180249226650510Nov. 180258806400490Nov. 190269106650530Nov. 1902712705850400Nov. 190289607000590Nov. 190291270NADamagedNov. 190309606350480Nov. 1903112705650367Nov. 1903212705650370Nov. 190338806750550 Several models of test projectiles were fired or designed during Project HARP: These projectiles were fired on the island of Barbados and some were fired by the US Army's Ballistic Research Lab. The tube's slender design, which contained the rocket's payload , was very narrow and long, limiting what objects could be inserted into the tube. This limitation on size was extremely inconvenient when considering the future proposed payloads of Martlet rockets , including satellites and space probes. The cannon-like design also eliminated the capacity for crewed space travel as well as the launching of satellites carrying extremely sensitive scientific instruments and payloads due to the extreme acceleration placed on the projectile during firing. Martlet 2A, 2B, 2C family The Martlet 2A, 2B, and 2C represented the earliest of the Martlet 2 16-inch (406mm) test projectiles. Martlet 2A was designed simultaneously with the Martlet 1 with a range of interest being 70 to 200 kilometers. Most carried multi-type research payloads studying the upper atmosphere and near-space conditions. Due to their low cost per missile launch, they were used to test out single payloads. Despite similarities in missile airframe , the Martlet 2A, 2B, and 2C featured differences in their structural materials and mechanical details. For the Martlet 2A, the liquid payload was loaded into an aluminum, tapered liner inside the missile body. But by the development of the Martlet 2C series, the aluminum insert was abandoned altogether to allow the liquid payload to be housed in contact with the steel body, increasing the quantity of the liquid payload that could be carried.", "edit_actions": [{"type": "R", "before": "High Velocity", "after": "High-Velocity", "start_char_pos": 715, "end_char_pos": 728, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "There were several", "after": "Several", "start_char_pos": 1728, "end_char_pos": 1746, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "were", "start_char_pos": 1774, "end_char_pos": 1774, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "slender design of the tube", "after": "tube's slender design,", "start_char_pos": 1934, "end_char_pos": 1960, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1998, "end_char_pos": 1998, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "which included", "after": ", including", "start_char_pos": 2196, "end_char_pos": 2210, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "Marlet", "after": "Martlet", "start_char_pos": 2621, "end_char_pos": 2627, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "air frame", "after": "airframe", "start_char_pos": 2948, "end_char_pos": 2957, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 219, 341, 478, 601, 734, 960, 1283, 1312, 1353, 1813, 1929, 2079, 2239, 2491, 2620, 2730, 2829, 2915, 3061, 3168], "domain": "wiki"}
{"doc_id": "292089", "revision_depth": 2, "before_revision": "Martlet 2G and 2G-1 The Martlet 2G was an advanced test projectile that had nearly all of its total 350lb (160kg) weight in the projectile. It was tested successfully with the Highwater gun and the Barbados gun but never proceeded beyond the engineering flight testing stage. The Martlet 2G-1 was a proposed space launch vehicle variant of Martlet 2G, which had a solid rocket motor in the projectile. The follow-on 2G-2 proposal was to have had a second rocket motor in order to be able to place the second stage in orbit, though with little or no payload. After development, it underwent horizontal development firings from the Highwater gun in 1966 but failed to be properly tested in time. Martlet 3 The Martlet 3 series consisted of advanced rocket-propelled projectiles. They were built and tested for the HARP project but were ultimately not successful due to restrictions in funding and a severe lack of technical information regarding the behavior of large rocket grains  under high acceleration loading. When testing these projectiles, the danger of in-bore detonation was considered a very serious potential problem. Martlet 3A The Martlet 3A was an diameter, gun-fired rocket projectile that theoretically was able to reach 500km altitude. As HARP's first attempt at a low-cost sabotted rocket system, the projectile was built with fiberglass or aluminum bodies. A standard 6-inch rocket was bonded to an aluminum case , and the rocket nozzle was supported by a pusher plate, which would impart the acceleration thrust to the missile through the aluminum wall casing. Fiberglass limited acceleration to 3600g (corresponding to a velocity of at rocket ignition). The Martlet 3A's original objective was to carry a 40-lbs payload to an altitude of 500km, which theoretically was feasible if the system could be launched at full gun pressures. The rocket motors' solid propellant deformed during firing and the design was never successful, despite several test firings.", "after_revision": "Martlet 2G and 2G-1 The Martlet 2G was an advanced test projectile with nearly all of its total 350lb (160kg) weight in the projectile. It was tested successfully with the Highwater gun and the Barbados gun but never proceeded beyond the engineering flight testing stage. The Martlet 2G-1 was a proposed space launch vehicle variant of Martlet 2G, which had a solid rocket motor in the projectile. The follow-on 2G-2 proposal was to have had a second rocket motor to place the second stage in orbit, though with little or no payload. After development, it underwent horizontal development firings from the Highwater gun in 1966 but failed to be properly tested in time. Martlet 3 The Martlet 3 series consisted of advanced rocket-propelled projectiles. They were built and tested for the HARP project but were ultimately not successful due to restrictions in funding and a severe lack of technical information regarding large rocket grains ' behaviour under high acceleration loading. When testing these projectiles, the danger of in-bore detonation was considered a severe potential problem. Martlet 3A The Martlet 3A was an diameter, gun-fired rocket projectile that theoretically could reach 500km altitude. As HARP's first attempt at a low-cost sabotted rocket system, the projectile was built with fiberglass or aluminum bodies. A standard 6-inch rocket was bonded to an aluminum case . The rocket nozzle was supported by a pusher plate, which would impart the missile's acceleration through the aluminum wall casing. Fiberglass limited acceleration to 3600g (corresponding to a velocity of at rocket ignition). The Martlet 3A's original objective was to carry a 40-lbs payload to an altitude of 500km, which theoretically was feasible if the system could be launched at full gun pressures. The rocket motors' solid propellant deformed during firing and the design was never successful, despite several test firings.", "edit_actions": [{"type": "R", "before": "that had", "after": "with", "start_char_pos": 67, "end_char_pos": 75, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "in order to be able to", "after": "to", "start_char_pos": 468, "end_char_pos": 490, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "the behavior of", "after": null, "start_char_pos": 944, "end_char_pos": 959, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "' behaviour", "start_char_pos": 980, "end_char_pos": 980, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "very serious", "after": "severe", "start_char_pos": 1096, "end_char_pos": 1108, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "was able to", "after": "could", "start_char_pos": 1218, "end_char_pos": 1229, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": ", and the", "after": ". The", "start_char_pos": 1431, "end_char_pos": 1440, "major_intent": "coherence", "raw_intents": ["coherence", "fluency", "coherence"]}, {"type": "R", "before": "acceleration thrust to the missile", "after": "missile's acceleration", "start_char_pos": 1511, "end_char_pos": 1545, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 139, 275, 401, 557, 693, 776, 1013, 1127, 1251, 1374, 1579, 1673, 1852], "domain": "wiki"}
{"doc_id": "292089", "revision_depth": 3, "before_revision": "Marlet 3B The Martlet 3B was similar to the Martlet 3A but using steel casings and attempting to solve some of the 3A model's other problems. The casings survived , but the propellant failed at . This was solved for later rockets by filling the propellant cavity with liquid, but only after development of the 3B model had ended. Martlet 3E The Martlet 3E was a suborbital solid rocket designed to be fired from a smaller, 7-inch (180mm) cannon used in the HARP project. Its basic concept revolved around packaging the rocket grain in a case with elastic properties to transmit the lateral strain to the gun tube. The 3E model utilized a new technique of rocket grain construction  that consisted of laminating sheet double-base propellant grain under hydraulic pressure. Martlet 4 By July 1964, the Marlet 4 program pursued the development of an orbit capable multi-stage rocket system to be launched from the 16\" Barbados gun. Two versions of full-scale orbital launch vehicle projectiles were proposed in the Martlet 4 series. The first was to have used three solid rocket motor stages and was planned to orbit approximately 50 pounds of payload. The second used liquid rocket motors and was planned to have orbited 200 pounds of payload. Both were about long and in diameter, weighing about at launch. However, no Martlet 4 vehicles were built; the project was halted before the design was completed. Martlet 4 Control Systems A guidance and control system was developed for the orbital mission by Aviation Electric Limited of Montreal under the direction of McGill-BRL-Harry Diamond Laboratory group. Infrared horizon sensors and sun sensors were included to calculate vehicle attitude. Information for on-board sensors was to be processed by the logic module  which provided commands to a cold gas thruster system which in turn adjusted the orientation of the vehicle . The components of the guidance and control assembly were integrated into a 6.25 inch diameter test projectile. Sun sensors, horizon sensors, telemetry packs, receiving/transmitting antenna, hydraulic systems, logic modules, and gas thruster attitude control systems were all test fired to approximately 10,000 g's.", "after_revision": "Marlet 3B The Martlet 3B was similar to the Martlet 3A but using steel casings and attempting to solve some of the 3A model's other problems. The casings survived , but the propellant failed at . This was solved for later rockets by filling the propellant cavity with liquid, but only after developing the 3B model had ended. Martlet 3E The Martlet 3E was a solid suborbital rocket designed to be fired from a smaller, 7-inch (180mm) cannon used in the HARP project. Its basic concept revolved around packaging the rocket grain in a case with elastic properties to transmit the lateral strain to the gun tube. The 3E model utilized a new rocket grain construction technique that consisted of laminating sheet double-base propellant grain under hydraulic pressure. Martlet 4 By July 1964, the Marlet 4 program pursued developing an orbit capable multi-stage rocket system to be launched from the 16\" Barbados gun. Two versions of full-scale orbital launch vehicle projectiles were proposed in the Martlet 4 series. The first was to have used three solid rocket motor stages and was planned to orbit approximately 50 pounds of payload. The second used liquid rocket motors and was planned to have orbited 200 pounds of payload. Both were about long and in diameter, weighing about at launch. However, no Martlet 4 vehicles were built; the project was halted before the design was completed. Martlet 4 Control Systems A guidance and control system were developed for the orbital mission by Aviation Electric Limited of Montreal under the direction of McGill-BRL-Harry Diamond Laboratory group. Infrared horizon sensors and sun sensors were included in calculating vehicle attitude. Information for on-board sensors was to be processed by the logic module , which provided commands to a cold gas thruster system which in turn adjusted the vehicle's orientation . The components of the guidance and control assembly were integrated into a 6.25 -inch diameter test projectile. Sun sensors, horizon sensors, telemetry packs, receiving/transmitting antenna, hydraulic systems, logic modules, and gas thruster attitude control systems were all test-fired to approximately 10,000 g's.", "edit_actions": [{"type": "R", "before": "development of", "after": "developing", "start_char_pos": 291, "end_char_pos": 305, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "suborbital solid", "after": "solid suborbital", "start_char_pos": 362, "end_char_pos": 378, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "D", "before": "technique of", "after": null, "start_char_pos": 642, "end_char_pos": 654, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "technique", "start_char_pos": 681, "end_char_pos": 681, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "the development of", "after": "developing", "start_char_pos": 825, "end_char_pos": 843, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "was", "after": "were", "start_char_pos": 1461, "end_char_pos": 1464, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "to calculate", "after": "in calculating", "start_char_pos": 1661, "end_char_pos": 1673, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1765, "end_char_pos": 1765, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "orientation of the vehicle", "after": "vehicle's orientation", "start_char_pos": 1847, "end_char_pos": 1873, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "inch", "after": "-inch", "start_char_pos": 1956, "end_char_pos": 1960, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "test fired", "after": "test-fired", "start_char_pos": 2151, "end_char_pos": 2161, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 141, 195, 329, 470, 613, 771, 928, 1029, 1149, 1241, 1305, 1348, 1404, 1605, 1691, 1875, 1986], "domain": "wiki"}
{"doc_id": "29326123", "revision_depth": 1, "before_revision": "Research has examined goal orientation as a motivation variable that is useful for recruitment, climate and culture, performance appraisal, and choice.DeGeest, D., & Brown, K. G. (2011). The role of goal orientation in leadership development. Human Resource Development Quarterly, 22(2), 157-175. Studies have also used goal orientation to predict sales performance, adaptive performanceStasielowicz, L. (2019). Goal orientation and performance adaptation: A meta-analysis. Journal of Research in Personality, 82., goal setting, learning and adaptive behaviors in training, and leadership. Historical perspective Dweck proposed that there are two types of goal orientation: learning orientation and performance orientation.Dweck, C.S. (1986). Motivational processes affecting learning. American Psychologist, 41, 1040-1048. Dweck postulated that children with learning goals were believed to approach situations to master the acquisition of new skills, while children with performance goals were believed to approach situations to gain approval from peers and teachers. Like Eison, Dweck conceptualized goal orientation as a two-dimension construct. Individuals with a learning goal orientation (sometimes referred to as mastery goal orientation; abbreviated as LGO) seek to develop their competence by acquiring new skills and mastering new situations. They are not concerned about their performance relative to others, but rather with furthering their understanding of a given topic or task.Hendricks, J.W., & Payne, S.C. (2007). Beyond the Big Five: Leader goal orientation as a predictor of leadership effectiveness. Human Performance, 20, 317-343. Individuals with a performance goal orientation seek to demonstrate and validate the adequacy of their competence to receive favorable compliments while avoiding negative judgments. Although Dweck's work in this area built on the foundation laid by Nicholls, the fundamental difference between the two scholars' works was the attribution of an individual's goal orientation: Nicholls believed that the goal orientation held by an individual was a result of the possession of either an internal or external referent, while Dweck considered the adoption of a particular goal orientation to be related to the theory of intelligence held by that individual. Subsequent work by Eison and colleagues in 1982 led to a change in the conceptualization of these orientations from two ends of a continuum to two separate constructs.Eison, J.A., Pollio, H., & Milton, O. (1982). LOGO II: A user's manual. Knoxville: University of Tennessee, Learning Research Center. More recently, researchers have embraced the idea that individuals can adopt the two orientation styles simultaneously: persons can be independently high or low in learning and performance orientations. Ultimately, individuals can entertain multiple competing goal orientations at the same time, and strive to both outperform competitors and improve their performance. This line of thinking led to the conceptualization of two separate continua: one for learning goal orientation and one for performance goal orientation.Eison, J. A., Pollio, H., & Milton, O. (1986). Educational and personal characteristics of four different types of learning- and grade-oriented students. Contemporary Educational Psychology, 11, 5467. State versus trait There has been great debate as to whether goal orientation should be operationalized as a state or as a trait. Throughout the goal orientation literature, there are inconsistencies about the conceptualization of the stability of the construct. For example, DeShon & Gillespie stated that in the literature, DeShon, R. P., & Gillespie, J. Z. (2005). A motivated action theory account of goal orientation. Journal of Applied Psychology, 90(6), 1096-1127. goal orientation has been conceptualized as a trait, quasi-trait, and state. They states that whether researchers conceptualize goal orientation as a trait or a state \"depends on the breadth of the inference that the researcher is attempting to support\". State goal orientation refers to the goal one has in a particular situation . It is similar to trait goal orientation in that it represents one's preference in an achievement situation. Button, Mathieu, & Zajac take an integrative view of the construct,Button, S. B., Mathieu, J. E., & Zajac, D. M. (1996, July). Goal orientation URLanizational research: A conceptual and empirical foundation. Organizational Behavior and Human Decision Processes, 67(1), 26-48. stating that goal orientation is best categorized as a relatively stable individual difference variable that can be influenced by situational and contextual characteristics. They found that when few situational cues are present, individuals will adopt their dispositional goal orientations. However, when \"dispositional goal orientations predispose individuals to adopt particular response patterns across situations, situational characteristics may cause them to adopt a different or less acute response pattern for a specific situation\". Therefore, trait and state goal orientations interact, and both should be considered simultaneously. Types Since the realization that performance goal orientation is most effectively split into two separate parts, researchers have conducted validation studies to demonstrate the statistical and conceptual distinction of three dimensions to goal orientation. Conceptual and empirical work by Elliot and Church and VandeWalle demonstrated that the factor structure of goal orientation lends itself to three distinct dimensions as summarized below .Elliot, A. J. Church, M. A. (1997). A hierarchical model of approach and avoidance achievement motivation. Journal of Personality and Social Psychology, 72, 218-232. An explanation of the learning-approach and learning-avoidance goal orientations are also included for completeness. Antecedents Throughout the goal orientation literature, many studies have examined relationships between goal orientation and various antecedents. These antecedents have been identified to have varying levels of importance. In a meta-analysis by Payne et al., both the need for achievement and the Big Five personality traits were identified as important antecedents of goal orientation, while cognitive ability was found to have almost no relationship with goal orientation. The following sections go into more detail about each antecedent. Payne and her colleagues did not distinguish between proximal and distal antecedents. Another factor to consider when examining the relationship between goal orientation and goal setting is the level of inherent complexity in the situation or task. In situations with more complex tasks, it appears that \"do your best\" goals may lead to higher performance than specific goals. It is possible that in complex tasks, a specific, difficult goal imposes greater cognitive demands on employees, making it difficult for them to learn the complex task due to this increased pressure.Latham, G.P & Seijts, G.H. (1999). The effects of proximal and distal goals on performance on a moderately complex task. Journal of Organizational Behavior, 2, 81-127. Kanfer and Ackerman found that in an air traffic controller simulation (a highly complex task),Kanfer, R. Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude-treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74, 657-690. performance-outcome goals interfered with acquiring the knowledge necessary to perform the task, and individuals performed better when they were asked to do their best, suggesting that adopting a learning orientation may be appropriate for complex tasks or in specific settings. However, it may be possible to set a specific, difficult learning goal. Latham and Brown found that when MBA students set specific, difficult learning goals such as mastering complex course material,Latham, G. P. Brown, T. C. (2006). The effect of learning vs. outcome goals on self-efficacy and satisfaction in an MBA program. Applied Psychology: An International Review, 55, 606-623. they outperformed MBA students who set a performance goal for GPA. Locke and Latham claim that creating a specific, difficult learning goal in this type of situation facilitates meta-cognition which is particularly helpful in complex environments with limited guidance, such as in an MBA program. Organizational outcomes Goal orientation has also been linked URLanizational outcomes, specifically job performance. Payne et al. found that individuals with high levels of trait and state LGO and low levels of trait APGO had better job performance, that PPGO was unrelated to performance, and that LGO predicted job performance above and beyond both cognitive ability and the Big Five personality characteristics. Their findings suggest that LGO is a valuable predictor of job performance and it may be in the best interest URLanizations to create a climate in which learning is valued over performance.", "after_revision": "Research has examined goal orientation as a motivation variable that is useful for recruitment, climate and culture, performance appraisal, and choice.DeGeest, D., & Brown, K. G. (2011). The role of goal orientation in leadership development. Human Resource Development Quarterly, 22(2), 157-175. It has also been used to predict sales performance, adaptive performanceStasielowicz, L. (2019). Goal orientation and performance adaptation: A meta-analysis. Journal of Research in Personality, 82., goal setting, learning and adaptive behaviors in training, and leadership. Historical perspective Dweck proposed that there are two types of goal orientation: learning orientation and performance orientation.Dweck, C.S. (1986). Motivational processes affecting learning. American Psychologist, 41, 1040-1048. Dweck postulated that children with learning goals were believed to approach situations to master the acquisition of new skills, while children with performance goals were believed to approach situations to gain approval from peers and teachers. Like Eison, Dweck conceptualized goal orientation as a two-dimension construct. Individuals with a learning goal orientation (sometimes referred to as mastery goal orientation; abbreviated as LGO) seek to develop their competence by acquiring new skills and mastering new situations. They are not concerned about their performance relative to others, but rather with furthering their understanding of a given topic or task.Hendricks, J.W., & Payne, S.C. (2007). Beyond the Big Five: Leader goal orientation as a predictor of leadership effectiveness. Human Performance, 20, 317-343. Individuals with a performance goal orientation seek to demonstrate and validate the adequacy of their competence to receive favorable compliments while avoiding negative judgments. Although Dweck's work in this area built on the foundation laid by Nicholls, the fundamental difference between the two scholars' works is the attribution of an individual's goal orientation: Nicholls believed that the goal orientation held by an individual was a result of the possession of either an internal or external referent, while Dweck considered the adoption of a particular goal orientation to be related to the theory of intelligence held by that individual. Subsequent work by Eison and colleagues in 1982 led to a change in the conceptualization of these orientations from two ends of a continuum to two separate constructs.Eison, J.A., Pollio, H., & Milton, O. (1982). LOGO II: A user's manual. Knoxville: University of Tennessee, Learning Research Center. More recently, researchers have embraced the idea that individuals can adopt the two orientation styles simultaneously: individuals can be independently high or low in learning and performance orientations. Ultimately, they can entertain multiple competing goal orientations at the same time, and strive to both outperform competitors and improve their performance. This line of thinking led to the conceptualization of two separate continua: one for learning goal orientation and one for performance goal orientation.Eison, J. A., Pollio, H., & Milton, O. (1986). Educational and personal characteristics of four different types of learning- and grade-oriented students. Contemporary Educational Psychology, 11, 5467. State versus trait There has been debate as to whether goal orientation should be operationalized as a state or as a trait. Throughout the goal orientation literature, there are inconsistencies about the conceptualization of the stability of the construct. For example, DeShon & Gillespie stated that goal orientation has been conceptualized as a trait, quasi-trait, and state. DeShon, R. P., & Gillespie, J. Z. (2005). A motivated action theory account of goal orientation. Journal of Applied Psychology, 90(6), 1096-1127. They assert that whether researchers conceptualize goal orientation as a trait or a state \"depends on the breadth of the inference that the researcher is attempting to support\". State goal orientation refers to the goal one has in a particular situation , and is similar to trait goal orientation in that it represents one's preference in an achievement situation. Button, Mathieu, & Zajac take an integrative view of the construct,Button, S. B., Mathieu, J. E., & Zajac, D. M. (1996, July). Goal orientation URLanizational research: A conceptual and empirical foundation. Organizational Behavior and Human Decision Processes, 67(1), 26-48. stating that goal orientation is best categorized as a relatively stable individual difference variable that can be influenced by situational and contextual characteristics. They found that when few situational cues are present, individuals adopt their dispositional goal orientations. However, when \"dispositional goal orientations predispose individuals to adopt particular response patterns across situations, situational characteristics may cause them to adopt a different or less acute response pattern for a specific situation\". Therefore, trait and state goal orientations interact, and both should be considered simultaneously. Types Since the realization that performance goal orientation is most effectively split into two separate parts, researchers have conducted validation studies to demonstrate the statistical and conceptual distinction of three dimensions to goal orientation. Conceptual and empirical work by Elliot and Church and VandeWalle demonstrated that the factor structure of goal orientation lends itself to three distinct dimensions .Elliot, A. J. Church, M. A. (1997). A hierarchical model of approach and avoidance achievement motivation. Journal of Personality and Social Psychology, 72, 218-232. An explanation of the learning-approach and learning-avoidance goal orientations are also included for completeness. Antecedents Many studies have examined relationships between goal orientation and various antecedents. These antecedents have been identified to have varying levels of importance. In a meta-analysis by Payne et al., both the need for achievement and the Big Five personality traits were identified as important antecedents of goal orientation, while cognitive ability was found to have almost no relationship with goal orientation. The following sections go into more detail about each antecedent. Payne and her colleagues did not distinguish between proximal and distal antecedents. Another factor to consider when examining the relationship between goal orientation and goal setting is the level of inherent complexity in the situation or task. In situations with more complex tasks, \"do your best\" goals may lead to higher performance than specific goals. It is possible that in complex tasks, a specific, difficult goal imposes greater cognitive demands on employees, making it difficult for them to learn the complex task due to this increased pressure.Latham, G.P & Seijts, G.H. (1999). The effects of proximal and distal goals on performance on a moderately complex task. Journal of Organizational Behavior, 2, 81-127. Kanfer and Ackerman found that in an air traffic controller simulation (a highly complex task),Kanfer, R. Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude-treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74, 657-690. performance-outcome goals interfered with acquiring the knowledge necessary to perform the task, and individuals performed better when they were asked to do their best, suggesting that adopting a learning orientation may be appropriate for complex tasks or in specific settings. However, it may be possible to set a specific, difficult learning goal. Latham and Brown found that when MBA students set specific, difficult learning goals such as mastering complex course material,Latham, G. P. Brown, T. C. (2006). The effect of learning vs. outcome goals on self-efficacy and satisfaction in an MBA program. Applied Psychology: An International Review, 55, 606-623. they outperformed MBA students who set a performance goal for GPA. Locke and Latham claim that creating a specific, difficult learning goal in this type of situation facilitates meta-cognition which is particularly helpful in complex environments with limited guidance, such as in an MBA program. Organizational outcomes Goal orientation has also been linked URLanizational outcomes, specifically job performance. Payne et al. found that individuals with high levels of trait and state LGO and low levels of trait APGO had better job performance, PPGO was unrelated to performance, and LGO predicted job performance more accurately than both cognitive ability and the Big Five personality characteristics. Their findings suggest that LGO is a valuable predictor of job performance and it may be in the best interest URLanizations to create a climate in which learning is valued over performance.", "edit_actions": [{"type": "R", "before": "Studies have also used goal orientation", "after": "It has also been used", "start_char_pos": 297, "end_char_pos": 336, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "was", "after": "is", "start_char_pos": 1971, "end_char_pos": 1974, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "persons", "after": "individuals", "start_char_pos": 2728, "end_char_pos": 2735, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "individuals", "after": "they", "start_char_pos": 2823, "end_char_pos": 2834, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "great", "after": null, "start_char_pos": 3364, "end_char_pos": 3369, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "in the literature,", "after": "goal orientation has been conceptualized as a trait, quasi-trait, and state.", "start_char_pos": 3637, "end_char_pos": 3655, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "goal orientation has been conceptualized as a trait, quasi-trait, and state. They states", "after": "They assert", "start_char_pos": 3802, "end_char_pos": 3890, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": ". It", "after": ", and", "start_char_pos": 4133, "end_char_pos": 4137, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}, {"type": "D", "before": "will", "after": null, "start_char_pos": 4760, "end_char_pos": 4764, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "as summarized below", "after": null, "start_char_pos": 5585, "end_char_pos": 5604, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Throughout the goal orientation literature, many", "after": "Many", "start_char_pos": 5901, "end_char_pos": 5949, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "it appears that", "after": null, "start_char_pos": 6719, "end_char_pos": 6734, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "that", "after": null, "start_char_pos": 8674, "end_char_pos": 8678, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "that", "after": null, "start_char_pos": 8718, "end_char_pos": 8722, "major_intent": "coherence", "raw_intents": ["coherence", "others", "fluency"]}, {"type": "R", "before": "above and beyond", "after": "more accurately than", "start_char_pos": 8753, "end_char_pos": 8769, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}], "sents_char_pos": [0, 151, 186, 242, 296, 411, 473, 589, 723, 742, 785, 823, 1069, 1149, 1246, 1353, 1493, 1531, 1620, 1652, 1834, 2306, 2474, 2519, 2545, 2607, 2810, 2976, 3129, 3175, 3282, 3329, 3459, 3592, 3697, 3752, 3878, 4056, 4134, 4242, 4369, 4450, 4692, 4809, 5058, 5159, 5417, 5641, 5712, 5771, 5888, 6035, 6112, 6364, 6430, 6516, 6679, 6807, 7007, 7041, 7127, 7174, 7304, 7417, 7740, 7812, 7974, 8068, 8193, 8423, 8540, 8838], "domain": "wiki"}
{"doc_id": "299609", "revision_depth": 1, "before_revision": "Psychological projection is a defense mechanism in which the ego defends itself against unconscious impulses or qualities (both positive and negative) by denying their existence in themselves by attributing them to others.Sigmund Freud, Case Histories II (PFL 9) p. 132 For example, a bully may project their own feelings of vulnerability onto the target, or a person who is confused will project their own feelings of confusion and inadequacy onto other people. Instead, Newman, Duff, and Baumeister (1997) proposed a new model of defensive projection . In this view, repressors try to suppress thoughts of their undesirable traits , and these efforts make those trait categories highly accessibleso that they are then used all the more often when forming impressions of others. The projection is then only a by-product of the real defensive mechanism.", "after_revision": "Psychological projection is a defense mechanism in which the ego defends itself against unconscious impulses or qualities (both positive and negative) by denying their existence in themselves and attributing them to others.Sigmund Freud, Case Histories II (PFL 9) p. 132 For example, a bully may project their own feelings of vulnerability onto the target, or a person who is confused will project their own feelings of confusion and inadequacy onto other people. Instead, Newman, Duff, and Baumeister (1997) proposed a new model of defensive projection in which the repressors efforts to suppress thoughts of their undesirable traits make those trait categories highly accessibleso that they are then used all the more often when forming impressions of others. The projection is then only a byproduct of the real defensive mechanism.", "edit_actions": [{"type": "R", "before": "by", "after": "and", "start_char_pos": 192, "end_char_pos": 194, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ". In this view, repressors try", "after": "in which the repressors efforts", "start_char_pos": 553, "end_char_pos": 583, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "D", "before": ", and these efforts", "after": null, "start_char_pos": 633, "end_char_pos": 652, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "by-product", "after": "byproduct", "start_char_pos": 810, "end_char_pos": 820, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 222, 462, 554, 779], "domain": "wiki"}
{"doc_id": "299609", "revision_depth": 2, "before_revision": "The method of managed projection is a type of projective technique. The basic principle of the method is that a subject is presented with their own verbal portrait named by the name of another person, as well as with a portrait of their fictional opposition (V. V. Stolin, 1981). Some studies were critical of Freud's theory. Research on social projection supports the existence of a false-consensus effect whereby humans have a broad tendency to believe that others are similar to themselves, and thus \"project\" their personal traits onto others. This applies to good traits as well as bad traitsand is not a defense mechanism for denying the existence of the trait within the self. Instead, Newman, Duff, and Baumeister (1997) proposed a new model of defensive projection in which the repressors efforts to suppress thoughts of their undesirable traits make those trait categories highly accessible\u2014so that they are then used all the more often when forming impressions of others. The projection is then only a byproduct of the real defensive mechanism.", "after_revision": "The method of managed projection is a projective technique. The basic principle of this method is that a subject is presented with their own verbal portrait named by the name of another person, as well as with a portrait of their fictional opposition (V. V. Stolin, 1981). Some studies were critical of Freud's theory. Research on social projection supports the existence of a false-consensus effect whereby humans have a broad tendency to believe that others are similar to themselves, and thus \"project\" their personal traits onto others. This applies to both good and bad traits; it is not a defense mechanism for denying the existence of the trait within the self. Instead, Newman, Duff, and Baumeister (1997) proposed a new model of defensive projection in which the repressor's efforts to suppress thoughts of their undesirable traits make those trait categories highly accessible\u2014so that they are then used all the more often when forming impressions of others. The projection is then only a byproduct of the real defensive mechanism.", "edit_actions": [{"type": "D", "before": "type of", "after": null, "start_char_pos": 38, "end_char_pos": 45, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "the", "after": "this", "start_char_pos": 91, "end_char_pos": 94, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "good traits as well as bad traitsand", "after": "both good and bad traits; it", "start_char_pos": 564, "end_char_pos": 600, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "repressors", "after": "repressor's", "start_char_pos": 787, "end_char_pos": 797, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 67, 279, 325, 547, 683, 982], "domain": "wiki"}
{"doc_id": "30730412", "revision_depth": 1, "before_revision": "President Michelle Bachelet with presidential sash and the O'Higgins Pioche, 11 March 2014 The O'Higgins Pioche In Chile,  O'Higgins Pioche is a piece regarded as the true symbol of presidential power that is placed in the bottom of the presidential sash of office. The O'Higgins Pioche is a five-pointed star of about 7cm in diameter, enameled in red. This name was given a medal O'Higgins ordered to bring in the presidential sash. This relic was given by the hero Jos Gregorio Argomedo (17671830) after his abdication (1823). During the inauguration of the monument in the Alameda (1872), the descendants of Argomedo gave it to President Federico Errzuriz Zaartu, who put  the presidential sash in inaugurating a tradition that continues. The original piocha remained intact until the 1973 coup, when it disappeared during the bombing of La Moneda Palace. During the military regime was drawn up a new in the bases of original photographs . Only used in conjunction with the presidential sash. The \"Pioche de O'Higgins\" is the element in the transfer of power from one president to another symbolizes the transfer of power .Sabas Que?", "after_revision": "President Michelle Bachelet with presidential sash and the O'Higgins Pioche, 11 March 2014 The O'Higgins Pioche In Chile, the O'Higgins Pioche is a piece regarded as the true symbol of presidential power , which is placed in the lower end of the presidential sash of office. The O'Higgins Pioche is a five-pointed star of about 7cm in diameter, enameled in red. This was the name given to a medal that Bernardo O'Higgins ordered to be put on the presidential sash. The relic was a gift by O'Higgins to Jos Gregorio Argomedo (17671830) after his abdication on January 1823. During the inauguration of the monument in the Alameda (1872), descendants of Argomedo gave it to President Federico Errzuriz Zaartu, who put it on the presidential sash in inaugurating a tradition that continues. The original piocha remained intact until the 1973 coup, when it disappeared during the bombing of La Moneda Palace. During the military regime , a replica was made in the base of photographs of the original. The piocha is only to be used in conjunction with the presidential sash. The \"Pioche de O'Higgins\" is the element that symbolizes the transfer of power from one president to the next one during the ceremony of change of command .Sabas Que?", "edit_actions": [{"type": "A", "before": null, "after": "the", "start_char_pos": 122, "end_char_pos": 122, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "that", "after": ", which", "start_char_pos": 201, "end_char_pos": 205, "major_intent": "style", "raw_intents": ["style", "coherence", "others"]}, {"type": "R", "before": "bottom", "after": "lower end", "start_char_pos": 223, "end_char_pos": 229, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "name was given a medal", "after": "was the name given to a medal that Bernardo", "start_char_pos": 358, "end_char_pos": 380, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "bring in", "after": "be put on", "start_char_pos": 402, "end_char_pos": 410, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "This relic was given by the hero", "after": "The relic was a gift by O'Higgins to", "start_char_pos": 434, "end_char_pos": 466, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "others"]}, {"type": "R", "before": "(1823).", "after": "on January 1823.", "start_char_pos": 521, "end_char_pos": 528, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 592, "end_char_pos": 595, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}, {"type": "A", "before": null, "after": "it on", "start_char_pos": 675, "end_char_pos": 675, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "was drawn up a new in the bases of original photographs . Only", "after": ", a replica was made in the base of photographs of the original. The piocha is only to be", "start_char_pos": 886, "end_char_pos": 948, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "R", "before": "in", "after": "that symbolizes", "start_char_pos": 1038, "end_char_pos": 1040, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "another symbolizes the transfer of power", "after": "the next one during the ceremony of change of command", "start_char_pos": 1085, "end_char_pos": 1125, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 265, 352, 433, 528, 741, 858, 943, 996], "domain": "wiki"}
{"doc_id": "30780790", "revision_depth": 1, "before_revision": " The philatelic business activities of the Tuvalu Philatelic Bureau are authorised by the Tuvalu Philatelic Bureau Ordinance (1982). Karl Tili was the first Tuvaluan appointed to the post of general manager of the Tuvalu Philatelic Bureau and was the general manager from 1989 to 31 December 2011. The Government of Tuvalu are proposing to merge the Tuvalu Philatelic Bureau with the Tuvalu Post Office . The Tuvaluan post office is regulated by the Tuvalu Post Office Act 1977.  Addresses in Tuvalu Addressing in Tuvalu has been virtually non-existent. As at 2018, fewer than 10 streets in the capital Funafuti were named, and only 100 homes and 10 businesses had a postal address. Most people had to travel to send or retrieve mail, and some did not have access to postal services at all. In 2018, Tuvalu Post, the countrys official postal operator, made what3words a national standard for addresses, enabling home deliveries for the first time.Tuvalu Post starts delivering more with what3words Tuvalu Philatelic Bureau Newsletter The first newsletter was published in March 1976, under the name News and Views. In 1999 the name was changed to the Tuvalu Philatelic Bureau Newsletter. The Tuvalu Philatelic Bureau Newsletter is published 3 to 4 times a year and provides information about new stamps issues and articles about Tuvalu.", "after_revision": "Tuvalu became a member of the Universal Postal Union on 3 February 1981. The philatelic business activities are authorised by the Tuvalu Philatelic Bureau Ordinance (1982). Karl Tili was the first Tuvaluan general manager of the Bureau from 1989 to 31 December 2011. In 2013, the Tuvalu government was proposing to merge the Bureau with the Tuvalu Post Office , which is regulated by the Tuvalu Post Office Act 1977. The Tuvalu Post Office is not separately constituted and is a government department. Addresses in Tuvalu Addressing in Tuvalu has been virtually non-existent. As at 2018, fewer than 10 streets in the capital Funafuti were named, and only 100 homes and 10 businesses had a postal address. Most people had to travel to send or retrieve mail, and some did not have access to postal services at all. In 2018, Tuvalu Post, the countrys official postal operator, made what3words a national standard for addresses, enabling home deliveries for the first time.Tuvalu Post starts delivering more with what3words ", "edit_actions": [{"type": "A", "before": null, "after": "Tuvalu became a member of the Universal Postal Union on 3 February 1981.", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "D", "before": "of the Tuvalu Philatelic Bureau", "after": null, "start_char_pos": 36, "end_char_pos": 67, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "appointed to the post of", "after": null, "start_char_pos": 166, "end_char_pos": 190, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Tuvalu Philatelic Bureau and was the general manager", "after": "Bureau", "start_char_pos": 214, "end_char_pos": 266, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "The Government of Tuvalu are", "after": "In 2013, the Tuvalu government was", "start_char_pos": 298, "end_char_pos": 326, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "D", "before": "Tuvalu Philatelic", "after": null, "start_char_pos": 350, "end_char_pos": 367, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ". The Tuvaluan post office", "after": ", which", "start_char_pos": 403, "end_char_pos": 429, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "The Tuvalu Post Office is not separately constituted and is a government department.", "start_char_pos": 479, "end_char_pos": 479, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "D", "before": "Tuvalu Philatelic Bureau Newsletter The first newsletter was published in March 1976, under the name News and Views. In 1999 the name was changed to the Tuvalu Philatelic Bureau Newsletter. The Tuvalu Philatelic Bureau Newsletter is published 3 to 4 times a year and provides information about new stamps issues and articles about Tuvalu.", "after": null, "start_char_pos": 998, "end_char_pos": 1336, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 132, 297, 404, 478, 553, 682, 790, 947, 1114, 1187]}
{"doc_id": "32047492", "revision_depth": 1, "before_revision": "In the USA Wilderness Areas (NWPS) are the strictest category of protected area. They are designated by law in accordance with the Wilderness Act of 1964 by the United States Congress. They must be large enough to adequately preserve its resources and may only be entered on foot, by canoe or on horseback. Exceptions are the extensive areas of wilderness in Alaska. At the beginning of 2010 there were over 750 recognised Wilderness Areas .", "after_revision": "In the USA Wilderness Areas (NWPS) are the strictest category of protected area. They are designated by law in accordance with the Wilderness Act of 1964 by the United States Congress. They must be large enough to adequately preserve their resources and may only be entered on foot, by canoe or on horseback. Limited exceptions are the extensive areas of wilderness in Alaska. There are over 800 recognised wilderness areas .", "edit_actions": [{"type": "R", "before": "its", "after": "their", "start_char_pos": 234, "end_char_pos": 237, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": "Exceptions", "after": "Limited exceptions", "start_char_pos": 307, "end_char_pos": 317, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "R", "before": "At the beginning of 2010 there were over 750 recognised Wilderness Areas", "after": "There are over 800 recognised wilderness areas", "start_char_pos": 367, "end_char_pos": 439, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 80, 184, 306, 366], "domain": "wiki"}
{"doc_id": "3213090", "revision_depth": 1, "before_revision": "In many Hindu temples, the temple structure reflects the symbolism of the Hindu association of the spiritual transition from daily life to spiritual perfection as a journey through stages. Ambulatory passageways for circumambulation are present through which worshipers move in a clockwise direction, starting at the sanctuary doorway and moving inward toward the inner sanctum where the deity is enshrined. This is a translation of the spiritual concept of transition through levels in life into bodily movements by the worshipers as they move inwardly through ambulatory halls to the most sacred centre of spiritual energy of the deity. Circumambulation is done in a clockwise direction and in an odd rather than even number of times. Circumambulatory walking around the shrine, by keeping time, is a common form of Hindu prayer. The circumambulatory pathway made of stone around the shrine is called the Pradakshina path.", "after_revision": "In many Hindu temples, the temple structure reflects the symbolism of the Hindu association of the spiritual transition from daily life to spiritual perfection as a journey through stages. Passageways for circumambulation are present through which worshipers move in a clockwise direction, starting at the sanctuary doorway and moving inward toward the inner sanctum where the deity is enshrined. This is a translation of the spiritual concept of transition through levels in life into bodily movements by the worshipers as they move inwardly through ambulatory halls to the most sacred centre of spiritual energy of the deity. It is done in a clockwise direction and in an odd rather than even number of times. Circumambulatory walking around the shrine, by keeping time, is a common form of Hindu prayer. The ambulatory pathway made of stone around the shrine is called the Pradakshina path.", "edit_actions": [{"type": "R", "before": "Ambulatory passageways", "after": "Passageways", "start_char_pos": 189, "end_char_pos": 211, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Circumambulation", "after": "It", "start_char_pos": 639, "end_char_pos": 655, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "circumambulatory", "after": "ambulatory", "start_char_pos": 836, "end_char_pos": 852, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 188, 407, 638, 736, 831], "domain": "wiki"}
{"doc_id": "3213090", "revision_depth": 2, "before_revision": "Christianity In the Catholic Church, a priest sometimes circumambulates an altar while incensing it with a thurible. Also, at some Catholic shrines, it is a tradition to circumambulate around the cult object of the place, usually relics of a saint or an image of Jesus or the Virgin Mary. Often this is performed three times, as a reference to the Trinity. In Romania, there is an Easter custom to circumambulate the church three times by singing priests leading the people, just before finishing Easter Liturgy. It symbolizes the funerary procession of the burial of Jesus URL Circumambulation is common in many Eastern Orthodox and Oriental Orthodox services. In the Coptic tradition, during the liturgy, the priest circumambulates while an acolyte (altar boy) holds a cross high on the opposite side of the altar . This is also a common practice in Lutheran, Roman Catholic, Anglican and Methodist churches during Lent when Stations of the Cross services are celebrated. The priest along with altar servers process around the circumambulate around the interior of the church visiting each of the 14 stations.", "after_revision": "Christianity In the Catholic Church, a priest sometimes circumambulates an altar while incensing it with a thurible. Also, at some Catholic shrines, it is a tradition to circle the cult object of the place, usually relics of a saint or an image of Jesus or the Virgin Mary. Often this is performed three times, as a reference to the Trinity. In Romania, there is an Easter custom to process around the church three times by singing priests leading the people, just before finishing Easter Liturgy. It symbolizes the funerary procession of the burial of Jesus URL Circumambulation is common in many Eastern Orthodox and Oriental Orthodox services. In the Coptic tradition, during the liturgy, the priest circles the altar while an acolyte (altar boy) holds a cross high on the opposite side . This is also a common practice in Lutheran, Roman Catholic, Anglican and Methodist churches during Lent when Stations of the Cross services are celebrated. The priest along with altar servers process around the interior of the church visiting each of the 14 stations.", "edit_actions": [{"type": "R", "before": "circumambulate around", "after": "circle", "start_char_pos": 170, "end_char_pos": 191, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "circumambulate", "after": "process around", "start_char_pos": 398, "end_char_pos": 412, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "circumambulates", "after": "circles the altar", "start_char_pos": 718, "end_char_pos": 733, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "of the altar", "after": null, "start_char_pos": 803, "end_char_pos": 815, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "around the circumambulate", "after": null, "start_char_pos": 1018, "end_char_pos": 1043, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 116, 288, 356, 512, 661, 817, 973], "domain": "wiki"}
{"doc_id": "325963", "revision_depth": 1, "before_revision": "Description Bee pollinating a sunflower. Pollen is transferred from anther of one plant to stigma of another as bee collects nectar Pollen tubes are unique to plants and their structures have evolved over the history of seed plants . Pollen tube formation is complex and the mechanism is not fully understood, but is of great interest to scientists . Pollen tube formation is important for sexual reproduction to occur in seed plants. Wind, water or pollinators transport pollen grains to the female gametophyte. Once a pollen grain has implanted on a compatible stigma via self-incompatibility mechanisms, the germination process is initiated. During this process, the pollen grain undergoes a conformational change whereupon a given section begins to protrude outwards to form a tube-like structure, known as the pollen tube. This structure rapidly descends down the length of the style via tip-directed growth, reaching rates of 1cm/h, whilst carrying two non-motile sperm cells. Upon reaching the ovule the pollen tube ruptures, thereby delivering the sperm cells to the female gametophyte, ultimately resulting in a double fertilization event. The first fertilization event produces a diploid zygote and the second fertilization event produces a triploid endosperm.", "after_revision": "Description Bee pollinating a sunflower. Pollen is transferred from anther of one plant to stigma of another as bee collects nectar Pollen tubes are unique to plants and their structures have evolved over the history of seed plants . Pollen tube formation is complex and the mechanism is not fully understood, but is of great interest to scientists because pollen tubes transport the male gametes produced by pollen grains to the female gametophyte. Once a pollen grain has implanted on a compatible stigma , its germination is initiated. During this process, the pollen grain begins to bulge outwards to form a tube-like structure, known as the pollen tube. The pollen tube structure rapidly descends down the length of the style via tip-directed growth, reaching rates of 1cm/h, whilst carrying two non-motile sperm cells. Upon reaching the ovule the pollen tube ruptures, thereby delivering the sperm cells to the female gametophyte, ultimately resulting in a double fertilization event. The first fertilization event produces a diploid zygote and the second fertilization event produces a triploid endosperm.", "edit_actions": [{"type": "R", "before": ". Pollen tube formation is important for sexual reproduction to occur in seed plants. Wind, water or pollinators transport", "after": "because pollen tubes transport the male gametes produced by", "start_char_pos": 349, "end_char_pos": 471, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "via self-incompatibility mechanisms, the germination process", "after": ", its germination", "start_char_pos": 570, "end_char_pos": 630, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "undergoes a conformational change whereupon a given section begins to protrude", "after": "begins to bulge", "start_char_pos": 683, "end_char_pos": 761, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "This", "after": "The pollen tube", "start_char_pos": 828, "end_char_pos": 832, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 40, 434, 512, 644, 827, 982, 1148], "domain": "wiki"}
{"doc_id": "33605176", "revision_depth": 1, "before_revision": "The Sullubawa are a Fulani clan in Northern Nigeria, the current ruling houses of Kano, Ringim and Katsina belong to the clan and another amongst the four ruling Houses of Zazzau Emirate . They are also found in Kano, Jigawa, Katsina and Sokoto states. They are said to have originated from Futa Toro, in what is now Senegal, and are associated with the Torodbe (Toronkawa) , the Sullubawa originated from Sissilo the husband of Cippowo  a sister of Uthman Toroddo ancestor of Usman dan Fodio as such they are regarded as the cousins of Toronkawa, they initially belong to the Wangarawa stock, they had Mandigo element in their ancestry and they were related to the Mandika, mandika are the  descendants of the Mali Empire , the Sullubawa spoke Wakore before they became absorbed into the Fulani group losing their original language and adopting the Fulani language. They played a key role in the Fulani Jihad led by Usman dan Fodio, which founded the Sokoto Caliphate. Sullubawa became \"hereditary beneficiaries of all positions of authority in all but one Hausa state\". In the 19th century, the Sullubawa controlled many of the fiefdoms of the Kano Emirate. The clan benefitted from British colonization and indirect rule which saw their influence increase. Sullubawa clansmen later attained positions of power following the fall of the first republic ; with one of them Umaru Musa Yar'Adua becoming President of Nigeria.", "after_revision": "The Sullubawa are a Fulani royal clan that historically featured prominently in the Fulani War, which founded the Sokoto Caliphate. The ruling houses of the ancient Sahelian city-states of Kano and Katsina belong to the clan and another amongst the four ruling Houses of Zazzau . They are also found in Kano, Jigawa, Katsina and Sokoto states. They are said to have originated from Futa Toro, in what is now Senegal, and are cousins with the Torodbe (Toronkawa) from Sissilo, the husband of Cippowo , a sister of Uthman Toroddo ancestor of Usman dan Fodio . The clan became culturally assimilated with the Fulani adopting the culture and language. The ancestral origin of the Sullubawa is bild as-sdn ( ) and they are descendants of the Mali Empire and the Wangarawa mercantile tribe which dominated the gold trade in the region. They played a key role in the Fulani Jihad led by Usman dan Fodio, which founded the Sokoto Caliphate. Sullubawa became \"hereditary beneficiaries of all positions of authority in all but one Hausa state\". In the 19th century, the Sullubawa controlled many of the fiefdoms of the Kano Emirate. The clan benefitted from British colonization and indirect rule which saw their influence increase. Sullubawa clansmen later attained positions of power following independence ; with one of them Umaru Musa Yar'Adua becoming President of Nigeria.", "edit_actions": [{"type": "R", "before": "clan in Northern Nigeria, the current", "after": "royal clan that historically featured prominently in the Fulani War, which founded the Sokoto Caliphate. The", "start_char_pos": 27, "end_char_pos": 64, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Kano, Ringim", "after": "the ancient Sahelian city-states of Kano", "start_char_pos": 82, "end_char_pos": 94, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": "Emirate", "after": null, "start_char_pos": 179, "end_char_pos": 186, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "associated", "after": "cousins", "start_char_pos": 334, "end_char_pos": 344, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", the Sullubawa originated from Sissilo", "after": "from Sissilo,", "start_char_pos": 374, "end_char_pos": 413, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 437, "end_char_pos": 437, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "as such they are regarded as the cousins of Toronkawa, they initially belong to the Wangarawa stock, they had Mandigo element in their ancestry and they were related to the Mandika, mandika are", "after": ". The clan became culturally assimilated with the Fulani adopting the culture and language. The ancestral origin of", "start_char_pos": 493, "end_char_pos": 686, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "others"]}, {"type": "A", "before": null, "after": "Sullubawa is bild as-sdn ( ) and they are", "start_char_pos": 691, "end_char_pos": 691, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": ", the Sullubawa spoke Wakore before they became absorbed into the Fulani group losing their original language and adopting the Fulani language.", "after": "and the Wangarawa mercantile tribe which dominated the gold trade in the region.", "start_char_pos": 723, "end_char_pos": 866, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "the fall of the first republic", "after": "independence", "start_char_pos": 1323, "end_char_pos": 1353, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 188, 252, 373, 866, 969, 1071, 1159, 1259, 1355], "domain": "wiki"}
{"doc_id": "33653387", "revision_depth": 2, "before_revision": "In this way, Persian became a second language to many across North India;  Alam contends that it neared the status of a first language. By the 18th century, many Indians in the north of the subcontinent had a \"native speaker's competence in Persian\".Faruqi, Shamsur Rahman (2003), \"A Long History of Urdu Literarature, Part 1\", in Pollock (ed.), Literary cultures in history: reconstructions from South Asia, p. 849, Decline Following Aurangzeb's death, Persian began to fall into decline, being displaced by Hindustani (Hindi-Urdu). The arrival and strengthening of British political power added a growing influence of English as well. However, for a long time Persian was still the dominant language of the subcontinent, used in education, Muslim rule, the judiciary, and literature. The East India Company acknowledged Persian as a \"language of command\", necessary to communicate with and control the Indian population. Hence many British officials arriving in India learned Persian in colleges established by the Company. The teachers in these colleges were often Indian. In some cases, Britishers even took over as Persian professors, sidelining the role of the Indians. Deccan A Persian poem produced in the Deccan. 17th century.|346x346px Although considerably distanced from North India, the Deccan was also a recipient of Persian's linguistic impact. Persianate culture was brought to the Deccan fleetingly through the efforts of the Delhi Sultanate in the early 14th century. Persian finally gained a foothold in the region with the establishment of the Bahmani Sultanate in 1347, which used the language for official purposes. The dynasty had a great interest in Persian culture, and several members were proficient in the language, producing their own literature. Literati from Northern India found themselves welcome at the court, and scholars from Iran were invited as well. Madrasas were built over the expanse of the kingdom, most notably the Mahmud Gawan Madrasa at Bidar, where Persian was taught.  In spite of this, Persian did not achieve the linguistic penetration it had in the north of the subcontinent. Richard Eaton writes that Persian was much less widely understood in the  region than vernacular languages  . During the turn of the 16th century, the Bahmani Sultanate splintered into the Deccan Sultanates, which were also Persianate in culture. They used Persian as a courtly language, as well as for official and administrative purposes. The language received literary patronage  . However, the sultans simultaneously promoted regional languages such as Telugu, Marathi, and Deccani (the southern variety of Hindustani), at times even using them in administration.  The reign of the Deccan Sultanates lasted until the late 17th century, when Aurangzeb brought the region into the Mughal Empire and established Aurangabad as the Deccan's capital. As was Mughal practice, Persian was used in many levels of administration. Major cities of this region that served as centres of Persian patronage included Gulbarga, Bidar, Bijapur, and Golconda. Notable Persian literature produced includes Futuh-us-Salatin, and the Bahmannamh . Influence on Indic languages As a prestige language and lingua franca over a period of 800 years in the Indian subcontinent, Classical Persian exerted a vast influence over numerous Indic languages. Generally speaking, the degree of impact is seen to increase the more one moves towards the north-west of the subcontinent, i.e. the Indo-Iranian frontier. For example, the Indo-Aryan languages have the most impact from Persian; this ranges from a high appearance in Punjabi, Sindhi, Kashmiri, and Gujarati, to more moderate representation in Bengali and Marathi. The largest foreign element in the Indo-Aryan languages is Persian. Conversely, the Dravidian languages have seen a low level of influence from Persian. They still feature loans from the language, some of which are direct, and some through Deccani (the southern variety of Hindustani), due to the Islamic rulers of the Deccan. Telugu is a good example of this (see Telangana Baasha for the Telugu dialect accentuated with these features). A categorised list of Persian vocabulary found in the Indic languages is provided below, and is far from exhaustive: +Loan categoryExamplesNounsProper namesMuslim names: Akhtar, Nawaz, Aftab, Dilshad Shah Bano, Zarina Non-Muslim names: Bahadur Shah, Chaman Lal, Iqbal Singh, Lal Bahadur, Roshan LalTitlesKhan Bahadur, Rai Bahadur, Yavar Jung, Salaar JungParts of the bodyjism (body), khn (blood), nkhn (nail, of fingers and toes), sn (chest), dil (heart), chehr (face), gardan (neck), zabn (tongue), halaq (throat)Place names (Suffixes)-bd, -stan, -ganj, -bagh, -sari (Hyderabad, Pakistan, Hazratganj, Arambagh, Mughalsarai)Kinship termsdmd (son-in-law), bb (father), shauhar (husband), birdar (brother)Foodsabz (vegetables), nn (bread), korm (curry) gsht (meat), km (minced meat), tandri (roasted) Clothingpaushk (dress), pajm (pyjamas), kamz (shirt), jeb (pocket), astar (inner, lining)Housegusalkhn (bathroom), pkhn (toilet), bwarchkhn (kitchen), darwz (door), diwr (wall)Ornamentszewar (ornaments), gulband (necklace), dastband (bracelet), pazeb (anklet)Fruitsseb (apple), anr (pomegranate), angr (grapes), nrang (tangerine), bdm (almond), kishmish (raisin)Vegetablesshalgam (turnip), qadd (pumpkin), sakarqand (sweet potato)Floracinr (plane tree), hin (henna), banafsh (pansy), gulb (rose), nlofar (water lily), ysmn (jasmine)Faunasher (lion), khargosh (rabbit), bulbul (nightingale), bz (falcon), kabutar (pigeon)Professionsdarz (tailor), hajjm (barber), sabz-farosh (greengrocer), khnsm (cook)Agriculturefasl (crop), rabi (spring), khrf (autumn), bpash (watering), nahar (canal), zamn (land)Timerz (day), sl (year), zamn (era)Lawadlat (court), qnn (law), muddai (plaintiff), vakl (lawyer), muakil (client)Administrationdarbr (court), pdshah (emperor), tehsldr (tax collector), zila (district)Writingqalam (pen), dawt (inkpot), syhi (ink), kgaz (paper)Religion (non-Arabic terms)rza (fasting), namz (ritual prayer), khud (God), paighambar (Prophet), bihisht (Paradise), pr (Sufi master)Measurementgaz (yard), ml (a mile), man (a mound), ser (a seer), murabb (square)Militarysiph (soldier), top (gun/cannon), topch (gunner), topkhn (artillery)Miscellaneousin (mirror), bzr (market), dst (friend), shehar (city)OtherModifiers:bilkul (surely/definitely), garm (hot), tz (fresh), zd (free/independent)Other Function words:khud (oneself), magar (but), lekin (but), afss (alas), shabsh (well done)Sources in order of importance: The Persian language is now largely defunct in the Indian subcontinent. However, it still lingers in some scholarly and literary circles; for example, the University of Kashmir in Srinagar has been publishing the Persian-language journal Dnish since 1969. Some colleges and universities in India, Pakistan, and Bangladesh offer Persian as a course of study.  Though Arabic largely dominates the realm of Islamic liturgy and theology in the Indian subcontinent, Persian can also be seen in some religious spheres: the dhikr sessions of Sufism often employ Persian poetry in song, and the Sufi devotional music genre of qawwali also uses Persian in parallel with local languages. Famed qawwali singer Nusrat Fateh Ali Khan sang sometimes in Persian.", "after_revision": "In this way, Persian became a second language to many across North India; Muzaffar Alam contends that it neared the status of a first language. By the 18th century, many Indians in the north of the subcontinent had a \"native speaker's competence in Persian\".Faruqi, Shamsur Rahman (2003), \"A Long History of Urdu Literarature, Part 1\", in Pollock (ed.), Literary cultures in history: reconstructions from South Asia, p. 849, Decline Following Aurangzeb's death, Persian began to fall into decline, being displaced by Hindustani (Hindi-Urdu). The arrival and strengthening of British political power added a growing influence of English as well. However, for a long time Persian was still the dominant language of the subcontinent, used in education, Muslim rule, the judiciary, and literature. The East India Company acknowledged Persian as a \"language of command\", necessary to communicate with and control the Indian population. Hence many British officials arriving in India learned Persian in colleges established by the Company. The teachers in these colleges were often Indian. In some cases, Britishers even took over as Persian professors, sidelining the role of the Indians. Deccan A Persian poem produced in the Deccan. 17th century.|346x346px Although considerably distanced from North India, the Deccan was also a recipient of Persian's linguistic impact. Persianate culture was brought to the Deccan fleetingly through the efforts of the Delhi Sultanate in the early 14th century. Persian finally gained a foothold in the region with the establishment of the Bahmani Sultanate in 1347, which used the language for official purposes. The dynasty had a great interest in Persian culture, and several members were proficient in the language, producing their own literature. Literati from Northern India found themselves welcome at the court, and scholars from Iran were invited as well. Madrasas were built over the expanse of the kingdom, such as the the Mahmud Gawan Madrasa at Bidar, where Persian was taught. A notable poet patronised by the Bahmans was Abdul Isami, who wrote the first Persian history of the Muslim conquest of India titled Futuh-us-Salatin. In spite of this, Richard Eaton writes that Persian was much less widely understood in the Deccan region than vernacular languages , and contrasts this situation with the Persian proficiency in the north of the subcontinent . During the turn of the 16th century, the Bahmani Sultanate splintered into the Deccan Sultanates, which were also Persianate in culture. They used Persian as a courtly language, as well as for official and administrative purposes. The language received literary patronage ; for example, Persian poet Muhammad Zuhuri, author of the Sqnmah, was a prominent figure at Ibrahim Adil Shah II's court in Bijapur . However, the sultans simultaneously promoted regional languages such as Telugu, Marathi, and Deccani (the southern variety of Hindustani), at times even using them in administration. For example, Alam writes that Telugu was the language of the sultan for the Qutub Shahis, and that Persian was removed from the Bijapur Sultanate's administrative system by Ibrahim Adil Shah I in favour of Marathi; these are corroborated by Eaton. The reign of the Deccan Sultanates lasted until the late 17th century, when Aurangzeb brought the region into the Mughal Empire and established Aurangabad as the Deccan's capital. In the late 18th century, while Delhi courts began to turn to Urdu, literary cultivation in Persian remained healthy in the Deccan. It was finally eclipsed by Urdu in the 19th century . Influence on Indic languages As a prestige language and lingua franca over a period of 800 years in the Indian subcontinent, Classical Persian exerted a vast influence over numerous Indic languages. Generally speaking, the degree of impact is seen to increase the more one moves towards the north-west of the subcontinent, i.e. the Indo-Iranian frontier. For example, the Indo-Aryan languages have the most impact from Persian; this ranges from a high appearance in Punjabi, Sindhi, Kashmiri, and Gujarati, to more moderate representation in Bengali and Marathi. The largest foreign element in the Indo-Aryan languages is Persian. Conversely, the Dravidian languages have seen a low level of influence from Persian. They still feature loans from the language, some of which are direct, and some through Deccani (the southern variety of Hindustani), due to the Islamic rulers of the Deccan. A categorised list of Persian vocabulary found in the Indic languages is provided below, and is far from exhaustive: +Loan categoryExamplesNounsProper namesMuslim names: Akhtar, Nawaz, Aftab, Dilshad Shah Bano, Zarina Non-Muslim names: Bahadur Shah, Chaman Lal, Iqbal Singh, Lal Bahadur, Roshan LalTitlesKhan Bahadur, Rai Bahadur, Yavar Jung, Salaar JungParts of the bodyjism (body), khn (blood), nkhn (nail, of fingers and toes), sn (chest), dil (heart), chehr (face), gardan (neck), zabn (tongue), halaq (throat)Place names (Suffixes)-bd, -stan, -ganj, -bagh, -sari (Hyderabad, Pakistan, Hazratganj, Arambagh, Mughalsarai)Kinship termsdmd (son-in-law), bb (father), shauhar (husband), birdar (brother)Foodsabz (vegetables), nn (bread), korm (curry) gsht (meat), km (minced meat), tandri (roasted) Clothingpaushk (dress), pajm (pyjamas), kamz (shirt), jeb (pocket), astar (inner, lining)Housegusalkhn (bathroom), pkhn (toilet), bwarchkhn (kitchen), darwz (door), diwr (wall)Ornamentszewar (ornaments), gulband (necklace), dastband (bracelet), pazeb (anklet)Fruitsseb (apple), anr (pomegranate), angr (grapes), nrang (tangerine), bdm (almond), kishmish (raisin)Vegetablesshalgam (turnip), qadd (pumpkin), sakarqand (sweet potato)Floracinr (plane tree), hin (henna), banafsh (pansy), gulb (rose), nlofar (water lily), ysmn (jasmine)Faunasher (lion), khargosh (rabbit), bulbul (nightingale), bz (falcon), kabutar (pigeon)Professionsdarz (tailor), hajjm (barber), sabz-farosh (greengrocer), khnsm (cook)Agriculturefasl (crop), rabi (spring), khrf (autumn), bpash (watering), nahar (canal), zamn (land)Timerz (day), sl (year), zamn (era)Lawadlat (court), qnn (law), muddai (plaintiff), vakl (lawyer), muakil (client)Administrationdarbr (court), pdshah (emperor), tehsldr (tax collector), zila (district)Writingqalam (pen), dawt (inkpot), syhi (ink), kgaz (paper)Religion (non-Arabic terms)rza (fasting), namz (ritual prayer), khud (God), paighambar (Prophet), bihisht (Paradise), pr (Sufi master)Measurementgaz (yard), ml (a mile), man (a mound), ser (a seer), murabb (square)Militarysiph (soldier), top (gun/cannon), topch (gunner), topkhn (artillery)Miscellaneousin (mirror), bzr (market), dst (friend), shehar (city)OtherModifiers:bilkul (surely/definitely), garm (hot), tz (fresh), zd (free/independent)Other function words:khud (oneself), magar (but), lekin (but), afss (alas), shabsh (well done)Sources in order of importance: The Persian language is now largely defunct in the Indian subcontinent. However, it still lingers in some scholarly and literary circles; for example, the University of Kashmir in Srinagar has been publishing the Persian-language journal Dnish since 1969. Some colleges and universities in India, Pakistan, and Bangladesh offer Persian as a course of study. Commenting on the state of the field in 2008, Abidi and Gargesh wrote that there was a \"general lack of interest\" in Persian studies. Though Arabic largely dominates the realm of Islamic liturgy and theology in the Indian subcontinent, Persian can be seen in some religious spheres: the dhikr sessions of Sufism often employ Persian poetry in song, and the Sufi devotional music genre of qawwali also uses Persian in parallel with local languages. Famed qawwali singer Nusrat Fateh Ali Khan sang sometimes in Persian.", "edit_actions": [{"type": "A", "before": null, "after": "Muzaffar", "start_char_pos": 74, "end_char_pos": 74, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "most notably the", "after": "such as the the", "start_char_pos": 1942, "end_char_pos": 1958, "major_intent": "clarity", "raw_intents": ["clarity", "style", "others"]}, {"type": "A", "before": null, "after": "A notable poet patronised by the Bahmans was Abdul Isami, who wrote the first Persian history of the Muslim conquest of India titled Futuh-us-Salatin.", "start_char_pos": 2016, "end_char_pos": 2016, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "Persian did not achieve the linguistic penetration it had in the north of the subcontinent.", "after": null, "start_char_pos": 2035, "end_char_pos": 2126, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "Deccan", "start_char_pos": 2200, "end_char_pos": 2200, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ", and contrasts this situation with the Persian proficiency in the north of the subcontinent", "start_char_pos": 2234, "end_char_pos": 2234, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "; for example, Persian poet Muhammad Zuhuri, author of the Sqnmah, was a prominent figure at Ibrahim Adil Shah II's court in Bijapur", "start_char_pos": 2509, "end_char_pos": 2509, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "For example, Alam writes that Telugu was the language of the sultan for the Qutub Shahis, and that Persian was removed from the Bijapur Sultanate's administrative system by Ibrahim Adil Shah I in favour of Marathi; these are corroborated by Eaton.", "start_char_pos": 2695, "end_char_pos": 2695, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "As was Mughal practice, Persian was used in many levels of administration.", "after": null, "start_char_pos": 2876, "end_char_pos": 2950, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Major cities of this region that served as centres of Persian patronage included Gulbarga, Bidar, Bijapur, and Golconda. Notable Persian literature produced includes Futuh-us-Salatin, and the Bahmannamh", "after": "In the late 18th century, while Delhi courts began to turn to Urdu, literary cultivation in Persian remained healthy in the Deccan. It was finally eclipsed by Urdu in the 19th century", "start_char_pos": 2951, "end_char_pos": 3153, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": "Telugu is a good example of this (see Telangana Baasha for the Telugu dialect accentuated with these features).", "after": null, "start_char_pos": 4046, "end_char_pos": 4157, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "Function", "after": "function", "start_char_pos": 6467, "end_char_pos": 6475, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "Commenting on the state of the field in 2008, Abidi and Gargesh wrote that there was a \"general lack of interest\" in Persian studies.", "start_char_pos": 6945, "end_char_pos": 6945, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "also", "after": null, "start_char_pos": 7060, "end_char_pos": 7064, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 73, 135, 250, 533, 636, 785, 922, 1025, 1075, 1175, 1221, 1235, 1359, 1485, 1637, 1775, 1888, 2015, 2126, 2373, 2467, 2511, 2694, 2875, 2950, 3071, 3155, 3354, 3510, 3583, 3718, 3786, 3871, 4045, 4157, 4274, 4700, 4714, 4908, 4956, 4980, 4996, 5010, 5024, 5038, 5071, 5086, 5107, 5121, 5160, 5180, 5201, 5234, 5253, 5268, 5287, 5301, 5346, 5362, 5410, 5423, 5440, 5453, 5474, 5506, 5525, 5547, 5560, 5602, 5618, 5645, 5681, 5696, 5711, 5729, 5744, 5769, 5780, 5808, 5819, 5839, 5854, 5898, 5916, 6388, 6658, 6724, 6842, 6944, 7264], "domain": "wiki"}
{"doc_id": "33780062", "revision_depth": 1, "before_revision": "After the explosion, initial reports indicated that it had happened at a CNG station, but this was later denied. Iranian officials said that the blast at the missile base was an accident, and ruled out any URLanized by the United States and its regional allies. The Army of the Guardians of the Islamic Revolution said that the explosion \"had taken place in an missile arms depot when munitions were being moved\". On the Iranian parliament's website, icana.ir, Parviz Soroori, an Iranian MP, was quoted as saying, \"No sabotage was involved in this incident. It has nothing to do with politics\". However, Time magazine cited a \"western intelligence source\" as saying that Mossad was behind the blast, and other reports, which were attributed to western intelligence services , also mentioned the involvement of Israel's intelligence services. Speaking to The Guardian on condition of anonymity, an official with close links to the Iranian Government in Iran said that he believed the Bid Kaneh explosion was \"part of the covert war against Iran, led by Israel\".", "after_revision": "After the explosion, initial reports indicated that it had happened at a CNG station, but this was later denied. Iranian officials said that the blast at the missile base was an accident, and ruled out any URLanized by the United States and its regional allies. The Army of the Guardians of the Islamic Revolution said that the explosion \"had taken place in an missile arms depot when munitions were being moved\". On the Iranian parliament's website, icana.ir, Parviz Soroori, an Iranian MP, was quoted as saying, \"No sabotage was involved in this incident. It has nothing to do with politics\". However, Time magazine reported claims from a \"western intelligence source\" that Israeli intelligence agency Mossad had orchestrated the explosion; other reports, which were also attributed to intelligence services in the West, likewise mentioned the involvement of Israel's intelligence services. An \" official with close links to the Iranian Government \" spoke to The Guardian anonymously and claimed that the blast was \"part of the covert war against Iran, led by Israel\".", "edit_actions": [{"type": "R", "before": "cited", "after": "reported claims from", "start_char_pos": 618, "end_char_pos": 623, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "as saying that Mossad was behind the blast, and", "after": "that Israeli intelligence agency Mossad had orchestrated the explosion;", "start_char_pos": 656, "end_char_pos": 703, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "R", "before": "attributed to western intelligence services , also", "after": "also attributed to intelligence services in the West, likewise", "start_char_pos": 730, "end_char_pos": 780, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": "Speaking to The Guardian on condition of anonymity, an", "after": "An \"", "start_char_pos": 842, "end_char_pos": 896, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "in Iran said that he believed the Bid Kaneh explosion", "after": "\" spoke to The Guardian anonymously and claimed that the blast", "start_char_pos": 949, "end_char_pos": 1002, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}], "sents_char_pos": [0, 112, 261, 413, 557, 594, 841], "domain": "wiki"}
{"doc_id": "33879797", "revision_depth": 1, "before_revision": "Resolution In the wake of attacks against the United Nations Interim Force in Lebanon (UNIFIL), the Security Council today extended its mandate for two years , until 31 August 2013, condemning all terrorist attacks against it in the strongest terms. Unanimously adopting resolution 2004 (2011), the 15-member body strongly called upon all parties concerned to respect the cessation of hostilities, prevent any violation of the Blue Line and cooperate fully with UNIFIL. It called for the rapid finalization of Lebanons investigation into the 27 May and 26 July attacks on the Force so as to bring the perpetrators to justice, and urged all parties to abide scrupulously by their obligation to respect the safety of UNIFIL and other United Nations personnel.", "after_revision": "Resolution In the wake of attacks against the United Nations Interim Force in Lebanon (UNIFIL), the Security Council extended its mandate for one year , until 31 August 2012, and condemned all terrorist attacks against it in the strongest terms. The resolution was unanimously adopted, and the Security Council urged all parties concerned to respect the cessation of hostilities, prevent any violation of the Blue Line and cooperate fully with UNIFIL. The resolution called for the rapid finalization of Lebanons investigation into the 27 May and 26 July attacks on the Force so as to bring the perpetrators to justice, and urged all parties to abide scrupulously by their obligation to respect the safety of UNIFIL and other United Nations personnel.", "edit_actions": [{"type": "D", "before": "today", "after": null, "start_char_pos": 117, "end_char_pos": 122, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "two years", "after": "one year", "start_char_pos": 148, "end_char_pos": 157, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "2013, condemning", "after": "2012, and condemned", "start_char_pos": 176, "end_char_pos": 192, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Unanimously adopting resolution 2004 (2011), the 15-member body strongly called upon", "after": "The resolution was unanimously adopted, and the Security Council urged", "start_char_pos": 250, "end_char_pos": 334, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "It", "after": "The resolution", "start_char_pos": 470, "end_char_pos": 472, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 249, 469], "domain": "wiki"}
{"doc_id": "34430743", "revision_depth": 1, "before_revision": "Fighting and outcome During the Civil War, Abdul Hakkul Mubin fled to Kinarut (in present-day Papar, Sabah, Malaysia) where, he stayed there for ten years, repelling repeated attacks by Sultan Muhyiddin. They returned to Brunei after a final attack by Muhyiddin's forces in which they failed to defeat Abdul Hakkul Mubin. Muhyiddin was concerned that the civil war was dragging on too long and asked the help of the sultan of Sulu to send forces. He reportedly promised the land of eastern Sabah as a reward for the Sulu's assistance. Muhyiddin eventually emerged victorious. Abdul Hakkul Mubin was killed in the civil war. It is not clear to historians whether Muhyiddin asked for Sulus help in the civil war. The Sultan of Sulu at that time on his part, however, claimed that he was asked by Brunei to help and was promised eastern Sabah as a reward. As promised, the Sultan of Sulu received eastern Sabah as an honorary gift from the Sultan of Brunei, for the Tausugs'  help during the civil war (an event which traced the roots of the North Borneo dispute between Malaysia and the Philippines in the present-day).", "after_revision": "Fighting and outcome During the Civil War, Abdul Hakkul Mubin fled to Kinarut (in present-day Papar, Sabah, Malaysia) where, he stayed there for ten years, repelling repeated attacks by Sultan Muhyiddin. They returned to Brunei after a final attack by Muhyiddin's forces in which they failed to defeat Abdul Hakkul Mubin. Muhyiddin was concerned that the civil war was dragging on too long and asked the help of the sultan of Sulu to send forces. He reportedly promised the land of eastern Sabah as a reward for the Sulu's assistance. Muhyiddin eventually emerged victorious. Abdul Hakkul Mubin was killed in the civil war. It is not clear to historians whether Muhyiddin asked for Sulus help in the civil war. The Sultan of Sulu at that time on his part, however, claimed that he was asked by Brunei to help and was promised eastern Sabah as a reward. Sultan of Sulu claimed eastern Sabah as a gift from the Sultan of Brunei, for the Tausugs' that never really help during the civil war (an event which traced the roots of the North Borneo dispute between Malaysia and the Philippines in the present-day).", "edit_actions": [{"type": "D", "before": "As promised, the", "after": null, "start_char_pos": 853, "end_char_pos": 869, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "received", "after": "claimed", "start_char_pos": 885, "end_char_pos": 893, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "an honorary", "after": "a", "start_char_pos": 911, "end_char_pos": 922, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "that never really", "start_char_pos": 972, "end_char_pos": 972, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 203, 321, 446, 534, 575, 623, 710, 852], "domain": "wiki"}
{"doc_id": "3541802", "revision_depth": 1, "before_revision": "A pilot study, pilot project, pilot test, or pilot experiment is a small scale preliminary study conducted in order to evaluate feasibility, duration, cost, adverse events, and improve upon the study design prior to performance of a full-scale research project. Implementation Pilot experiments are frequently carried out before large-scale quantitative research, in an attempt to avoid time and money being used on an inadequately designed project. A pilot study is usually carried out on members of the relevant population. A pilot study is often used to test the design of the full-scale experiment which then can be adjusted. It is a potentially valuable insight and, should anything be missing in the pilot study, it can be added to the full-scale (and more expensive) experiment to improve the chances of a clear outcome. Other applications In the health researchcontext , studies conducted in preparation for a future randomized controlled trial are known as pilot and feasibility studies, where pilot studies are a subset of feasibility studies. A feasibility study asks whether something can be done, should we proceedwith it , and if so, how. A pilot study asks the same questions, but also has a specific design feature: in a pilot study, a future study or part of a future study is conducted on a smaller scale . Sandra M Eldridge, Gillian A Lancaster, Michael J Campbell, Lehana Thabane, Sally Hopewell, Claire L Coleman, Christine M Bond. Defining Feasibility and Pilot Studies in Preparation for Randomised Controlled Trials: Development of a Conceptual Framework. PLoS ONE. 2016; 11(3) ", "after_revision": "A pilot study, pilot project, pilot test, or pilot experiment is a small-scale preliminary study conducted to evaluate feasibility, duration, cost, adverse events, and improve upon the study design prior to performance of a full-scale research project. Implementation Pilot experiments are frequently carried out before large-scale quantitative research, in an attempt to avoid time and money being used on an inadequately designed project. A pilot study is usually carried out on members of the relevant population. A pilot study is used to formulate the design of the full-scale experiment which then can be adjusted. The pilot study is potentially a critical insight to clinical trial design, recruitment and sample size of participants, treatment testing, and statistical analysis to improve the power of testing the hypothesis of the study. Analysis from the pilot experiment can be added to the full-scale (and more expensive) experiment to improve the chances of a clear outcome. Applications In clinical research , studies conducted in preparation for a future randomized controlled trial are known as \"pilot\" and \"feasibility\" studies, where pilot studies are a subset of feasibility studies. A feasibility study asks whether the study should proceed , and if so, how. A pilot study asks the same questions, but also has a specific design feature: in a pilot study, a future study is conducted on a smaller scale , Sandra M Eldridge, Gillian A Lancaster, Michael J Campbell, Lehana Thabane, Sally Hopewell, Claire L Coleman, Christine M Bond. Defining Feasibility and Pilot Studies in Preparation for Randomised Controlled Trials: Development of a Conceptual Framework. PLoS ONE. 2016; 11(3) which, if having produced positive results, may lead to a Phase I clinical trial.", "edit_actions": [{"type": "R", "before": "small scale", "after": "small-scale", "start_char_pos": 67, "end_char_pos": 78, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "in order", "after": null, "start_char_pos": 107, "end_char_pos": 115, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "often used to test", "after": "used to formulate", "start_char_pos": 543, "end_char_pos": 561, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "It is a potentially valuable insight and, should anything be missing in the pilot study, it", "after": "The pilot study is potentially a critical insight to clinical trial design, recruitment and sample size of participants, treatment testing, and statistical analysis to improve the power of testing the hypothesis of the study. Analysis from the pilot experiment", "start_char_pos": 630, "end_char_pos": 721, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "R", "before": "Other applications", "after": "Applications", "start_char_pos": 828, "end_char_pos": 846, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "R", "before": "the health researchcontext", "after": "clinical research", "start_char_pos": 850, "end_char_pos": 876, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "pilot and feasibility", "after": "\"pilot\" and \"feasibility\"", "start_char_pos": 966, "end_char_pos": 987, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "something can be done, should we proceedwith it", "after": "the study should proceed", "start_char_pos": 1087, "end_char_pos": 1134, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "or part of a future study", "after": null, "start_char_pos": 1265, "end_char_pos": 1290, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ".", "after": ",", "start_char_pos": 1323, "end_char_pos": 1324, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "which, if having produced positive results, may lead to a Phase I clinical trial.", "start_char_pos": 1602, "end_char_pos": 1602, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 261, 449, 525, 629, 827, 1053, 1152, 1452, 1579], "domain": "wiki"}
{"doc_id": "38306871", "revision_depth": 1, "before_revision": "He got married in 1982 and had three sons (Tahir, Abdullah Saqib and Muaz) and four daughters. He was a Multilingual (able to speak many languages); in addition to his native language of Hindko language and the Arabiclanguage from his graduate study, He was also fluent in English, Urdu, Pashto and Greek, and could read and understand Persian.Who is Zubair Alizai, his family, his sons - URL Death Hafiz Zubair Alizai died on the 10th of November 2013 after short admission in Benazir Bhutto Hospital  Rawalpindi, Pakistan , and the cause of his death was lung failure.", "after_revision": "He got married in 1982 and had three sons (Tahir, Abdullah Saqib and Muaz) and four daughters. In addition to his native language of Hindko and Arabic, he was also fluent in English, Urdu, Pashto and Greek, and could read and understand Persian.Who is Zubair Alizai, his family, his sons - URL Death Hafiz Zubair Alizai died on the 10th of November 2013 in Benazir Bhutto Hospital in Rawalpindi, Pakistan . The cause of his death was lung failure.", "edit_actions": [{"type": "R", "before": "He was a Multilingual (able to speak many languages); in", "after": "In", "start_char_pos": 95, "end_char_pos": 151, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "language and the Arabiclanguage from his graduate study, He", "after": "and Arabic, he", "start_char_pos": 194, "end_char_pos": 253, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "after short admission", "after": null, "start_char_pos": 453, "end_char_pos": 474, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "in", "start_char_pos": 502, "end_char_pos": 502, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": ", and the", "after": ". The", "start_char_pos": 524, "end_char_pos": 533, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}], "sents_char_pos": [0, 94, 148, 344], "domain": "wiki"}
{"doc_id": "38569161", "revision_depth": 1, "before_revision": "170px|Textile doll (cotton, wool, wood) The Chancay were a pre-Columbian archeological civilization which developed between the valleys of Fortaleza, Pativilca, Supe, Huaura, Chancay, Chilln, Rimac and Lurn, on the central coast of Peru, from about CE 1000 to 1470. History 200px|Llama effigy (earthenware, slip paint) Not much is known about the Chancay civilization, which developed in the later part of the Inca Empire. This culture emerged after the fall of the Wari civilization. Parts of the southern Chancay area were conquered by the Chim in the early fifteenth century and in about 1450 A.D. the Incas were occupying both areas. It is believed that the Chancay had a centralized political structure , forming a small regional state. Thus the Chancay culture declined in the fifteenth centuryto make way for the territorial expansion of the Inca Empire . The settlements in Lauri, Lumbra, Tambo Blanco, Handrail, Pisquillo Chico and Tronconal focused mainly on artisans producing large-scale ceramics and textiles. The Chancay culture is the first of the Peruvian cultures that had mass production of ceramics, textiles and metals such as gold and silver which were ritualistic and domestic goods. They were also noted for their wood carved items. In type of fabric used include llama wool, cotton, chiffon  and feathers. Their technique involved were decorated open weave, brocade, embroidery, and painting. Brushes were used to paint anthropomorphic, zoomorphic, geometric and other creative designs directly on the canvases. The Chancay are known for the quality of their painted tapestries. The typically geometric designs also included drawings of plants, animals such as fish, cats, birds, monkeys and dogs (most notably the hairless Peruvian dog) as well as human figures. Some of the human fiber sculptures are elaborate and include such scenes as a mother teaching her daughter to weave on a backstrap loom. Birds and deities wearing crescent-like headdresses were one of the more common decorative features. They produced a variety of goods such as clothing, bags, and funeral masks.", "after_revision": "170px|Textile doll (cotton, wool, wood) The Chancay were a pre-Hispanic archeological civilization that developed between the valleys of Fortaleza, Pativilca, Supe, Huaura, Chancay, Chilln, Rimac and Lurn, on the central coast of Peru, from about 1000 to 1470 CE. History 200px|Llama effigy (earthenware, slip paint) Not much is known about the Chancay civilization, which developed in the later part of the Inca Empire. This culture emerged after the fall of the Wari civilization. Parts of the southern Chancay area were conquered by the Chim in the early 1400s, and by about 1450 CE the Incas were occupying both areas. The Chancay likely had a centralized political structure and formed a small regional state. Thus the Chancay culture declined in the later 15th century, as the Inca Empire expanded into their lands . The settlements in Lauri, Lumbra, Tambo Blanco, Handrail, Pisquillo Chico and Tronconal focused mainly on artisans producing large-scale ceramics and textiles. The Chancay culture is the first of the Peruvian cultures that had mass production of ceramics, textiles and metals such as gold and silver which were ritualistic and domestic goods. They were also noted for their wood carved items. In type of fabric used include llama wool, cotton, chiffon , and feathers. Their technique involved were decorated open weave, brocade, embroidery, and painting. Brushes were used to paint anthropomorphic, zoomorphic, geometric and other creative designs directly on the canvases. The Chancay are known for the quality of their painted tapestries. The typically geometric designs also included drawings of plants, animals such as fish, cats, birds, monkeys and dogs (most notably the hairless Peruvian dog) as well as human figures. Some of the human fiber sculptures are elaborate and include such scenes as a mother teaching her daughter to weave on a backstrap loom. Birds and deities wearing crescent-like headdresses were one of the more common decorative features. They produced a variety of goods such as clothing, bags, and funeral masks.", "edit_actions": [{"type": "R", "before": "pre-Columbian archeological civilization which", "after": "pre-Hispanic archeological civilization that", "start_char_pos": 59, "end_char_pos": 105, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}, {"type": "D", "before": "CE", "after": null, "start_char_pos": 249, "end_char_pos": 251, "major_intent": "style", "raw_intents": ["style", "clarity", "fluency"]}, {"type": "R", "before": "1470.", "after": "1470 CE.", "start_char_pos": 260, "end_char_pos": 265, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "fifteenth century and in", "after": "1400s, and by", "start_char_pos": 560, "end_char_pos": 584, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "A.D.", "after": "CE", "start_char_pos": 596, "end_char_pos": 600, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "It is believed that the Chancay", "after": "The Chancay likely", "start_char_pos": 638, "end_char_pos": 669, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "R", "before": ", forming", "after": "and formed", "start_char_pos": 708, "end_char_pos": 717, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "fifteenth centuryto make way for the territorial expansion of the Inca Empire", "after": "later 15th century, as the Inca Empire expanded into their lands", "start_char_pos": 783, "end_char_pos": 860, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1315, "end_char_pos": 1315, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 265, 422, 484, 637, 741, 862, 1022, 1205, 1255, 1329, 1416, 1535, 1602, 1787, 1924, 2025], "domain": "wiki"}
{"doc_id": "39091628", "revision_depth": 1, "before_revision": "Although debatable,  many agricultural historians also believe that the introduction of convertible husbandry was brought about with the introduction of the turnip. They argue that \"the lowly turnip made possible a change in crop rotation which did not require much capital, but which brought about a tremendous rise in agricultural productivity.\"Timmer, C. Peter, \"The Turnip, New Husbandry, and The English Agricultural Revolution\", The Quarterly Journal of Economics, Vol. 83, No. 3 (Aug., 1969), pp. 375-395 They believe that this \"fodder\" crop pushed agriculture in a direction in which \"alternating\" husbandry was seen as more efficient than traditional permanent pasture farming and jump-started the improvement of crop rotation and agricultural output versus capital. Although the turnip was popularized by Lord Townshend during the mid-18th century, the use of turnips being grown as fodder was seen as early as the 16th century. Convertible husbandry has been praised as the \"best way to keep high fertility on both arable and pasture and to retain excellent soil texture and composition.\"Slicher van Bath, B.H., \"Agrarian History of Western Europe\", 1963, pp 249-54. It is adaptable to the type of land or soil at the location, and it provided a rotation between pasture and arable land which \"not only produced the same amount of grain on a much reduced area, but broke the agrarian cycle of diminished returns by allowing more sheep and cattle to be kept, animals whose dung maintained\" fertility.", "after_revision": "Although often praised as increasing yields and efficiency, this is debatable. Yields were often poor in the late Middle Ages and approximately equivalent to regular common fields. It was not always adaptable to the type of land or soil at all locations. Kitsikopoulos claims it was the introduction of fodder legumes, such as clover, in the early modern period which eventually truly achieved greater agricultural productivity. Although debatable, many agricultural historians also believe that the introduction of convertible husbandry was brought about with the introduction of the turnip. They argue that \"the lowly turnip made possible a change in crop rotation which did not require much capital, but which brought about a tremendous rise in agricultural productivity.\"Timmer, C. Peter, \"The Turnip, New Husbandry, and The English Agricultural Revolution\", The Quarterly Journal of Economics, Vol. 83, No. 3 (Aug., 1969), pp. 375-395 They believe that this \"fodder\" crop pushed agriculture in a direction in which \"alternating\" husbandry was seen as more efficient than traditional permanent pasture farming and jump-started the improvement of crop rotation and agricultural output versus capital. Although the turnip was popularized by Lord Townshend during the mid-18th century, the use of turnips being grown as fodder was seen as early as the 16th century. Convertible husbandry has been praised as the \"best way to keep high fertility on both arable and pasture and to retain excellent soil texture and composition.\"Slicher van Bath, B.H., \"Agrarian History of Western Europe\", 1963, pp 249-54. According to one author, the rotation it provided between pasture and arable land \"not only produced the same amount of grain on a much reduced area, but broke the agrarian cycle of diminished returns by allowing more sheep and cattle to be kept, animals whose dung maintained\" fertility.", "edit_actions": [{"type": "R", "before": "debatable,", "after": "often praised as increasing yields and efficiency, this is debatable. Yields were often poor in the late Middle Ages and approximately equivalent to regular common fields. It was not always adaptable to the type of land or soil at all locations. Kitsikopoulos claims it was the introduction of fodder legumes, such as clover, in the early modern period which eventually truly achieved greater agricultural productivity.", "start_char_pos": 9, "end_char_pos": 19, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Although debatable,", "start_char_pos": 20, "end_char_pos": 20, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "It is adaptable to the type of land or soil at the location, and it provided a rotation", "after": "According to one author, the rotation it provided", "start_char_pos": 1178, "end_char_pos": 1265, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "which", "after": null, "start_char_pos": 1298, "end_char_pos": 1303, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}], "sents_char_pos": [0, 164, 346, 775, 938, 1098, 1177], "domain": "wiki"}
{"doc_id": "39942", "revision_depth": 1, "before_revision": "19881993: Beginnings Aniston first worked in Off-Broadway productions such as For Dear Life and Dancing on Checker's Grave, and supported herself with part-time jobs which included working as a telemarketer, waitress , and bike messenger. In 1988 , she had an uncredited minor role in the critically panned sci-fi adventure film Mac and Me. The next year she appeared on The Howard Stern Show as a spokesmodel for Nutrisystem, and moved back to LosAngeles. Aniston also appeared in two more failed television comedy series , The Edge and Muddling Through, and guest-starred in Quantum Leap, Herman's Head , and Burke's Law. 19942004: Friends and worldwide recognition Depressed over her four unsuccessful television shows, Aniston approached Warren Littlefield at a Los Angeles gas station asking for reassurance. The head of NBC entertainment encouraged her to continue acting, and a few months later helped cast her in Friends, a sitcom that was set to debut on NBC's 19941995 fall lineup. The producer originally wanted Aniston to audition for the role of Monica Geller, but Courteney Cox was deemed more suitable . Thus, Aniston was cast as Rachel Green. She was also offered a spot as a featured player on Saturday Night Live, but turned it down in favor of Friends. She played Rachel until the show ended in 2004. upAniston at the 2008 Toronto International Film Festival The program was a massive hit and Aniston, along with her co-stars, gained worldwide recognition among television viewers . Her character was widely popularand is now regarded as one of the greatest female characters in American television history. Aniston received five Primetime Emmy Award nominations (two for Supporting Actress, three for Lead Actress), including a win for Lead Actress. She was also nominated for two Golden Globe Awards and won in 2003 as Best Actress Television Series Musical or Comedy. According to the Guinness Book of World Records, Aniston (along with her female co-stars) became the highest-paid television actress of all time with her $1million-per-episode paycheck for the final season of Friends. Her character's relationship with Ross Geller, portrayed by David Schwimmer in the show, was widely popular among audiences, and the couple was frequently voted as television's favorite couple in polls and magazines. Following a four-year hiatus, Aniston returned to film work in 1996, when she performed in the ensemble cast of romantic comedy She's the One. Aniston's first starring vehicle was Picture Perfect (1997), in which she played a struggling young advertising executive opposite Kevin Bacon and Jay Mohr. While the film received mixed reviews, it was a moderate commercial success, and Aniston's performance was more warmly received, with many critics suggesting that she had screen presence. In 1998, she appeared as a woman who falls for a gay man (played by Paul Rudd) in the romantic comedy The Object of My Affection, and the next year she starred as a restaurant waitress in the cult film Office Space. Aniston appeared in the dramedy Rock Star (2001) opposite Mark Wahlberg and Dominic West. She starred in the independent dramedy The Good Girl (2002) as an unglamorous cashier who cheats on her husband. The film was a commercial success in limited release, taking in over $ 14million in North America. Film critic Roger Ebert declared it her breakthrough: Aniston's biggest commercial success in film has been the comedy Bruce Almighty (2003), in which she played the girlfriend of a television field reporter (Jim Carrey) offered the chance to be God for one week. With a worldwide box office gross of $484million, the film was the fifth-highest-grossing big screen production of the year. Aniston next starred as the old classmate of a tightly wound newly-wed in the romantic comedy Along Came Polly (2004), opposite Ben Stiller, which placed number one at the North American box office, earning $ 27.7million in its opening weekend; it eventually made $172million worldwide.", "after_revision": "19881993: Beginnings Aniston first worked in Off-Broadway productions such as For Dear Life and Dancing on Checker's Grave, and supported herself with part-time jobs including work as a telemarketer, waitress and bike messenger. In 1988 she had an uncredited minor role in the critically panned sci-fi adventure film Mac and Me. The next year she appeared on The Howard Stern Show as a spokesmodel for Nutrisystem, and moved back to LosAngeles. Aniston also appeared in the two failed television comedy series The Edge and Muddling Through, and guest-starred in Quantum Leap, Herman's Head and Burke's Law. 19942004: Friends and worldwide recognition Depressed over her four unsuccessful television shows, Aniston approached Warren Littlefield at a Los Angeles gas station asking for reassurance. The head of NBC entertainment encouraged her to continue acting, and a few months later helped cast her in Friends, a sitcom set to debut on NBC's 19941995 fall lineup. The producer wanted Aniston to audition for the role of Monica Geller, but Courteney Cox was deemed more suitable , and Aniston was cast as Rachel Green. She was also offered a spot as a featured player on Saturday Night Live, but turned it down in favor of Friends. She played Rachel until the show ended in 2004. upAniston at the 2008 Toronto International Film Festival The program was a massive hit and Aniston, along with her co-stars, gained worldwide recognition . Her character was especially popular. She received five Primetime Emmy Award nominations (two for Supporting Actress, three for Lead Actress), and won for Lead Actress. She was also nominated for two Golden Globe Awards and won in 2003 as Best Actress Television Series Musical or Comedy. According to the Guinness Book of World Records, Aniston (along with her female co-stars) became the highest-paid television actress of all time with her $1million-per-episode paycheck during the final season of Friends. Her character's relationship with Ross Geller, portrayed by David Schwimmer, was widely popular among audiences; they were frequently voted television's favorite couple in polls and magazines. After a four-year hiatus, Aniston returned to film work in 1996, when she performed in the ensemble cast of romantic comedy She's the One. Her first starring film vehicle was Picture Perfect (1997), where she played a struggling young advertising executive opposite Kevin Bacon and Jay Mohr. It received mixed reviews and was only a moderate commercial success; but Aniston's performance was more warmly received, with many critics suggesting that she had screen presence. In 1998, she appeared as a woman who falls for a gay man (played by Paul Rudd) in the romantic comedy The Object of My Affection, and the next year she starred as a restaurant waitress in the cult film Office Space. Aniston appeared in the dramedy Rock Star (2001) opposite Mark Wahlberg and Dominic West. She starred in the independent dramedy The Good Girl (2002) as an unglamorous cashier who cheats on her husband. The film was a commercial success in limited release, taking in over $ 14million in North America. Film critic Roger Ebert declared it her breakthrough: Aniston's biggest commercial success in film has been the comedy Bruce Almighty (2003), where she played the girlfriend of a television field reporter (Jim Carrey) offered the chance to be God for one week. With a worldwide box office gross of $484million, it was the fifth-highest-grossing feature film of the year. Aniston next starred as the old classmate of a tightly wound newlywed in the romantic comedy Along Came Polly (2004) opposite Ben Stiller, which placed number one at the North American box office, earning $ 27.7million in its opening weekend; it eventually made $172million worldwide.", "edit_actions": [{"type": "R", "before": "which included working", "after": "including work", "start_char_pos": 166, "end_char_pos": 188, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 217, "end_char_pos": 218, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 247, "end_char_pos": 248, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "two more", "after": "the two", "start_char_pos": 482, "end_char_pos": 490, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 523, "end_char_pos": 524, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 605, "end_char_pos": 606, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "that was", "after": null, "start_char_pos": 939, "end_char_pos": 947, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "originally", "after": null, "start_char_pos": 1005, "end_char_pos": 1015, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": ". Thus,", "after": ", and", "start_char_pos": 1117, "end_char_pos": 1124, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}, {"type": "D", "before": "among television viewers", "after": null, "start_char_pos": 1475, "end_char_pos": 1499, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "widely popularand is now regarded as one of the greatest female characters in American television history. Aniston", "after": "especially popular. She", "start_char_pos": 1520, "end_char_pos": 1634, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "including a win", "after": "and won", "start_char_pos": 1736, "end_char_pos": 1751, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "$1million-per-episode paycheck for the final season of Friends. Her character's relationship with Ross Geller, portrayed by David Schwimmer in the show, was widely popular among audiences, and the couple was frequently voted as television's favorite couple in polls and magazines. Following a four-year hiatus, Aniston returned to film work in 1996, when she performed in the ensemble cast of romantic comedy She's the One. Aniston's first starring vehicle was Picture Perfect (1997), in which she played a struggling young advertising executive opposite Kevin Bacon and Jay Mohr. While the film received mixed reviews, it was a moderate commercial success, and Aniston's performance was more warmly received, with many critics suggesting that she had screen presence. In 1998, she appeared as a woman who falls for a gay man (played by Paul Rudd) in the romantic comedy The Object of My Affection, and the next year she starred as a restaurant waitress in the cult film Office Space. Aniston appeared in the dramedy Rock Star (2001) opposite Mark Wahlberg and Dominic West. She starred in the independent dramedy The Good Girl (2002) as an unglamorous cashier who cheats on her husband. The film was a commercial success in limited release, taking in over $", "after": "$1million-per-episode paycheck during the final season of Friends. Her character's relationship with Ross Geller, portrayed by David Schwimmer, was widely popular among audiences; they were frequently voted television's favorite couple in polls and magazines. After a four-year hiatus, Aniston returned to film work in 1996, when she performed in the ensemble cast of romantic comedy She's the One. Her first starring film vehicle was Picture Perfect (1997), where she played a struggling young advertising executive opposite Kevin Bacon and Jay Mohr. It received mixed reviews and was only a moderate commercial success; but Aniston's performance was more warmly received, with many critics suggesting that she had screen presence. In 1998, she appeared as a woman who falls for a gay man (played by Paul Rudd) in the romantic comedy The Object of My Affection, and the next year she starred as a restaurant waitress in the cult film Office Space. Aniston appeared in the dramedy Rock Star (2001) opposite Mark Wahlberg and Dominic West. She starred in the independent dramedy The Good Girl (2002) as an unglamorous cashier who cheats on her husband. The film was a commercial success in limited release, taking in over $", "start_char_pos": 2044, "end_char_pos": 3302, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "in which", "after": "where", "start_char_pos": 3473, "end_char_pos": 3481, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "$484million, the film was the fifth-highest-grossing big screen production of the year. Aniston next starred as the old classmate of a tightly wound newly-wed in the romantic comedy Along Came Polly (2004), opposite Ben Stiller, which placed number one at the North American box office, earning $", "after": "$484million, it was the fifth-highest-grossing feature film of the year. Aniston next starred as the old classmate of a tightly wound newlywed in the romantic comedy Along Came Polly (2004) opposite Ben Stiller, which placed number one at the North American box office, earning $", "start_char_pos": 3632, "end_char_pos": 3928, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 238, 340, 456, 813, 991, 1118, 1158, 1271, 1501, 1626, 1769, 1889, 2107, 2324, 2467, 2624, 2812, 3028, 3118, 3231, 3330, 3594, 3719, 3964], "domain": "wiki"}
{"doc_id": "4059354", "revision_depth": 1, "before_revision": "Jane Toppan (March 31, 1854 August 17, 1938), born Honora Kelley, was an American serial killer, nicknamed \"Jolly Jane\". After her arrest in 1901, she confessed to 31 murders, but only 12 were confirmed. Toppan is quoted as saying that her ambition was \"to have killed more peoplehelpless peoplethan any other man or woman who ever lived\". Toppan worked as a nurse from c. 1885 to 1902. Her victims consisted of her patients and their family members. In 1860, only a few years after his wife's death, Kelley took his two youngest children, eight-year-old Delia Josephine and six-year-old Honora, to the Boston Female Asylum, an orphanage for indigent female children. Kelley surrendered the two girls, never to see them again. Documents from the asylum note that they were \"rescued from a very miserable home\". No records exist of Delia and Honora's experiences during their time in the asylum, but reportedly, Delia became a prostitute while their older sister Nellie (who was not committed to the orphanage) was committed to an insane asylum. In November 1862, less than two years after her father had left them , Honora Kelley was placed as an indentured servant in the home of Mrs. Ann C. Toppan of Lowell, Massachusetts. Though never formally adopted by the Toppans, Honora took on the surname of her benefactors and eventually became known as Toppan. The original Toppan family already had a daughter, Elizabeth, with whom Honora was on good terms. The surviving members of the Davis family ordered a toxicology exam on Alden Davis' youngest daughter, Minnie. The report found that she had been poisoned, and local authorities assigned a police detail on Toppan to watch her. On October 29, 1901, she was arrested for murder. By 1902, she had confessed to 31 murders. Soon after the trial, one of William Randolph Hearst's newspapers, the New York Journal, printed what was purported to be Toppan's confession to her lawyer, claiming that she had killed more than 31 people, and that she wanted the jury to find her sane so she could eventually have a chance at being released. Toppan insisted upon her own sanity in court, claiming that she could not be insane if she knew what she was doing and knew that it was wrong, but nonetheless she was declared insane and committed. On June 23, in the Barnstable County Courthouse, she was found not guilty by reason of insanity and committed for life in the Taunton Insane Hospital.", "after_revision": "Jane Toppan (March 31, 1854 August 17, 1938), born Honora Kelley, was an American serial killer, nicknamed \"Jolly Jane\". After her arrest in 1901, she confessed to thirty-one murders, but only twelve were confirmed. Toppan is quoted as saying that her ambition was \"to have killed more peoplehelpless peoplethan any other man or woman who ever lived\". Toppan worked as a nurse from around 1885 to 1902. Her victims consisted of her patients and their family members. In 1860, only a few years after his wife's death, Kelley took his two youngest children, eight-year-old Delia Josephine and six-year-old Honora, to the Boston Female Asylum, an orphanage for indigent female children. Kelley surrendered the two girls, never to see them again. Documents from the asylum note that they were \"rescued from a very miserable home\". No records exist of Delia and Honora's experiences during their time in the asylum, but reportedly, Delia became a prostitute while their older sister Nellie (who was not committed to the orphanage) was committed to an insane asylum. In November 1862, less than two years after being abandoned by her father , Honora Kelley was placed as an indentured servant in the home of Mrs. Ann C. Toppan of Lowell, Massachusetts. Though never formally adopted by the Toppans, Honora took on the surname of her benefactors and eventually became known as Toppan. The original Toppan family already had a daughter, Elizabeth, with whom Honora was on good terms. The surviving members of the Davis family ordered a toxicology exam on Alden Davis' youngest daughter, Minnie. The report found that she had been poisoned, and local authorities assigned a police detail on Toppan to watch her. On October 29, 1901, she was arrested for murder. By 1902, she had confessed to thirty-one murders. Soon after the trial, one of William Randolph Hearst's newspapers, the New York Journal, printed what was purported to be Toppan's confession to her lawyer, claiming that she had killed more than thirty-one people, and that she wanted the jury to find her sane so she could eventually have a chance at being released. Toppan insisted upon her own sanity in court, claiming that she could not be insane if she knew what she was doing and knew that it was wrong, but nonetheless she was declared insane and committed. On June 23, in the Barnstable County Courthouse, she was found not guilty by reason of insanity and committed for life in the Taunton Insane Hospital.", "edit_actions": [{"type": "R", "before": "31", "after": "thirty-one", "start_char_pos": 164, "end_char_pos": 166, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "12", "after": "twelve", "start_char_pos": 185, "end_char_pos": 187, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "c.", "after": "around", "start_char_pos": 370, "end_char_pos": 372, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "her father had left them", "after": "being abandoned by her father", "start_char_pos": 1089, "end_char_pos": 1113, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "31", "after": "thirty-one", "start_char_pos": 1762, "end_char_pos": 1764, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "31", "after": "thirty-one", "start_char_pos": 1970, "end_char_pos": 1972, "major_intent": "fluency", "raw_intents": ["fluency", "others", "style"]}], "sents_char_pos": [0, 120, 203, 339, 386, 450, 667, 726, 810, 1044, 1185, 1225, 1356, 1454, 1565, 1681, 1731, 1773, 2083, 2281], "domain": "wiki"}
{"doc_id": "40918428", "revision_depth": 1, "before_revision": "The Brussels Agreement ( / Briselski sporazum, ), formally the First Agreement of Principles Governing the Normalisation of Relations, was a multilateral treaty proposed to normalize relations between the nations of Serbia and KosovoAs to the agreement not having been signed, see text of agreement. . The agreement was negotiated and concluded in Brussels under the auspices of the European Union and signed on 19 April 2013. Negotiations were led by Serbian Prime Minister Ivica Dai and Kosovo Prime Minister Hashim Thai and mediated by EU High Representative Catherine Ashton. The government of Serbia does not recognize Kosovo as a sovereign state, but began normalizing relations with the government of Kosovo nonetheless, in accordance with the Brussels URL Background Following the Kosovo War and NATO bombing of Yugoslavia in 1999, Kosovo, as part of Serbia and Federal Republic of Yugoslavia, was placed under United Nations administration under UNSC Resolution 1244. In 2008, Kosovo declared independence and has since been recognized by 96 countries. However, Serbia maintains that Kosovo continues to be its territory. Kosovo has become a member of several URLanizations, while according to Serbia and about half of the UN member states, Resolution 1244 remains legally binding on all parties. The EU-mediated dialogue between Kosovo and Serbia was launched in March 2011, three years after Kosovo declared URL Belgrade and Pristina were urged to continue talks in Brussels, but Serbia was not obliged to recognize Kosovo at any point in the URL Talks Until preset, ten rounds of talks have been held at the European External Action Service office in Brussels. Baroness Ashton, the EU High Representative, had been chairing talks for two years, followed by Federica Mogherini. Normalization of relations with neighboring states is a key precondition for states wishing to join the EU; the Brussels Agreement brought Serbia close to EU accession talks and Kosovo to initialing a Stabilisation and Association Agreement (SAA). The SAA was signed by HR Mogherini and PM Mustafa in October 2015. Removing the parallel Serbian structures in the Northern part of Kosovo remains challenging. 330px|Ivica Dai, Prime Minister of Serbia (left), Catherine Ashton (middle), and Hashim Thai, Prime Minister of Kosovo (right) in Munich, 2014. United States diplomats have supported the EU-led dialogue from the beginning. U.S. Secretary of State Hillary Clinton invited Baroness Ashton to undertake joint travel in the Balkans, and the two made joint visits to Belgrade and Pristina (as well as Sarajevo) in October 2012. Clinton assigned Deputy Assistant Secretary of State ambassador Philip Reeker to support Ashton's efforts. Reeker and his team coordinated closely with EU colleagues and met with Serbian and Kosovar representatives on the margins of every dialogue session in Brussels, and during visits to the region. The result was the 15-point \"First Agreement of Principles Governing the Normalisation of Relations\", initialed but not signed on 19 April 2013. After the agreement was concluded, the European Commission officially advised that work start on an SAA with Kosovo, and accession negotiations began with Serbia. The agreement was supported by the European Union, NATO, the OSCE, and the United Nations. In October 2013, arrangements were made for Serbian officials to visit North Kosovo. It was agreed that electoral bodies in Kosovo would include Kosovo Serb representatives and that international dialing code +383 would be attributed to Kosovo. The new code was expected to go into effect in late 2015 but has again been delayed by Serbia until March 2016. The geographic phone code was implemented in December 2016. The National Assembly of Serbia has not treated the agreement as international and has not ratified it into a binding law, which is a procedure required for international agreements in Serbia.Zakon o zakljucivanju i izvrsenju medjunarodnih ugovora It has, however, merely accepted the government report about the \"hitherto process of political and technical dialog with the temporary institutions in Pristina with the mediation of EU, including the process of implementation of the achieved agreements\".O PRIHVATANjU IZVETAJA O DOSADANjEM PROCESS POLITIKOG I TEHNIKOG DIJALOGA SA PRIVREMENIM INSTITUCIJAMA SAMOUPRAVE U PRITINI UZ POSREDOVANjE EVROPSKE UNIJE, UKLjUUJUI PROCES IMPLEMENTACIJE POSTIGNUTIH DOGOVORA The constitutional court in Belgrade did not answer the question about constitutionality of the agreement, stating in December 2014 that the issue raised is a political question, and not a legal one. Serbia has accepted, however, that progress in implementing the Brussels agreement is a prerequisite for EU accession talks within the opening of Chapter 35 negotiations in December 2015.", "after_revision": "The Brussels Agreement ( / Briselski sporazum, ), formally the First Agreement of Principles Governing the Normalisation of Relations, was a multilateral treaty proposed to normalize relations between the governments of Serbia and Kosovo. The agreement was negotiated and concluded in Brussels under the auspices of the European Union and signed on 19 April 2013. Negotiations were led by Serbian Prime Minister Ivica Dai and Kosovo Prime Minister Hashim Thai and mediated by EU High Representative Catherine Ashton. The government of Serbia does not recognize Kosovo as a sovereign state, but began normalizing relations with the government of Kosovo nonetheless, in accordance with the Brussels URL Background Following the Kosovo War and NATO bombing of Yugoslavia in 1999, Kosovo, as part of Serbia and Federal Republic of Yugoslavia, was placed under United Nations administration under UNSC Resolution 1244. In 2008, Kosovo declared independence and has since been recognized by 96 countries. However, Serbia maintains that Kosovo continues to be its territory. The European Union-mediated dialogue between Kosovo and Serbia was launched in March 2011, three years after Kosovo declared URL Belgrade and Pristina were urged to continue talks in Brussels, but Serbia was not obliged to recognize Kosovo at any point in the URL Talks Until preset, ten rounds of talks have been held at the European External Action Service office in Brussels. Baroness Ashton, the EU High Representative, had been chairing talks for two years, followed by Federica Mogherini. Normalization of relations with neighboring states is a key precondition for states wishing to join the EU; the Brussels Agreement brought Serbia close to EU accession talks and Kosovo to initialing a Stabilisation and Association Agreement (SAA). The SAA was signed by HR Mogherini and PM Mustafa in October 2015. 330px|Ivica Dai, Prime Minister of Serbia (left), Catherine Ashton (middle), and Hashim Thai, Prime Minister of Kosovo (right) in Munich, 2014. United States diplomats have supported the EU-led dialogue from the beginning. U.S. Secretary of State Hillary Clinton invited Baroness Ashton to undertake joint travel in the Balkans, and the two made joint visits to Belgrade and Pristina (as well as Sarajevo) in October 2012. After the agreement was concluded, the European Commission officially advised that work start on an SAA with Kosovo, and accession negotiations began with Serbia. The agreement was supported by the European Union, NATO, the OSCE, and the United Nations. In October 2013, arrangements were made for Serbian officials to visit North Kosovo. It was agreed that electoral bodies in Kosovo would include Kosovo Serb representatives and that international dialing code +383 would be attributed to Kosovo. After some delays, the new geographic phone code was implemented in December 2016. The National Assembly of Serbia has not treated the agreement as international and has not ratified it into a binding law, which is a procedure required for international agreements in Serbia.Zakon o zakljucivanju i izvrsenju medjunarodnih ugovora It has, however, merely accepted the government report about the \"hitherto process of political and technical dialogue with the temporary institutions in Pristina with the mediation of EU, including the process of implementation of the achieved agreements\".O PRIHVATANjU IZVETAJA O DOSADANjEM PROCESS POLITIKOG I TEHNIKOG DIJALOGA SA PRIVREMENIM INSTITUCIJAMA SAMOUPRAVE U PRITINI UZ POSREDOVANjE EVROPSKE UNIJE, UKLjUUJUI PROCES IMPLEMENTACIJE POSTIGNUTIH DOGOVORA The constitutional court in Belgrade did not answer the question about constitutionality of the agreement, stating in December 2014 that the issue raised is a political question, and not a legal one. ", "edit_actions": [{"type": "R", "before": "nations", "after": "governments", "start_char_pos": 205, "end_char_pos": 212, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "KosovoAs to the agreement not having been signed, see text of agreement. .", "after": "Kosovo.", "start_char_pos": 227, "end_char_pos": 301, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Kosovo has become a member of several URLanizations, while according to Serbia and about half of the UN member states, Resolution 1244 remains legally binding on all parties. The EU-mediated", "after": "The European Union-mediated", "start_char_pos": 1131, "end_char_pos": 1321, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "D", "before": "Removing the parallel Serbian structures in the Northern part of Kosovo remains challenging.", "after": null, "start_char_pos": 2104, "end_char_pos": 2196, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "Clinton assigned Deputy Assistant Secretary of State ambassador Philip Reeker to support Ashton's efforts. Reeker and his team coordinated closely with EU colleagues and met with Serbian and Kosovar representatives on the margins of every dialogue session in Brussels, and during visits to the region.", "after": null, "start_char_pos": 2620, "end_char_pos": 2921, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "The result was the 15-point \"First Agreement of Principles Governing the Normalisation of Relations\", initialed but not signed on 19 April 2013.", "after": null, "start_char_pos": 2922, "end_char_pos": 3066, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "The new code was expected to go into effect in late 2015 but has again been delayed by Serbia until March 2016. The", "after": "After some delays, the new", "start_char_pos": 3566, "end_char_pos": 3681, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "dialog", "after": "dialogue", "start_char_pos": 4096, "end_char_pos": 4102, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "D", "before": "Serbia has accepted, however, that progress in implementing the Brussels agreement is a prerequisite for EU accession talks within the opening of Chapter 35 negotiations in December 2015.", "after": null, "start_char_pos": 4650, "end_char_pos": 4837, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 426, 579, 976, 1061, 1130, 1305, 1672, 1788, 1896, 2036, 2103, 2196, 2340, 2419, 2619, 2726, 2921, 3066, 3229, 3320, 3405, 3565, 3677, 3737, 3930, 4241, 4649], "domain": "wiki"}
{"doc_id": "40918428", "revision_depth": 2, "before_revision": "The National Assembly of Serbia has not treated the agreement as international and has not ratified it into a binding law, which is a procedure required for international agreements in Serbia .Zakon o zakljucivanju i izvrsenju medjunarodnih ugovora It has, however, merely accepted the government report about the \"hitherto process of political and technical dialogue with the temporary institutions in Pristina with the mediation of EU, including the process of implementation of the achieved agreements\".O PRIHVATANjU IZVE\u0160TAJA O DOSADA\u0160NjEM PROCESS POLITI\u010cKOG I TEHNI\u010cKOG DIJALOGA SA PRIVREMENIM INSTITUCIJAMA SAMOUPRAVE U PRI\u0160TINI UZ POSREDOVANjE EVROPSKE UNIJE, UKLjU\u010cUJU\u0106I PROCES IMPLEMENTACIJE POSTIGNUTIH DOGOVORA The constitutional court in Belgrade did not answer the question about constitutionality of the agreement, stating in December 2014 that the issue raised is a political question, and not a legal one.", "after_revision": "The National Assembly of Serbia has not treated the agreement as international and has not ratified it , which would be the procedure required in Serbia for approving an international agreement .Zakon o zakljucivanju i izvrsenju medjunarodnih ugovora It has, however, merely accepted the government report about the \"hitherto process of political and technical dialogue with the temporary institutions in Pristina with the mediation of EU, including the process of implementation of the achieved agreements\".O PRIHVATANjU IZVE\u0160TAJA O DOSADA\u0160NjEM PROCESS POLITI\u010cKOG I TEHNI\u010cKOG DIJALOGA SA PRIVREMENIM INSTITUCIJAMA SAMOUPRAVE U PRI\u0160TINI UZ POSREDOVANjE EVROPSKE UNIJE, UKLjU\u010cUJU\u0106I PROCES IMPLEMENTACIJE POSTIGNUTIH DOGOVORA The constitutional court in Belgrade did not answer a question about the constitutionality of the agreement, stating in December 2014 that the issue raised was a political question, and not a legal one.", "edit_actions": [{"type": "R", "before": "into a binding law, which is a procedure required for international agreements in Serbia", "after": ", which would be the procedure required in Serbia for approving an international agreement", "start_char_pos": 103, "end_char_pos": 191, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "the question about", "after": "a question about the", "start_char_pos": 774, "end_char_pos": 792, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "is", "after": "was", "start_char_pos": 876, "end_char_pos": 878, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}], "sents_char_pos": [0, 122, 506], "domain": "wiki"}
{"doc_id": "40918428", "revision_depth": 3, "before_revision": "The Brussels Agreement ( / Briselski sporazum, ), formally the First Agreement of Principles Governing the Normalisation of Relations, was a multilateral treaty proposed to normalize relations between the nations of Serbia and KosovoAs to the agreement not having been signed, see text of agreement. . The agreement was negotiated and concluded in Brussels under the auspices of the European Union and signed on 19 April 2013. Negotiations were led by Serbian Prime Minister Ivica Da\u010di\u0107 and Kosovo Prime Minister Hashim Tha\u00e7i and mediated by EU High Representative Catherine Ashton. The government of Serbia does not recognize Kosovo as a sovereign state, but began normalizing relations with the government of Kosovo nonetheless, in accordance with the Brussels URL In 2018, Serbia's president Aleksandar Vucic commented that the agreement is a difficult compromise for Serbia, which according to him fulfilled all of its obligations. Background Following the Kosovo War and NATO bombing of Yugoslavia in 1999, Kosovo, as part of Serbia and Federal Republic of Yugoslavia, was placed under United Nations administration under UNSC Resolution 1244. In 2008, Kosovo declared independence and has since been recognized by 98 countries. However, Serbia maintains that Kosovo continues to be its territory. Kosovo has become a member of several international organizations, while according to Serbia and about half of the UN member states, Resolution 1244 remains legally binding on all parties. The EU-mediated dialogue between Kosovo and Serbia was launched in March 2011, three years after Kosovo declared URL Belgrade and Pristina were urged to continue talks in Brussels, but Serbia was not obliged to recognize Kosovo at any point in the URL Talks Until preset, ten rounds of talks have been held at the European External Action Service office in Brussels. Baroness Ashton, the EU High Representative, had been chairing talks for two years, followed by Federica Mogherini. Normalization of relations with neighboring states is a key precondition for states wishing to join the EU; the Brussels Agreement brought Serbia close to EU accession talks and Kosovo to initialing a Stabilisation and Association Agreement (SAA). The SAA was signed by HR Mogherini and PM Mustafa in October 2015. Removing the parallel Serbian structures in the Northern part of Kosovo remains challenging. 330px|Ivica Da\u010di\u0107, Prime Minister of Serbia (left), Catherine Ashton (middle), and Hashim Tha\u00e7i, Prime Minister of Kosovo (right) in Munich, 2014. United States diplomats have supported the EU-led dialogue from the beginning. U.S. Secretary of State Hillary Clinton invited Baroness Ashton to undertake joint travel in the Balkans, and the two made joint visits to Belgrade and Pristina (as well as Sarajevo) in October 2012. Clinton assigned Deputy Assistant Secretary of State ambassador Philip Reeker to support Ashton's efforts. Reeker and his team coordinated closely with EU colleagues and met with Serbian and Kosovar representatives on the margins of every dialogue session in Brussels, and during visits to the region. The result was the 15-point \"First Agreement of Principles Governing the Normalisation of Relations\", initialed but not signed on 19 April 2013. After the agreement was concluded, the European Commission officially advised that work start on an SAA with Kosovo, and accession negotiations began with Serbia. The agreement was supported by the European Union, NATO, the OSCE, and the United Nations. In October 2013, arrangements were made for Serbian officials to visit North Kosovo. It was agreed that electoral bodies in Kosovo would include Kosovo Serb representatives and that international dialing code +383 would be attributed to Kosovo. The new code was expected to go into effect in late 2015 but has again been delayed by Serbia until March 2016. The geographic phone code was implemented in December 2016. The National Assembly of Serbia has not treated the agreement as international and has not ratified it into a binding law, which is a procedure required for international agreements in Serbia .Zakon o zakljucivanju i izvrsenju medjunarodnih ugovora It has, however, merely accepted the government report about the \"hitherto process of political and technical dialog with the temporary institutions in Pristina with the mediation of EU, including the process of implementation of the achieved agreements\".O PRIHVATANjU IZVE\u0160TAJA O DOSADA\u0160NjEM PROCESS POLITI\u010cKOG I TEHNI\u010cKOG DIJALOGA SA PRIVREMENIM INSTITUCIJAMA SAMOUPRAVE U PRI\u0160TINI UZ POSREDOVANjE EVROPSKE UNIJE, UKLjU\u010cUJU\u0106I PROCES IMPLEMENTACIJE POSTIGNUTIH DOGOVORA The constitutional court in Belgrade did not answer the question about constitutionality of the agreement, stating in December 2014 that the issue raised is a political question, and not a legal one. Serbia has accepted, however, that progress in implementing the Brussels agreement is a prerequisite for EU accession talks within the opening of Chapter 35 negotiations in December 2015.", "after_revision": "The Brussels Agreement ( / Briselski sporazum, ), formally the First Agreement of Principles Governing the Normalisation of Relations, was a multilateral treaty proposed to normalize relations between the governments of Serbia and Kosovo. The agreement was negotiated and concluded in Brussels under the auspices of the European Union and signed on 19 April 2013. Negotiations were led by Serbian Prime Minister Ivica Da\u010di\u0107 and Kosovo Prime Minister Hashim Tha\u00e7i and mediated by EU High Representative Catherine Ashton. The government of Serbia does not recognize Kosovo as a sovereign state, but began normalizing relations with the government of Kosovo nonetheless, in accordance with the Brussels URL In 2018, Serbia's president Aleksandar Vu\u010di\u0107 commented that the agreement is a difficult compromise for Serbia, which according to him fulfilled all of its obligations. Background Following the Kosovo War and NATO bombing of Yugoslavia in 1999, Kosovo, as part of Serbia and Federal Republic of Yugoslavia, was placed under United Nations administration under UNSC Resolution 1244. In 2008, Kosovo declared independence and has since been recognized by 96 countries. However, Serbia maintains that Kosovo continues to be its territory. The European Union-mediated dialogue between Kosovo and Serbia was launched in March 2011, three years after Kosovo declared URL Belgrade and Pristina were urged to continue talks in Brussels, but Serbia was not obliged to recognize Kosovo at any point in the URL Talks Until preset, ten rounds of talks have been held at the European External Action Service office in Brussels. Baroness Ashton, the EU High Representative, had been chairing talks for two years, followed by Federica Mogherini. Normalization of relations with neighboring states is a key precondition for states wishing to join the EU; the Brussels Agreement brought Serbia close to EU accession talks and Kosovo to initialing a Stabilisation and Association Agreement (SAA). The SAA was signed by HR Mogherini and PM Mustafa in October 2015. 330px|Ivica Da\u010di\u0107, Prime Minister of Serbia (left), Catherine Ashton (middle), and Hashim Tha\u00e7i, Prime Minister of Kosovo (right) in Munich, 2014. United States diplomats have supported the EU-led dialogue from the beginning. U.S. Secretary of State Hillary Clinton invited Baroness Ashton to undertake joint travel in the Balkans, and the two made joint visits to Belgrade and Pristina (as well as Sarajevo) in October 2012. After the agreement was concluded, the European Commission officially advised that work start on an SAA with Kosovo, and accession negotiations began with Serbia. The agreement was supported by the European Union, NATO, the OSCE, and the United Nations. In October 2013, arrangements were made for Serbian officials to visit North Kosovo. It was agreed that electoral bodies in Kosovo would include Kosovo Serb representatives and that international dialing code +383 would be attributed to Kosovo. After some delays, the new geographic phone code was implemented in December 2016. The National Assembly of Serbia has not treated the agreement as international and has not ratified it , which would be the procedure required in Serbia for approving an international agreement .Zakon o zakljucivanju i izvrsenju medjunarodnih ugovora It has, however, merely accepted the government report about the \"hitherto process of political and technical dialogue with the temporary institutions in Pristina with the mediation of EU, including the process of implementation of the achieved agreements\".O PRIHVATANjU IZVE\u0160TAJA O DOSADA\u0160NjEM PROCESS POLITI\u010cKOG I TEHNI\u010cKOG DIJALOGA SA PRIVREMENIM INSTITUCIJAMA SAMOUPRAVE U PRI\u0160TINI UZ POSREDOVANjE EVROPSKE UNIJE, UKLjU\u010cUJU\u0106I PROCES IMPLEMENTACIJE POSTIGNUTIH DOGOVORA The constitutional court in Belgrade did not answer a question about the constitutionality of the agreement, stating in December 2014 that the issue raised was a political question, and not a legal one. ", "edit_actions": [{"type": "R", "before": "nations", "after": "governments", "start_char_pos": 205, "end_char_pos": 212, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "KosovoAs to the agreement not having been signed, see text of agreement. .", "after": "Kosovo.", "start_char_pos": 227, "end_char_pos": 301, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "Vucic", "after": "Vu\u010di\u0107", "start_char_pos": 806, "end_char_pos": 811, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "98", "after": "96", "start_char_pos": 1220, "end_char_pos": 1222, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Kosovo has become a member of several international organizations, while according to Serbia and about half of the UN member states, Resolution 1244 remains legally binding on all parties. The EU-mediated", "after": "The European Union-mediated", "start_char_pos": 1303, "end_char_pos": 1507, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "D", "before": "Removing the parallel Serbian structures in the Northern part of Kosovo remains challenging.", "after": null, "start_char_pos": 2290, "end_char_pos": 2382, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "D", "before": "Clinton assigned Deputy Assistant Secretary of State ambassador Philip Reeker to support Ashton's efforts. Reeker and his team coordinated closely with EU colleagues and met with Serbian and Kosovar representatives on the margins of every dialogue session in Brussels, and during visits to the region.", "after": null, "start_char_pos": 2809, "end_char_pos": 3110, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "D", "before": "The result was the 15-point \"First Agreement of Principles Governing the Normalisation of Relations\", initialed but not signed on 19 April 2013.", "after": null, "start_char_pos": 3111, "end_char_pos": 3255, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "The new code was expected to go into effect in late 2015 but has again been delayed by Serbia until March 2016. The", "after": "After some delays, the new", "start_char_pos": 3755, "end_char_pos": 3870, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "into a binding law, which is a procedure required for international agreements in Serbia", "after": ", which would be the procedure required in Serbia for approving an international agreement", "start_char_pos": 4030, "end_char_pos": 4118, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "dialog", "after": "dialogue", "start_char_pos": 4286, "end_char_pos": 4292, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the question about", "after": "a question about the", "start_char_pos": 4699, "end_char_pos": 4717, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "is", "after": "was", "start_char_pos": 4801, "end_char_pos": 4803, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "D", "before": "Serbia has accepted, however, that progress in implementing the Brussels agreement is a prerequisite for EU accession talks within the opening of Chapter 35 negotiations in December 2015.", "after": null, "start_char_pos": 4847, "end_char_pos": 5034, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 426, 582, 935, 1148, 1233, 1302, 1491, 1858, 1974, 2082, 2222, 2289, 2382, 2529, 2608, 2808, 2915, 3110, 3255, 3418, 3509, 3594, 3754, 3866, 3926, 4049, 4431, 4846], "domain": "wiki"}
{"doc_id": "413204", "revision_depth": 2, "before_revision": "Cities that installed their sewage collection systems before the 1930s typically used single piping systems to transport both urban runoff and sewage. This type of collection system is referred to as a combined sewer system (CSS). The cities' rationale when combined sewers were built was that it would be cheaper to build just a single system. Burrian, Steven J., et al. (1999).\"The Historical Development of Wet-Weather Flow Management.\" EPA, National Risk Management Research Laboratory, Cincinnati, OH. Document No. EPA/600/JA-99/275. In these systems a sudden large rainfall that exceeds sewage treatment capacity is allowed to overflow directly from storm drains into receiving waters via structures called combined sewer overflows.EPA. \"Combined Sewer Overflow (CSO) Control Policy.\" Federal Register, 59 FR 18688. April 19, 1994. Storm drains are typically  at shallower depths than combined sewers ; because , while storm drains are designed to accept surface runoff from streets, combined sewers were designed to also accept sewage flows from buildings with basements  .", "after_revision": "Cities that installed their sewage collection systems before the 1930s typically used single piping systems to transport both urban runoff and sewage. This type of collection system is referred to as a combined sewer system (CSS). The cities' rationale when combined sewers were built was that it would be cheaper to build just a single system. In these systems a sudden large rainfall that exceeds sewage treatment capacity is allowed to overflow directly from storm drains into receiving waters via structures called combined sewer overflows.EPA. \"Combined Sewer Overflow (CSO) Control Policy.\" Federal Register, 59 FR 18688. April 19, 1994. Storm drains are typically installed at shallower depths than combined sewers . This is because combined sewers were designed to accept sewage flows from buildings with basements , in addition to receiving surface runoff from streets .", "edit_actions": [{"type": "D", "before": "Burrian, Steven J., et al. (1999).\"The Historical Development of Wet-Weather Flow Management.\" EPA, National Risk Management Research Laboratory, Cincinnati, OH. Document No. EPA/600/JA-99/275.", "after": null, "start_char_pos": 345, "end_char_pos": 538, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "installed", "start_char_pos": 865, "end_char_pos": 865, "major_intent": "clarity", "raw_intents": ["clarity", "others", "coherence"]}, {"type": "R", "before": "; because , while storm drains are designed to accept surface runoff from streets,", "after": ". This is because", "start_char_pos": 907, "end_char_pos": 989, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "D", "before": "also", "after": null, "start_char_pos": 1023, "end_char_pos": 1027, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "A", "before": null, "after": ", in addition to receiving surface runoff from streets", "start_char_pos": 1078, "end_char_pos": 1078, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 150, 230, 344, 379, 439, 538, 738, 790, 821, 837, 908], "domain": "wiki"}
{"doc_id": "437772", "revision_depth": 1, "before_revision": "Euthymia is defined as a normal, tranquil mental state or mood. It is often used to describe a stable mental state or mood in those affected with bipolar disorder that is neither manic nor depressive, yet is distinguishable from healthy controls. Euthymia is also used to describe the baseline of other cyclical mood disorders like major depressive disorder (MDD), borderline personality disorder (BPD) and narcissistic personality disorder (NPD). This state is the goal of psychiatric and psychological interventions. The word euthymia traces its roots to the Greek words eu, meaning well, and thymo, meaning soul or emotion. The word thymos also had four additional meanings: life energy; feelings and passions; desires and inclinations; and thought or intelligence. Euthymia is also derived from a verb, euthymeo, that means both I am happy, in good spirits and I make others happy, I reassure and encourage. This is the basis on which the first formal definition of euthymia was built.", "after_revision": "In psychiatry and psychology, euthymia is a normal, tranquil mental state or mood. In those with bipolar disorder, euthymia is a stable mental state or mood that is neither manic nor depressive, yet is distinguishable from healthy controls. Euthymia is also the baseline of other cyclical mood disorders like major depressive disorder (MDD), borderline personality disorder (BPD) and narcissistic personality disorder (NPD). This state is the goal of psychiatric and psychological interventions. The word euthymia is derived from the Greek words eu, well, and thymo, soul or emotion. The word thymos also had four additional meanings: life energy; feelings and passions; desires and inclinations; and thought or intelligence. Euthymia is also derived from a verb, euthymeo, that means both I am happy, in good spirits and I make others happy, I reassure and encourage. This is the basis on which the first formal definition of euthymia was built.", "edit_actions": [{"type": "R", "before": "Euthymia is defined as", "after": "In psychiatry and psychology, euthymia is", "start_char_pos": 0, "end_char_pos": 22, "major_intent": "others", "raw_intents": ["others", "others", "clarity"]}, {"type": "R", "before": "It is often used to describe", "after": "In those with bipolar disorder, euthymia is", "start_char_pos": 64, "end_char_pos": 92, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "in those affected with bipolar disorder", "after": null, "start_char_pos": 123, "end_char_pos": 162, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "used to describe", "after": null, "start_char_pos": 264, "end_char_pos": 280, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "traces its roots to", "after": "is derived from", "start_char_pos": 537, "end_char_pos": 556, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "meaning", "after": null, "start_char_pos": 577, "end_char_pos": 584, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "meaning", "after": null, "start_char_pos": 602, "end_char_pos": 609, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 63, 246, 447, 518, 626, 690, 713, 739, 768, 911]}
{"doc_id": "43854", "revision_depth": 1, "before_revision": "Reality-Virtuality Continuum.The Virtuality Continuum is a continuous scale ranging between the completely virtual, a Virtuality , and the completely real: Reality. The reality-virtuality continuum therefore encompasses all possible variations and compositions of real and virtual objects. It has been described as a concept in new media and computer science, but in fact it could be considered a matter of anthropology. The concept was first introduced by Paul Milgram. The area between the two extremes, where both the real and the virtual are mixed, is the so-called Mixed reality. This in turn is said to consist of both Augmented Reality , where the virtual augments the real, and Augmented virtuality, where the real augments the virtual. Cyberspace, the world's computer systems considered as an interconnected whole, can be thought of as a virtual reality; for instance, it is portrayed as such in the cyberpunk fiction of William Gibson and others. Second life and MMORPGs such as World of Warcraft are examples of artificial environments or virtual worlds (falling some way short of full virtual reality) in cyberspace.", "after_revision": "Reality-virtuality continuum.The virtuality continuum is a continuous scale ranging between the completely virtual, a virtuality , and the completely real: reality. The realityvirtuality continuum therefore encompasses all possible variations and compositions of real and virtual objects. It has been described as a concept in new media and computer science, but in fact it could be considered a matter of anthropology. The concept was first introduced by Paul Milgram. The area between the two extremes, where both the real and the virtual are mixed, is the so-called mixed reality. This in turn is said to consist of both augmented reality , where the virtual augments the real, and augmented virtuality, where the real augments the virtual. Cyberspace, the world's computer systems considered as an interconnected whole, can be thought of as a virtual reality; for instance, it is portrayed as such in the cyberpunk fiction of William Gibson and others. Second Life and MMORPGs such as World of Warcraft are examples of artificial environments or virtual worlds (falling some way short of full virtual reality) in cyberspace.", "edit_actions": [{"type": "R", "before": "Reality-Virtuality Continuum.The Virtuality Continuum", "after": "Reality-virtuality continuum.The virtuality continuum", "start_char_pos": 0, "end_char_pos": 53, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "Virtuality", "after": "virtuality", "start_char_pos": 118, "end_char_pos": 128, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Reality. The reality-virtuality", "after": "reality. The realityvirtuality", "start_char_pos": 156, "end_char_pos": 187, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "Mixed", "after": "mixed", "start_char_pos": 570, "end_char_pos": 575, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Augmented Reality", "after": "augmented reality", "start_char_pos": 625, "end_char_pos": 642, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Augmented", "after": "augmented", "start_char_pos": 686, "end_char_pos": 695, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "life", "after": "Life", "start_char_pos": 965, "end_char_pos": 969, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 29, 164, 289, 420, 470, 584, 744, 864, 957], "domain": "wiki"}
{"doc_id": "4409682", "revision_depth": 1, "before_revision": "Saraswata was brought up by Sage Dadhichi who found him as a child floating in the Saraswati River . Deva king Indra made a weapon named the Vajra using the skeleton of Dadhichi, which he used in his battle against his enemies, the Asuras. In this battle 99 heroes in the army of the Daityas, (a clan of Asuras) were slain. After this great war there was a drought extending for 12 years. During that drought extending for twelve years, the great sages (who practiced Vedic traditions and who were settled on the banks of Saraswati River) for the sake of sustenance, migrated from the river valley (9:51). However, Sage Saraswata continued to live on the banks of the Saraswati. After the drought of 12 years had died, the great sages solicited one another for lectures on the Vedas. While wandering with famished stomachs, the sages had lost the knowledge of the Vedas. There was, indeed, not one amongst them that could understand the scriptures. It chanced that someone amongst them encountered Saraswata, that foremost of sages, while the latter was reading the Vedas with concentrated attention. Coming back to the conclave of rishis, he spoke to them of Saraswata, of unrivalled splendour and god-like form engaged in reading the Vedas in a solitary forest. Then all the great Rishis came to that spot, and jointly asked him to teach them the Vedas. Those sages duly became his disciples and obtaining from him their knowledge of the Vedas, once more began to practice their rites. A total of 60,000 sages became disciples of the venerable Rishi Saraswata for the sake of acquiring their knowledge of the Vedas from him (9:51). The present day Goud Saraswat Brahmins who are Goud saraswat , Shenvis, Chitrapur saraswats, bhalvalikar, Rajapur Saraswats, and pednekars, have this event as part of their culture. This event is repeated again at (3:85):- At the forest of Tungaka in olden days sage Saraswata taught the Vedas to the ascetics. When the Vedas had been lost in consequence of the sages having URLotten them. Here (3:85) he is mentioned as the son of Angirasa. The asylum of Aditya, the place of Soma and the hermitage of Dadhichi were mentioned as adjacent places on the banks of Saraswati at (3:83). Here it is mentioned that Angirasa - a great sage belonging to the Saraswata race - was born in the ashram of Dadhichi . At (13:50) Saraswata is  mentioned as Sage Atris son. Saraswata is mentioned as a sage from the western regions of India at (12:207, 13:165).", "after_revision": "Saraswata was born to family of sage brighu his father was sage (some time prince) dhadichi ( not one who donate his bones for vajra) . Deva king Indra made a weapon named the Vajra using the skeleton of Dadhichi, which he used in his battle against his enemies, the Asuras. In this battle 99 heroes in the army of the Daityas, (a clan of Asuras) were slain. After this great war there was a drought extending for 12 years. During that drought extending for twelve years, the great sages (who practiced Vedic traditions and who were settled on the banks of Saraswati River) for the sake of sustenance, migrated from the river valley (9:51). However, Sage Saraswata continued to live on the banks of the Saraswati. After the drought of 12 years had died, the great sages solicited one another for lectures on the Vedas. While wandering with famished stomachs, the sages had lost the knowledge of the Vedas. There was, indeed, not one amongst them that could understand the scriptures. It chanced that someone amongst them encountered Saraswata, that foremost of sages, while the latter was reading the Vedas with concentrated attention. Coming back to the conclave of rishis, he spoke to them of Saraswata, of unrivalled splendour and god-like form engaged in reading the Vedas in a solitary forest. Then all the great Rishis came to that spot, and jointly asked him to teach them the Vedas. Those sages duly became his disciples and obtaining from him their knowledge of the Vedas, once more began to practice their rites. A total of 60,000 sages became disciples of the venerable Rishi Saraswata for the sake of acquiring their knowledge of the Vedas from him (9:51). The present day Saraswat Brahmins who are kashmiri pandits, punjabi brahmins, Goud saraswat brahmins , Shenvis, Chitrapur saraswats, bhalvalikar, Rajapur Saraswats, and pednekars, have this event as part of their culture. This event is repeated again at (3:85):- At the forest of Tungaka in olden days sage Saraswata taught the Vedas to the ascetics. When the Vedas had been lost in consequence of the sages having URLotten them. Here (3:85) he is mentioned as the son of Angirasa. The asylum of Aditya, the place of Soma and the hermitage of Dadhichi were mentioned as adjacent places on the banks of Saraswati at (3:83). Here it is mentioned that Angirasa - a great sage belonging to the Saraswata race . At (13:50) Saraswata is sometimes mentioned as Sage Atris son. Saraswata is mentioned as a sage from the western regions of India at (12:207, 13:165).", "edit_actions": [{"type": "R", "before": "brought up by Sage Dadhichi who found him as a child floating in the Saraswati River", "after": "born to family of sage brighu his father was sage (some time prince) dhadichi ( not one who donate his bones for vajra)", "start_char_pos": 14, "end_char_pos": 98, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "D", "before": "Goud", "after": null, "start_char_pos": 1650, "end_char_pos": 1654, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "Goud saraswat", "after": "kashmiri pandits, punjabi brahmins, Goud saraswat brahmins", "start_char_pos": 1681, "end_char_pos": 1694, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "- was born in the ashram of Dadhichi", "after": null, "start_char_pos": 2299, "end_char_pos": 2335, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "sometimes", "start_char_pos": 2362, "end_char_pos": 2362, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}], "sents_char_pos": [0, 239, 323, 388, 605, 678, 783, 870, 948, 1100, 1263, 1355, 1487, 1633, 1815, 1944, 2023, 2075, 2216, 2337, 2391], "domain": "wiki"}
{"doc_id": "44390924", "revision_depth": 1, "before_revision": "Children of the Open Road () is a Swiss feature/drama film that was produced in 1992, focusing on the fictitious family, the Kesslers, and the historically real attempt of the Swiss to eliminate the Yeniche culture in Switzerland by the Kinder der Landstrasse foundation  between 1926 and 1973. Plot The Yeniche Kessel family Theresa, Paul and their five-year-old daughter Jana - escapes the Nazi terror and returns to Switzerland in 1939. Trusting to have escaped the hell, they get into their supposedly safe home into the clutches of the auxiliary Kinder der Landstrasse and victims of racial persecution in the name of so-called children's welfare: Jana Kessler (Martina Straessler, Jara Weiss as a child, and Jasmin Tabatabai as adult Jana), in 1939 five years old, has been snatched from her parents , and the following years she's consigned to a life of orphanages and foster homes, in order to sever her ties with her culture and to 'assimilate' Jana to a 'better way of life'.", "after_revision": "Children of the Open Road () is a Swiss feature/drama film that was produced in 1992. Its topic is the Kinder der Landstrasse foundation , active between 1926 and 1973, which controversially attempted to assimilate the itinerant Yeniche population of Switzerland by forcibly moving their children to foster homes or orphanages. The historical topic is presented in fictionalized account. Plot The Yeniche Kessel family Theresa, Paul and their five-year-old daughter Jana - escapes the Nazi terror and returns to Switzerland in 1939. They become victims of the Kinder der Landstrasse activities as Jana Kessler (Martina Straessler, Jara Weiss as a child, and Jasmin Tabatabai as adult Jana), in 1939 five years old, is snatched from her parents and consigned to a life of orphanages and foster homes, in order to sever her ties with her culture and to 'assimilate' Jana to a 'better way of life'.", "edit_actions": [{"type": "R", "before": "1992, focusing on the fictitious family, the Kesslers, and the historically real attempt of the Swiss to eliminate the Yeniche culture in Switzerland by the", "after": "1992. Its topic is the", "start_char_pos": 80, "end_char_pos": 236, "major_intent": "coherence", "raw_intents": ["coherence", "others", "clarity"]}, {"type": "A", "before": null, "after": ", active", "start_char_pos": 271, "end_char_pos": 271, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "1973.", "after": "1973, which controversially attempted to assimilate the itinerant Yeniche population of Switzerland by forcibly moving their children to foster homes or orphanages. The historical topic is presented in fictionalized account.", "start_char_pos": 289, "end_char_pos": 294, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Trusting to have escaped the hell, they get into their supposedly safe home into the clutches of the auxiliary", "after": "They become victims of the", "start_char_pos": 440, "end_char_pos": 550, "major_intent": "coherence", "raw_intents": ["coherence", "others", "clarity"]}, {"type": "R", "before": "and victims of racial persecution in the name of so-called children's welfare:", "after": "activities as", "start_char_pos": 574, "end_char_pos": 652, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "has been", "after": "is", "start_char_pos": 771, "end_char_pos": 779, "major_intent": "fluency", "raw_intents": ["fluency", "style", "fluency"]}, {"type": "R", "before": ", and the following years she's", "after": "and", "start_char_pos": 806, "end_char_pos": 837, "major_intent": "coherence", "raw_intents": ["coherence", "others", "clarity"]}], "sents_char_pos": [0, 294, 439, 652], "domain": "wiki"}
{"doc_id": "483672", "revision_depth": 1, "before_revision": "Most land trusts are URLanizations which owns and manages land, and sometimes waters. There are three main types of land trust, distinguished from one another by the ways in which they are legally structured and by the purposes for which they URLanized and operated . A real estate investment trust is a fiduciary arrangement whereby one party (the trustee) agrees to own and to manage real property for the benefit of a limited number of beneficiaries. Aldrich, R. Wyerman, J. 2005 National Land Trust Census Report (2006), Land Trust Alliance A community land trust (CLT) is a private, nonprofit corporation that acquires, manages, and develops land for a variety of purposes, primarily for the production and stewardship of affordable housing, although many CLTs are also engaged in non-residential buildings and uses. A conservation land trust is a private, non-profit corporation in the US that acquires land or conservation easements for the purpose of limiting commercial development and preserving open space, natural areas, waterways, and/or productive farms and forests. Another of landholding is sometimes referred to as a \"trust\" in the United States ; approximately of land are owned by the United States government and held in trust for Native American tribes and individuals .U.S. Department of the Interior. Office of the Special Trustee. History. In Australia, Aboriginal land trusts are a type of URLanisation that holds the freehold title to an area of land on behalf of a community of Aboriginal and/or Torres Strait Islander people. History Ancient example Possible earliest concept of equity in land held in trust is the depiction of this ancient king (trustor) which grants property back to its previous owner (beneficiary) during his absence, supported by witness testimony (trustee). In essence and in this case, the king, in place of the later state (trustor and holder of assets at highest position) issues ownership along with past proceeds (equity) back to the beneficiary:  Land trusts have been actively used in Illinois for over a hundred years and in recent decades have begun to be used in other states. The declaration of a trust is through a \"deed to trustee\". If the trust is filed as a public document, it removes all of the asset protection provided by the formation of the land trust. Robert Pless pioneered the use of the land trust that has been used by many firms throughout the United States since the early 1990s.   Conservation land trusts The Land Trust Alliance, formed in 1981, provides technical support to land trusts in the United States. The Alliance performs a National Land Trust Census that keeps track of the land protected by local and regional land trusts. The last Census, conducted in 2003, reported that these trusts have protected almost 9.4million acres (38,000km2) of land in the United States, double the 4.7million acres (19,000km2) recorded in the 1998 survey. Over 5million acres (20,000km2) of that was protected by conservation easement in 2003. Although it does not include national or international land trusts in its Census, the LTA estimates another 25million acres (100,000km2) in the U.S. have been protected by URLanizations. The largest amount of land protected by local and regional trusts is in the Northeast with 2.9million acres (12,000km2), while the fastest growing region between 1998 and 2003 was the Pacific (consisting of California, Nevada, and Hawaii), with protected land increasing 147\\% to 1.5million acres (6,100km2) in 2003.  ", "after_revision": "Land trusts are URLanizations which own and manage land, and sometimes waters. There are three common types of land trust, distinguished from one another by the ways in which they are legally structured and by the purposes for which they URLanized and operated : A real estate investment trust is a fiduciary arrangement whereby one party (the trustee) agrees to own and to manage real property for the benefit of a limited number of beneficiaries. A community land trust (CLT) is a private, nonprofit corporation that acquires, manages, and develops land for a variety of purposes, primarily for the production and stewardship of affordable housing, although many CLTs are also engaged in non-residential buildings and uses. A conservation land trust is a private, non-profit corporation in the US that acquires land or conservation easements for the purpose of limiting commercial development and preserving open space, natural areas, waterways, and/or productive farms and forests. In the United States , the land owned by the United States government and held in trust for Native American tribes and individuals is sometimes referred to as a land trust. In Australia, Aboriginal land trusts are a type of URLanisation that holds the freehold title to an area of land on behalf of a community of Aboriginal and/or Torres Strait Islander people. History Ancient example Possible earliest concept of equity in land held in trust is the depiction of this ancient king (trustor) which grants property back to its previous owner (beneficiary) during his absence, supported by witness testimony (trustee). In essence and in this case, the king, in place of the later state (trustor and holder of assets at highest position) issues ownership along with past proceeds (equity) back to the beneficiary: Ben-Barak, Zafrira. \"Meribaal and the System of Land Grants in Ancient Israel.\" Biblica (1981): 73-91. Land trusts have been actively used in Illinois for over a hundred years and in recent decades have begun to be used in other states. The declaration of a trust is through a \"deed to trustee\". If the trust is filed as a public document, it removes all of the asset protection provided by the formation of the land trust. Robert Pless pioneered the use of the land trust that has been used by many firms throughout the United States since the early 1990s. A real estate investment trust is a fiduciary arrangement whereby one party (the trustee) agrees to own and to manage real property for the benefit of a limited number of beneficiaries.Aldrich, R. Wyerman, J. 2005 National Land Trust Census Report (2006), Land Trust Alliance Conservation land trusts The Land Trust Alliance, formed in 1981, provides technical support to land trusts in the United States. The Alliance performs a National Land Trust Census that keeps track of the land protected by local and regional land trusts. The last Census, conducted in 2003, reported that these trusts have protected almost 9.4million acres (38,000km2) of land in the United States, double the 4.7million acres (19,000km2) recorded in the 1998 survey. Over 5million acres (20,000km2) of that was protected by conservation easement in 2003. Although it does not include national or international land trusts in its Census, the LTA estimates another 25million acres (100,000km2) in the U.S. have been protected by URLanizations. The largest amount of land protected by local and regional trusts is in the Northeast with 2.9million acres (12,000km2), while the fastest growing region between 1998 and 2003 was the Pacific (consisting of California, Nevada, and Hawaii), with protected land increasing 147\\% to 1.5million acres (6,100km2) in 2003. Other types In the United States; approximately (as of 2008) of land are owned by the United States government and held in trust for Native American tribes and individuals.U.S. Department of the Interior. Office of the Special Trustee. History. In Australia, Aboriginal land trusts are a type of URLanisation that holds the freehold title to an area of land on behalf of a community of Aboriginal and/or Torres Strait Islander people.", "edit_actions": [{"type": "R", "before": "Most land", "after": "Land", "start_char_pos": 0, "end_char_pos": 9, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "owns and manages", "after": "own and manage", "start_char_pos": 41, "end_char_pos": 57, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "main", "after": "common", "start_char_pos": 102, "end_char_pos": 106, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ".", "after": ":", "start_char_pos": 266, "end_char_pos": 267, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "D", "before": "Aldrich, R.", "after": null, "start_char_pos": 454, "end_char_pos": 465, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "Wyerman, J. 2005 National Land Trust Census Report (2006), Land Trust Alliance", "after": null, "start_char_pos": 466, "end_char_pos": 544, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "Another of landholding is sometimes referred to as a \"trust\" in", "after": "In", "start_char_pos": 1081, "end_char_pos": 1144, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "; approximately of land are", "after": ", the land", "start_char_pos": 1163, "end_char_pos": 1190, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": ".U.S. Department of the Interior. Office of the Special Trustee. History.", "after": "is sometimes referred to as a land trust.", "start_char_pos": 1290, "end_char_pos": 1363, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Ben-Barak, Zafrira. \"Meribaal and the System of Land Grants in Ancient Israel.\" Biblica (1981): 73-91.", "start_char_pos": 2003, "end_char_pos": 2003, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "A real estate investment trust is a fiduciary arrangement whereby one party (the trustee) agrees to own and to manage real property for the benefit of a limited number of beneficiaries.Aldrich, R.", "start_char_pos": 2459, "end_char_pos": 2459, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Wyerman, J. 2005 National Land Trust Census Report (2006), Land Trust Alliance", "start_char_pos": 2460, "end_char_pos": 2460, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "A", "before": null, "after": "Other types In the United States; approximately (as of 2008) of land are owned by the United States government and held in trust for Native American tribes and individuals.U.S. Department of the Interior. Office of the Special Trustee. History.", "start_char_pos": 3521, "end_char_pos": 3521, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "In Australia, Aboriginal land trusts are a type of URLanisation that holds the freehold title to an area of land on behalf of a community of Aboriginal and/or Torres Strait Islander people.", "start_char_pos": 3522, "end_char_pos": 3522, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 85, 267, 453, 821, 1080, 1164, 1323, 1354, 1553, 1808, 2137, 2196, 2324, 2458, 2590, 2715, 2928, 3016, 3203], "domain": "wiki"}
{"doc_id": "49907", "revision_depth": 1, "before_revision": "Provincial tour 350px|The Roman Empire during the reign of Caracalla In spring 214, Caracalla departed for the eastern provinces, travelling through the Danubian provinces and arriving in Asia and Bithynia. The winter of 214/215 he spent at Nicomedia. Before 4 April 214 he had left Nicomedia, and in the summer he was at Antioch on the Orontes. From December 215 he was at Alexandria in the Nile Delta, where he stayed until March or April 216. When the inhabitants of Alexandria heard of Caracalla's claims that he had killed his brother Geta in self-defence, they produced a satire mocking this as well as Caracalla's other pretensions. In 215 Caracalla travelled to Alexandria and responded to this insult by slaughtering the deputation of leading citizens who had unsuspectingly assembled before the city to greet his arrival  , before setting his troops against Alexandria for several days of looting and plunder. In spring 216 he arrived again at Antioch and before 27 May had set out for his Persian War. For the winter of 215/216 he was at Edessa. Caracalla  moved east into Armenia. By 216 he had pushed through Armenia and south into Parthia.", "after_revision": "Provincial tour 350px|The Roman Empire during the reign of Caracalla In spring 214, Caracalla departed for the eastern provinces, travelling through the Danubian provinces and the Anatolian provinces of Asia and Bithynia. He spent the winter of 214/215 in Nicomedia. By 4 April 215 he had left Nicomedia, and in the summer he was at Antioch on the Orontes. From December 215 he was in Alexandria in the Nile Delta, where he stayed until March or April 216. When the inhabitants of Alexandria heard of Caracalla's claims that he had killed his brother Geta in self-defence, they produced a satire mocking this as well as Caracalla's other pretensions. Caracalla responded to this insult by slaughtering the deputation of leading citizens who had unsuspectingly assembled before the city to greet his arrival in December 215 , before setting his troops against Alexandria for several days of looting and plunder. In spring 216 he returned to Antioch and before 27 May had set out to lead his Roman army against the Parthians. During the winter of 215/216 he was in Edessa. Caracalla then moved east into Armenia. By 216 he had pushed through Armenia and south into Parthia.", "edit_actions": [{"type": "R", "before": "arriving in", "after": "the Anatolian provinces of", "start_char_pos": 176, "end_char_pos": 187, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "The", "after": "He spent the", "start_char_pos": 207, "end_char_pos": 210, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "he spent at Nicomedia. Before", "after": "in Nicomedia. By", "start_char_pos": 229, "end_char_pos": 258, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "214", "after": "215", "start_char_pos": 267, "end_char_pos": 270, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "at", "after": "in", "start_char_pos": 371, "end_char_pos": 373, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "In 215 Caracalla travelled to Alexandria and", "after": "Caracalla", "start_char_pos": 640, "end_char_pos": 684, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "in December 215", "start_char_pos": 831, "end_char_pos": 831, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "arrived again at", "after": "returned to", "start_char_pos": 937, "end_char_pos": 953, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "for his Persian War. For", "after": "to lead his Roman army against the Parthians. During", "start_char_pos": 992, "end_char_pos": 1016, "major_intent": "others", "raw_intents": ["others", "others", "clarity"]}, {"type": "R", "before": "at", "after": "in", "start_char_pos": 1046, "end_char_pos": 1048, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "others", "fluency"]}, {"type": "A", "before": null, "after": "then", "start_char_pos": 1067, "end_char_pos": 1067, "major_intent": "others", "raw_intents": ["others", "others", "coherence"]}], "sents_char_pos": [0, 206, 251, 345, 445, 639, 919, 1012, 1056, 1092], "domain": "wiki"}
{"doc_id": "49949965", "revision_depth": 1, "before_revision": "Kulbhushan Sudhir Jadhav (also spelled Kulbhushan Yadav, alleged alias Hussain Mubarak Patel) (born 16 April 1970) is an Indian national. It is alleged by the Pakistani government that he was arrested in the Pakistani province of Balochistan on charges of terrorism and spying for India's  Research and Analysis Wing. On 10 April 2017, India's Ministry of External Affairs stated that he had been \"kidnapped last year from Iran and his subsequent presence in Pakistan has never been explained credibly\". The Pakistani government claimed that he was a commander in the Indian Navy who was involved in subversive activities inside Pakistan and was arrested on 3 March 2016 during a counter-intelligence operation in Balochistan. The Indian government recognised Jadhav as a former naval officer but denied any current links with him and maintained that he took premature retirement and operated a business in Iranian city of Chabahar and was abducted from there .Revealed: 'Spy' Kulbhushan Yadav not caught but abducted by extremist Sunni group Jaishul Adil , India Today, 30 March 2016. Pakistani security forces reported Jadhav as a serving officer in the Indian Navy and stated that he was commissioned to the Research and Analysis Wing  , India's external intelligence agency. They believed him to be involved in subversive activities in Balochistan and Karachi. Jadhav was soon shifted to Islamabad for interrogation. While according to Indian sources, Jadhav was kidnapped by Mullah Shahozi of Jaish ul-Adl from Chabahar and was transported via Sarbaz to IranPakistan border and where he handed Jadhav over to the Pakistani Army. In a statement issued in the Parliament of India on 11 April 2017, Rajnath Singh, India's Minister of Home Affairs, reiterated that Jadhav was kidnapped by Pakistani agencies from Iran and put through trial as a R AW agent. Sushma Swaraj, India's Minister of External Affairs said there was no evidence of any wrongdoing by Jadhav and termed his sentencing an act of \"premeditated murder\". Swaraj said that if Pakistan implemented the death sentence, the bilateral relations between both countries would face dire consequences. The ICJ proceedings began in The Hague on 15 May to review the case. India and Pakistan both sent their legal teams to put forward their arguments, led by Harish Salve and Khawar Qureshi QC respectively. On 18 May 2017, the International Court of Justice stayed the hanging of Jadhav by Pakistan . On 22 June 2017, Pakistani sources confirmed that Jadhav had sought clemency from the country's army chief following his conviction. Pakistan also released a new confessional video of Jadhav, in which he stated that he visited Karachi twice for gathering intelligence on naval facilities. He also admitted to supporting and funding, on behalf of R AW , Baloch militants affiliated with the BLA and BRA, in addition to infiltrating and establishing \"30 to 40 R AW operatives along the Makran Coast\" for involvement in terrorist activities. Jadhav said that R AW 's activities in Balochistan and Sindh were conducted under the direction of Anil Kumar Dhasmana. India's Foreign Ministry again dismissed the confession as \"false propaganda\", stating that Pakistan was trying to influence ICJ proceedings while denying the consular rights to Jadhav. ICJ did not allow Pakistan to play the video during the hearing. On 17 July 2019, the ICJ, in its verdict, rejected India's appeal for Jadhav's release and directed Pakistan to allow consular access and effectively Stayed the death sentence. The court ordered that Pakistan should grant consular access to Jhadav. Pakistan and India both claimed victory and welcomed the judgement of ICJ.", "after_revision": "Kulbhushan Sudhir Jadhav (also spelled Kulbhushan Yadav, alleged alias Hussain Mubarak Patel) (born 16 April 1970) is an Indian national. It is alleged by the Pakistani government that he was arrested in the Pakistani province of Balochistan on charges of terrorism and spying for India's intelligence agency, the Research and Analysis Wing. On 10 April 2017, Indian foreign ministry alleged he had been \"kidnapped last year from Iran and his subsequent presence in Pakistan has never been explained credibly\". The Pakistani government stated that he was a commander in the Indian Navy who was involved in subversive activities inside Pakistan and was arrested on 3 March 2016 during a counter-intelligence operation in Balochistan. The Indian government recognised Jadhav as a former naval officer but denied any current links with him and maintained that he took premature retirement and was abducted from Iran .Revealed: 'Spy' Kulbhushan Yadav not caught but abducted by extremist Sunni group Jaishul Adil , India Today, 30 March 2016. Pakistani security forces reported Jadhav as a serving officer in the Indian Navy and stated that he was commissioned to the Research and Analysis Wing (RAW) , India's external intelligence agency. They believed him to be involved in subversive activities in Balochistan and Karachi. Jadhav was soon shifted to Islamabad for interrogation. While according to Indian sources, Jadhav was kidnapped by Mullah Omar Irani of Jaish ul-Adl from Sarbaz City, Iran who later handed over Jadhav to the Pakistani Army. In a statement issued in the Parliament of India on 11 April 2017, Rajnath Singh, India's Minister of Home Affairs, reiterated that Jadhav was kidnapped by Pakistani agencies from Iran and put through trial as a RAW agent. Sushma Swaraj, India's Minister of External Affairs said there was no evidence of any wrongdoing by Jadhav and termed his sentencing an act of \"premeditated murder\". Swaraj said that if Pakistan implemented the death sentence, the bilateral relations between both countries would face dire consequences. The ICJ proceedings began in The Hague on 15 May to review the case. India and Pakistan both sent their legal teams to put forward their arguments, led by Harish Salve and Khawar Qureshi respectively. On 18 May 2017, the International Court of Justice stayed the hanging of Jadhav . On 22 June 2017, Pakistani sources confirmed that Jadhav had sought clemency from the country's army chief following his conviction. Pakistan also released a new confessional video of Jadhav, in which he stated that he visited Karachi twice for gathering intelligence on naval facilities. He also admitted to supporting and funding, on behalf of India's RAW , Baloch militants affiliated with the BLA and BRA, in addition to infiltrating and establishing \"30 to 40 RAW operatives along the Makran Coast\" for involvement in terrorist activities. Jadhav said that RAW 's activities in Balochistan and Sindh were conducted under the direction of Anil Kumar Dhasmana. India's Foreign Ministry again dismissed the confession as \"false propaganda\", stating that Pakistan was trying to influence ICJ proceedings while denying the consular rights to Jadhav. ICJ did not allow Pakistan to play the video during the hearing. On 17 July 2019, the ICJ, in its verdict, rejected India's appeal for Jadhav's release and directed Pakistan to allow consular access and effectively review the death sentence. The court ordered that Pakistan should grant consular access to Jhadav. Pakistan and India both claimed victory and welcomed the judgement of ICJ.", "edit_actions": [{"type": "A", "before": null, "after": "intelligence agency, the", "start_char_pos": 289, "end_char_pos": 289, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "India's Ministry of External Affairs stated that", "after": "Indian foreign ministry alleged", "start_char_pos": 336, "end_char_pos": 384, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "claimed", "after": "stated", "start_char_pos": 529, "end_char_pos": 536, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "operated a business in Iranian city of Chabahar and", "after": null, "start_char_pos": 884, "end_char_pos": 935, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "there", "after": "Iran", "start_char_pos": 954, "end_char_pos": 959, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "(RAW)", "start_char_pos": 1238, "end_char_pos": 1238, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "others", "style"]}, {"type": "R", "before": "Shahozi", "after": "Omar Irani", "start_char_pos": 1487, "end_char_pos": 1494, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Chabahar and was transported via Sarbaz to IranPakistan border and where he handed Jadhav over", "after": "Sarbaz City, Iran who later handed over Jadhav", "start_char_pos": 1516, "end_char_pos": 1610, "major_intent": "others", "raw_intents": ["others", "others", "clarity"]}, {"type": "D", "before": "R", "after": null, "start_char_pos": 1846, "end_char_pos": 1847, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": "AW", "after": "RAW", "start_char_pos": 1848, "end_char_pos": 1850, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "D", "before": "QC", "after": null, "start_char_pos": 2349, "end_char_pos": 2351, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "by Pakistan", "after": null, "start_char_pos": 2446, "end_char_pos": 2457, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "R", "after": null, "start_char_pos": 2806, "end_char_pos": 2807, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "AW", "after": "India's RAW", "start_char_pos": 2808, "end_char_pos": 2810, "major_intent": "others", "raw_intents": ["others", "others", "clarity"]}, {"type": "D", "before": "R", "after": null, "start_char_pos": 2918, "end_char_pos": 2919, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "AW", "after": "RAW", "start_char_pos": 2920, "end_char_pos": 2922, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "D", "before": "R", "after": null, "start_char_pos": 3016, "end_char_pos": 3017, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "AW", "after": "RAW", "start_char_pos": 3018, "end_char_pos": 3020, "major_intent": "others", "raw_intents": ["others", "others", "others"]}, {"type": "R", "before": "Stayed", "after": "review", "start_char_pos": 3520, "end_char_pos": 3526, "major_intent": "others", "raw_intents": ["others", "others", "style"]}], "sents_char_pos": [0, 137, 317, 503, 726, 961, 1085, 1278, 1364, 1420, 1633, 1857, 2023, 2161, 2230, 2365, 2459, 2592, 2748, 2998, 3118, 3304, 3369, 3546, 3618], "domain": "wiki"}
{"doc_id": "50178403", "revision_depth": 1, "before_revision": "Green hunting (  eco-hunting, green bullet concept, green darting , darting safari) is the practice of hunting game animals with tranquilizer dart guns or bows and the subsequent release of the live animals . Green hunting would typically be performed when tranquilization of the animal is necessary for veterinary, monitoring , or translocation purposes. Green hunting has been advocated as a conservation-minded alternative to destructive sport hunting, because it allows the hunter to experience the thrill of a traditional hunt without killing the animal, leaving the wildlife abundance undiminished and  contributing directly to conservation initiatives. The measurements and details of a tranquilized animal are accepted in URLanization's record books and fiberglass replicas can be made for the hunter. There are some  advantages of eco-hunting . The hunter may have the additional option to be kept regularly updated on the animal's future movements if the darting was part of a GPS collaring project. Though still costly (sometimes upwards of $25,000 USD), green hunting is more economical as fees are lower than trophy hunting (up to $ 60,000 USD). The fees finance the conservation project that necessitates an immobilized animal for micro-chipping , ear-notching, tissue collection or GPS tracking and any additional funds can be used to support the management of the protected area involved. The hunter must get within 30 meters from the animal to successfully dart it, often much closer than required for traditional hunting. Group darting safaris also exist, where a group witnesses the work, without participating in the procedures.", "after_revision": "Green hunting ( also eco-hunting, green bullet concept, green darting or darting safari) is the practice of hunting game animals with non-lethal tranquilizer guns or bows and subsequently releasing the captured animals alive . Green hunting would typically be performed when tranquilization of the animal is necessary for veterinary, monitoring or translocation purposes. Green hunting has been advocated as a conservation-minded alternative to destructive sport hunting, because it allows the hunter to experience the thrill of a traditional hunt without killing the animal, leaving the wildlife abundance undiminished and thus contributing directly to conservation initiatives. The measurements and details of a tranquilized animal are generally accepted by URLanization's logbooks and fiberglass trophies can be made to accolade the hunter. There are some other advantages of eco-hunting , as the hunter has the additional option to be kept regularly updated on the animal's future movements if the darting was part of a GPS tracking project. Though still costly (sometimes upwards of $25,000 USD), green hunting is more economical as fees are typically lower than trophy hunting (up to $ 60,000 USD). The fees finance the conservation project that necessitates an immobilized animal for microchipping , ear-notching, tissue collection or GPS tracking and any additional funds can be used to support the management of the protected area involved. Due to the significantly reduced terminal ballistics, the hunter must get within from the target animal to successfully dart it, often much closer than required for traditional hunting. Group darting safaris also exist, where a group of spectators witness the hunt without participating in the procedures.", "edit_actions": [{"type": "A", "before": null, "after": "also", "start_char_pos": 16, "end_char_pos": 16, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": ",", "after": "or", "start_char_pos": 66, "end_char_pos": 67, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "tranquilizer dart", "after": "non-lethal tranquilizer", "start_char_pos": 129, "end_char_pos": 146, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "the subsequent release of the live animals", "after": "subsequently releasing the captured animals alive", "start_char_pos": 164, "end_char_pos": 206, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 327, "end_char_pos": 328, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "thus", "start_char_pos": 608, "end_char_pos": 608, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "R", "before": "accepted in", "after": "generally accepted by", "start_char_pos": 718, "end_char_pos": 729, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "record books and fiberglass replicas", "after": "logbooks and fiberglass trophies", "start_char_pos": 745, "end_char_pos": 781, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "for", "after": "to accolade", "start_char_pos": 794, "end_char_pos": 797, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "other", "start_char_pos": 825, "end_char_pos": 825, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": ". The hunter may have", "after": ", as the hunter has", "start_char_pos": 852, "end_char_pos": 873, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "collaring", "after": "tracking", "start_char_pos": 991, "end_char_pos": 1000, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "$25,000 USD), green hunting is more economical as fees are lower than trophy hunting (up to $", "after": "$25,000 USD), green hunting is more economical as fees are typically lower than trophy hunting (up to $", "start_char_pos": 1052, "end_char_pos": 1145, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "micro-chipping", "after": "microchipping", "start_char_pos": 1245, "end_char_pos": 1259, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "The", "after": "Due to the significantly reduced terminal ballistics, the", "start_char_pos": 1405, "end_char_pos": 1408, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "30 meters from the", "after": "from the target", "start_char_pos": 1432, "end_char_pos": 1450, "major_intent": "style", "raw_intents": ["style", "style", "others"]}, {"type": "R", "before": "witnesses the work,", "after": "of spectators witness the hunt", "start_char_pos": 1588, "end_char_pos": 1607, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 355, 659, 809, 853, 1009, 1158, 1404, 1539], "domain": "wiki"}
{"doc_id": "513965", "revision_depth": 1, "before_revision": " In the 1800s, rammed earth was popularized in the United States by the book Rural Economy by S. W. Johnson. The technique was used to construct the Borough House Plantation and the Church of the Holy Cross in Stateburg, South Carolina, both being National Historic Landmarks. From the 1920s through the 1940s rammed-earth construction in the US was studied. South Dakota State College extensively researched and constructed almost one hundred weathering walls of rammed earth. For over 30 years the college investigated the use of paints and plasters in relation to colloids in soil. In 1945, Clemson Agricultural College of South Carolina published the results of their research of rammed earth in a pamphlet titled \"Rammed Earth Building Construction\". In 1936, on a homestead near Gardendale, Alabama, the United States Department of Agriculture constructed an experimental community of rammed-earth edifices with architect Thomas Hibben. The houses were inexpensively constructed and were sold to the public along with sufficient land for gardens and small plots for livestock. The project successfully provided valuable homes to low-income families. The US Agency for International Development is working with undeveloped countries to improve the engineering of rammed-earth houses. It also financed the authorship of the Handbook of Rammed Earth by Texas A&M University and the Texas Transportation Institute. The Handbook was unavailable for purchase by the public until the Rammed Earth Institute International gained permission to reprint it.", "after_revision": "United States and Canada In the 1800s, rammed earth was popularized in the United States by the book Rural Economy by S. W. Johnson. The technique was used to construct the Borough House Plantation and the Church of the Holy Cross in Stateburg, South Carolina, both being National Historic Landmarks. From the 1920s through the 1940s rammed-earth construction in the US was studied. South Dakota State College extensively researched and constructed almost one hundred weathering walls of rammed earth. For over 30 years the college investigated the use of paints and plasters in relation to colloids in soil. In 1945, Clemson Agricultural College of South Carolina published the results of their research of rammed earth in a pamphlet titled \"Rammed Earth Building Construction\". In 1936, on a homestead near Gardendale, Alabama, the United States Department of Agriculture constructed an experimental rammed-earth edifices with architect Thomas Hibben. The houses were inexpensively constructed and were sold to the public along with sufficient land for gardens and small plots for livestock. The project was successful provided homes to low-income families. The US Agency for International Development is working with developing countries to improve the engineering of rammed-earth houses. It also financed the authorship of the Handbook of Rammed Earth by Texas A&M University and the Texas Transportation Institute. The Handbook was unavailable for purchase by the public until the Rammed Earth Institute International gained permission to reprint it.", "edit_actions": [{"type": "A", "before": null, "after": "United States and Canada", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "community of", "after": null, "start_char_pos": 878, "end_char_pos": 890, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "successfully provided valuable", "after": "was successful provided", "start_char_pos": 1095, "end_char_pos": 1125, "major_intent": "others", "raw_intents": ["clarity", "others", "others"]}, {"type": "R", "before": "undeveloped", "after": "developing", "start_char_pos": 1216, "end_char_pos": 1227, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 108, 276, 358, 477, 584, 755, 942, 1082, 1155, 1288, 1416], "domain": "wiki"}
{"doc_id": "51908", "revision_depth": 1, "before_revision": "Publicity and witness accounts Multiple news reports would have details of family, friends, and acquaintances calling him normal and without a propensity to hurt people, not believing the bombings were an \" anti-government \" message, as speculated by the magnitude of the attacks and Helder's manifesto. Helder also reportedly asked for a Bible to look through the contents of the pages. Helder was notably a major fan of Kurt Cobain , even after Cobain committed suicide, having posters and merchandise of Cobain and his band, Nirvana, in his room when he was arrested. A bandmate in Apathy remarked Helder had \"an interesting style of singing\" and would write much of the lyrics of the group's tracks, not having a full understanding of their meaning.", "after_revision": "Publicity and witness accounts News reports detailed family, friends, and acquaintances calling him normal and without a propensity to hurt people, disbelieving the bombings were an anti-government message. Though the same articles mentioned his roommate noting he had become obsessed with death and astral projection, and had begun telling them extended monologues about his new spiritual beliefs. Helder was notably a fan of Kurt Cobain ; merchandise of Cobain and his band, Nirvana, decorated his room when he was arrested. A bandmate in Apathy remarked Helder had \"an interesting style of singing\" and would write much of the lyrics of the group's tracks, not having a full understanding of their meaning.", "edit_actions": [{"type": "R", "before": "Multiple news reports would have details of", "after": "News reports detailed", "start_char_pos": 31, "end_char_pos": 74, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "not believing", "after": "disbelieving", "start_char_pos": 170, "end_char_pos": 183, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "D", "before": "\"", "after": null, "start_char_pos": 205, "end_char_pos": 206, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "\" message, as speculated by the magnitude of the attacks and Helder's manifesto. Helder also reportedly asked for a Bible to look through the contents of the pages.", "after": "message. Though the same articles mentioned his roommate noting he had become obsessed with death and astral projection, and had begun telling them extended monologues about his new spiritual beliefs.", "start_char_pos": 223, "end_char_pos": 387, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "major", "after": null, "start_char_pos": 409, "end_char_pos": 414, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": ", even after Cobain committed suicide, having posters and", "after": ";", "start_char_pos": 434, "end_char_pos": 491, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "in", "after": "decorated", "start_char_pos": 537, "end_char_pos": 539, "major_intent": "clarity", "raw_intents": ["others", "clarity", "clarity"]}], "sents_char_pos": [0, 303, 387, 570], "domain": "wiki"}
{"doc_id": "51908", "revision_depth": 2, "before_revision": "Writings and theories Helder performing with Apathy in his hometown of Pine Island, Minnesota While Helder was not at the top of his college class,  his teachers described him as a reasonably good, quiet and polite student, and at first there was confusion as to what the motivation for the bombings could be. Within the year prior to his arrest, Helder had become passionate about astral projection techniques and came to believe that death of the flesh and body is not the end of existence, as evidenced by the manifesto he sent to The Badger Herald of the University of WisconsinMadison at the beginning of the bombing spree. The essay also includes statements such as \"I'm taking very drastic measure in attempt to provide this information to you... I will die/change in the end for this, but that's ok, hahaha paradise awaits! I'm dismissing a few individuals from reality, to change all of you for the better\"Lagorio, Christine. \"News Analysis: Newspapers receipt of shady mail a daily occurrence.\" The Badger Herald May 10, 2002 Musical background Prior to his bombing spree, Helder was a member of Apathy, a Rochester, Minnesota, three-piece grunge band. Although the band was successful only on a local scale, it recorded a CD named Sacks of People at the end of their first summer together, which they funded and released themselves.  When the news broke that Helder was the pipe bomber, the media made significant mention of his status as a musician. Many music critics scrambled to acquire copies of Apathy's CD, some of which were auctioned on eBay for as much as $200. Publicity and witness accounts News reports detailed family, friends, and acquaintances calling him normal and without a propensity to hurt people, disbelieving the bombings were an anti-government message. Though the same articles mentioned his roommate noting he had become obsessed with death and astral projection, and had begun telling them extended monologues about his new spiritual beliefs. Helder was notably a fan of Kurt Cobain; merchandise of Cobain and his band, Nirvana, decorated his room when he was arrested. A bandmate in Apathy remarked Helder had \"an interesting style of singing\" and would write much of the lyrics of the group's tracks, not having a full understanding of their meaning.", "after_revision": "Writings and theories Helder performing with Apathy in his hometown of Pine Island, Minnesota Family, friends, and acquaintances referred to Helder as normal and without a propensity to hurt people, though in the same articles, his roommate noted he had become obsessed with death and astral projection, and had begun reciting extended monologues about his new spiritual beliefs to his friends. Helder was not at the top of his college class, but his teachers described him as a reasonably good, quiet and polite student, and at first there was confusion as to what the motivation for the bombings could be. Within the year prior to his arrest, Helder had become passionate about astral projection techniques and came to believe that death of the flesh and body is not the end of existence, as evidenced by the manifesto he sent to The Badger Herald of the University of WisconsinMadison at the beginning of the bombing spree. The essay also includes statements such as \"I'm taking very drastic measure in attempt to provide this information to you... I will die/change in the end for this, but that's ok, hahaha paradise awaits! I'm dismissing a few individuals from reality, to change all of you for the better\"Lagorio, Christine. \"News Analysis: Newspapers receipt of shady mail a daily occurrence.\" The Badger Herald May 10, 2002 Musical background Prior to his bombing spree, Helder was a member of Apathy, a Rochester, Minnesota, three-piece grunge band. Although the band was successful only on a local scale, it recorded a CD named Sacks of People at the end of their first summer together, which they funded and released themselves. Helder, a fan of Kurt Cobain, had heavily decorated his dorm room with Nirvana posters and ephemera. A bandmate in Apathy remarked Helder had \"an interesting style of singing\" and would write much of the lyrics of the group's tracks, not having a full understanding of their meaning. When the news broke that Helder was the pipe bomber, the media made significant mention of his status as a musician. Many music critics scrambled to acquire copies of Apathy's CD, some of which were auctioned on eBay for as much as $200. ", "edit_actions": [{"type": "R", "before": "While Helder", "after": "Family, friends, and acquaintances referred to Helder as normal and without a propensity to hurt people, though in the same articles, his roommate noted he had become obsessed with death and astral projection, and had begun reciting extended monologues about his new spiritual beliefs to his friends. Helder", "start_char_pos": 94, "end_char_pos": 106, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "A", "before": null, "after": "but", "start_char_pos": 148, "end_char_pos": 148, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "A", "before": null, "after": "Helder, a fan of Kurt Cobain, had heavily decorated his dorm room with Nirvana posters and ephemera. A bandmate in Apathy remarked Helder had \"an interesting style of singing\" and would write much of the lyrics of the group's tracks, not having a full understanding of their meaning.", "start_char_pos": 1344, "end_char_pos": 1344, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "D", "before": "Publicity and witness accounts News reports detailed family, friends, and acquaintances calling him normal and without a propensity to hurt people, disbelieving the bombings were an anti-government message. Though the same articles mentioned his roommate noting he had become obsessed with death and astral projection, and had begun telling them extended monologues about his new spiritual beliefs.", "after": null, "start_char_pos": 1583, "end_char_pos": 1981, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "Helder was notably a fan of Kurt Cobain; merchandise of Cobain and his band, Nirvana, decorated his room when he was arrested. A bandmate in Apathy remarked Helder had \"an interesting style of singing\" and would write much of the lyrics of the group's tracks, not having a full understanding of their meaning.", "after": null, "start_char_pos": 1982, "end_char_pos": 2291, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 309, 628, 753, 831, 934, 1004, 1162, 1343, 1461, 1582, 1789, 1981, 2022, 2108], "domain": "wiki"}
{"doc_id": "51908", "revision_depth": 3, "before_revision": "Personality, writings and theories Helder performing with Apathy in his hometown of Pine Island, Minnesota  Family, friends, and  acquaintances referred to Helder as normal and without a propensity to hurt people, though in the same articles, his roommate noted he had  become obsessed with death and astral projection, and had begun reciting extended monologues about his new spiritual beliefs to his friends . Helder was not at the top of his college class, but his teachers described him as a reasonably good, quiet and polite student, and at first there was confusion as to what the motivation for the bombings could be .", "after_revision": "Personality, writings and theories Helder performing with Apathy in his hometown of Pine Island, Minnesota Helder was not at the top of his college class, but his teachers described him as a reasonably good, quiet and polite student, and at first there was confusion as to what the motivation for the bombings could be. Family, friends, and many acquaintances referred to Helder as normal and without a propensity to hurt people, though his roommate noted he had recently become obsessed with death , had begun smoking marijuana, and had been reciting extended monologues about his new spiritual beliefs to his friends (which they had mostly \"laughed off\") .", "edit_actions": [{"type": "A", "before": null, "after": "Helder was not at the top of his college class, but his teachers described him as a reasonably good, quiet and polite student, and at first there was confusion as to what the motivation for the bombings could be.", "start_char_pos": 107, "end_char_pos": 107, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "A", "before": null, "after": "many", "start_char_pos": 129, "end_char_pos": 129, "major_intent": "style", "raw_intents": ["meaning-changed", "style", "style"]}, {"type": "D", "before": "in the same articles,", "after": null, "start_char_pos": 221, "end_char_pos": 242, "major_intent": "coherence", "raw_intents": ["coherence", "style", "coherence"]}, {"type": "A", "before": null, "after": "recently", "start_char_pos": 269, "end_char_pos": 269, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "and astral projection, and had begun", "after": ", had begun smoking marijuana, and had been", "start_char_pos": 297, "end_char_pos": 333, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ". Helder was not at the top of his college class, but his teachers described him as a reasonably good, quiet and polite student, and at first there was confusion as to what the motivation for the bombings could be", "after": "(which they had mostly \"laughed off\")", "start_char_pos": 410, "end_char_pos": 623, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 12, 96, 115, 213], "domain": "wiki"}
{"doc_id": "5233272", "revision_depth": 1, "before_revision": "History In the late 19th century, Portugal's dominance of Mozambique was threatened by the expansionist colonial ambitions of Great Britain and Germany . Although the borders of Mozambique had nominally been fixed by the Berlin Conference of 1884-1885, Portugal lacked the capital to exert effective control or  exploitation of the territory. To help overcome this, in 1891 the Portuguese government authorised three private companies by royal charter to manage portions of Mozambique: the Mozambique Company, the Zambezi Company and the Niassa Company. In 1913-14, a German banking consortium bought a majority of the shares of the Niassa Company, aiming at a partition of  Mozambique between Germany and Great Britain. With the outbreak of World War I , the British government confiscated  German equities and handed them over to an English financial group led by Owen Philipps, chairman of the Union-Castle Line, which did much business in Portuguese East Africa, but which found the concession to be more of a liability. During the war, the Niassa Company's territory was the scene of several  resistance operations by the local chiefs aided by the Germans (including in the Kionga Triangle). To resist this invasion, more than 300 kilometers of roads were opened. This also meant the effective occupation of the Mueda Plateau, which was occupied by the heavily-armed Makonde people. The Niassa Company only managed to suppress the Makonde by the early 1920s, and the tribe later became the backbone of the FRELIMO movement in the 1960s and 1970s against continued Portuguese colonial rule. In addition, the company relied on a hut tax , modelled after the system in British East Africa . Although theoretically a tax on each dwelling, in reality the tax was on each adult person, which could be paid in cash or in produce. The hut tax, which was two escudo in 1921 was increased to 50 escudo in 1927 and to 85 escudo in 1929. This was party due to devaluation of the escudo, but was also due to a dwindling tax base, as thousands of Mozambicans sought refuge in Tanganyika or Nyasaland to avoid the tax. As a tax of 50 escudo was the equivalent of three months paid labor, many people fell deeply into debt or forced labor.", "after_revision": "History In the late 19th century, Portugal's dominance of Mozambique began to be challenged by the establishment of neighboring British and German colonies . Although the borders of Mozambique had nominally been fixed by the Berlin Conference of 1884-1885, Portugal lacked the capital to exert effective control or economic exploitation of the territory. To help overcome this, in 1891 the Portuguese government authorised three private companies by royal charter to manage portions of Mozambique: the Mozambique Company, the Zambezi Company and the Niassa Company. In 1913-14, a German banking consortium bought a majority of the shares of the Niassa Company, aiming to bring about a partition of Portuguese Mozambique between Germany and Great Britain. With the outbreak of the First World War , the British government confiscated the German equities and handed them over to an English financial group led by Owen Philipps, chairman of the Union-Castle Line, which conducted several businesses in Portuguese East Africa, but which found the concession to be more of a liability. During the war, the Niassa Company's territory was the scene of several anti-Portuguese resistance operations by the local chiefs aided by the Germans (including in the Kionga Triangle). To resist this invasion, more than 300 kilometers of roads were opened. This also meant the effective occupation of the Mueda Plateau, which was occupied by the heavily-armed Makonde people. The Niassa Company only managed to suppress the Makonde by the early 1920s, and the tribe later became the backbone of the FRELIMO movement in the 1960s and 1970s against continued Portuguese colonial rule. In addition, the company relied on a hut tax to fund its operations . Although theoretically a tax on each dwelling, in reality the tax was on each adult person, which could be paid in cash or in produce. The hut tax, which was two escudo in 1921 was increased to 50 escudo in 1927 and to 85 escudo in 1929. This was party due to devaluation of the escudo, but was also due to a dwindling tax base, as thousands of Mozambicans sought refuge in Tanganyika or Nyasaland to avoid the tax. As a tax of 50 escudo was the equivalent of three months paid labor, many people fell deeply into debt or forced labor.", "edit_actions": [{"type": "R", "before": "was threatened by the expansionist colonial ambitions of Great Britain and Germany", "after": "began to be challenged by the establishment of neighboring British and German colonies", "start_char_pos": 69, "end_char_pos": 151, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "economic", "start_char_pos": 311, "end_char_pos": 311, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "at", "after": "to bring about", "start_char_pos": 656, "end_char_pos": 658, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Portuguese", "start_char_pos": 674, "end_char_pos": 674, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "World War I", "after": "the First World War", "start_char_pos": 742, "end_char_pos": 753, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 791, "end_char_pos": 791, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "did much business", "after": "conducted several businesses", "start_char_pos": 922, "end_char_pos": 939, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "anti-Portuguese", "start_char_pos": 1097, "end_char_pos": 1097, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", modelled after the system in British East Africa", "after": "to fund its operations", "start_char_pos": 1640, "end_char_pos": 1690, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 153, 342, 553, 720, 1024, 1196, 1268, 1387, 1594, 1692, 1827, 1930, 2108], "domain": "wiki"}
{"doc_id": "52589868", "revision_depth": 1, "before_revision": "To its north, Chile imposed its dominance by eliminating the threat posed by the union of Bolivia with Peru during the War of the Confederation (1836-1839), followed by the conquest of these countries' mineral-rich territories in the Atacama during the War of the Pacific (1879-1884)in the process leaving Bolivia landlocked and pillaging the Peruvian capital. To its south and east, Chile used military force and colonization to occupy Araucana , a still undefeated indigenous nation (1861-1883) and although Chile retained westernmost Patagonia and the Strait of Magellan with the treaty of 1881, it yielded to Argentina most of Patagonia and its access to the Atlantic to avoid the outbreak of war in multiple fronts during the War of the Pacific . Chile's expansionist drive practically culminated with Easter Island's annexation in 1888, but vestiges of it continued well into the 20th centurye.g., the country's claim over Antarctic territory in 1940.", "after_revision": "To its north, Chile imposed its dominance by eliminating the threat posed by the union of Bolivia with Peru during the War of the Confederation (1836-1839), followed by the conquest of these countries' mineral-rich territories in the Atacama during the War of the Pacific (1879-1884)in the process leaving Bolivia landlocked and pillaging the Peruvian capital. To its south and east, Chile used military force and colonization to occupy Araucana (1861-1883) and successfully dispute Argentine claims over westernmost Patagonia and the Strait of Magellanthe outbreak of war only narrowly avoided in multiple occasions . Chile's expansionist drive practically culminated with Easter Island's annexation in 1888, but vestiges of it continued well into the 20th centurye.g., the country's claim over Antarctic territory in 1940.", "edit_actions": [{"type": "D", "before": ", a still undefeated indigenous nation", "after": null, "start_char_pos": 446, "end_char_pos": 484, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "although Chile retained", "after": "successfully dispute Argentine claims over", "start_char_pos": 501, "end_char_pos": 524, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Magellan with the treaty of 1881, it yielded to Argentina most of Patagonia and its access to the Atlantic to avoid the", "after": "Magellanthe", "start_char_pos": 565, "end_char_pos": 684, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "style"]}, {"type": "R", "before": "in multiple fronts during the War of the Pacific", "after": "only narrowly avoided in multiple occasions", "start_char_pos": 701, "end_char_pos": 749, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}], "sents_char_pos": [0, 360, 751], "domain": "wiki"}
{"doc_id": "52589868", "revision_depth": 2, "before_revision": "Chile's Polynesian aspirations would again be awakened in the aftermath of the country's decisive victory against Peru in the War of the Pacific, which left the Chilean fleet as the dominant maritime force in the Pacific coast of the Americas. Valparaso had also become the most important port in the Pacific coast of South America, providing Chilean merchants with the capacity to find markets in the Pacific for its new mineral wealth acquired from the Atacama. During this period, the Chilean intellectual and politician Benjamn Vicua Mackenna (who served as senator in the National Congress from 1876 to 1885) was an influential voice in favor of Chilean expansionism into the Pacifiche considered that Spain's discoveries in the Pacific had been stolen by the British, and envisioned that Chile's duty was to create an empire in the Pacific that would reach Asia . In the context of this imperialist fervor is that, in 1886, Captain Policarpo Toro of the Chilean Navy proposed to his superiors the annexation of Easter Island; a proposal that attained support from President Jos Manuel Balmaceda because of the island's apparent strategic location and economic value. After Toro transferred the rights to the island's sheep ranching operations from Tahiti-based businesses to the Chilean-based Williamson-Balfour Company in 1887, Easter Island's annexation process was culminated with the signing of the \"Agreement of Wills\" between Rapa Nui chieftains and Toro, in name of the Chilean government, in 1888.See:", "after_revision": "Chile's Polynesian aspirations would again be awakened in the aftermath of the country's decisive victory against Peru in the War of the Pacific, which left the Chilean fleet as the dominant maritime force in the Pacific coast of the Americas. Valparaso had also become the most important port in the Pacific coast of South America, providing Chilean merchants with the capacity to find markets in the Pacific for its new mineral wealth acquired from the Atacama. During this period, the Chilean intellectual and politician Benjamn Vicua Mackenna (who served as senator in the National Congress from 1876 to 1885) was an influential voice in favor of Chilean expansionism into the Pacifiche considered that Spanish discoveries in the Pacific had been taken by the British, and envisioned that Chile's duty was to create an empire in the Pacific that would reach Asian shores . In the context of this imperialist fervor is that, in 1886, Captain Policarpo Toro of the Chilean Navy proposed to his superiors the annexation of Easter Island; a proposal that attained support from President Jos Manuel Balmaceda because of the island's apparent strategic location and economic value. After Toro transferred the rights to the island's sheep ranching operations from Tahiti-based businesses to the Chilean-based Williamson-Balfour Company in 1887, Easter Island's annexation process was culminated with the signing of the \"Agreement of Wills\" between Rapa Nui chieftains and Toro, in name of the Chilean government, in 1888.See:", "edit_actions": [{"type": "R", "before": "Spain's", "after": "Spanish", "start_char_pos": 707, "end_char_pos": 714, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "stolen", "after": "taken", "start_char_pos": 751, "end_char_pos": 757, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "Asia", "after": "Asian shores", "start_char_pos": 863, "end_char_pos": 867, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}], "sents_char_pos": [0, 243, 463, 869, 1031, 1172], "domain": "wiki"}
{"doc_id": "53199994", "revision_depth": 1, "before_revision": "James Robert Phillips (1863 - 1897) was the deputy commissioner and consul for the Niger Coast Protectorate. He is remembered for his part in the events that led to the Benin Expedition of 1897. In 1897  Phillips set out to depose the Oba of Benin, although his reasons for doing so remain unclear. He and many of his party were killed as they approached Benin City . Although Phillips had acted without authority , after his death a British punitive force deposed the Benin monarchy, pillaged Benin City and took many riches including the Benin Bronzes back to Britain . Aftermath The British Admiralty responded swiftly to the Benin Massacre by authorizing the Benin Punitive Expedition of February 1897. On 18 February, Benin City was taken by the British and the kingdom of Benin overthrown . The British Government revoked the Royal Niger Company's charter (granted in 1886) as a result of these events and ongoing complaints against the company from the Brass people of the Benin Rivers, the killings of Hausa princes during the Bida Battle ( 1897 ) and ongoing territorial rivalry with the French. Following the revoking of its charter on 31 December 1899, the Royal Niger Company sold its holdings to the British government for 865,000 (). On 1 January 1900 all its territories and assets passed to the British crown . The surrendered territories together with the Niger Coast Protectorate were formed into Northern and Southern Protectorates of the Niger River. In 1914, the two protectorates were formally united and amalgamated as the Colony and Protectorate of Nigeria by Governor Lord Frederick Lugard. Memorials An eagle lectern in Christ Church, Ivegill, was inscribed in his memory; it is now in Carlisle Cathedral. At the site of the massacre, a memorial cross was erected to the seven British killed.", "after_revision": "James Robert Phillips (1863 - 1897) was the deputy commissioner and consul for the Niger Coast Protectorate. He is remembered for his part in the events that led to the Benin Expedition of 1897. In 1897 , Phillips set out to depose the Oba of Benin, although his reasons for doing so remain unclear. He and his party was ambushed and slaughtered as they approached Benin City , with Philips being among the casualties. Though Philips had acted without consulting the Royal Niger Company authorities , after his death the British government dispatched an punitive force against the Benin monarchy, which the force defeated and deposed, leading to the kingdom's eventual absorption into colonial Nigeria . Aftermath The British Admiralty responded swiftly to the Benin Massacre by authorizing the punitive expedition which departed in February 1897. On 18 February, Benin City was captured by the expedition and the Kingdom of Benin was overthrown; it was eventually absorbed into colonial Nigeria . The British government revoked the Royal Niger Company's charter (granted in 1886) as a result of these events and ongoing complaints against the company from the Brass people of the Benin Rivers, the deaths of Hausa princes during skirmishes in the Bida Emirate in 1897 and ongoing territorial rivalry with the French. Following the revoking of its charter on 31 December 1899, the Royal Niger Company sold its holdings to the British government for 865,000 (). On 1 January 1900 all its territories and assets passed to the British Crown . The surrendered territories together with the Niger Coast Protectorate were formed into Northern and Southern Protectorates of the Niger River. In 1914, the two protectorates were formally united and amalgamated as the Colony and Protectorate of Nigeria by Governor Lord Frederick Lugard. Memorials An eagle lectern in Christ Church, Ivegill, was inscribed in his memory; it is now in Carlisle Cathedral. At the site of the massacre, a memorial cross was erected to the seven Britons killed.", "edit_actions": [{"type": "A", "before": null, "after": ",", "start_char_pos": 203, "end_char_pos": 203, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "many of his party were killed", "after": "his party was ambushed and slaughtered", "start_char_pos": 306, "end_char_pos": 335, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": ". Although Phillips", "after": ", with Philips being among the casualties. Though Philips", "start_char_pos": 366, "end_char_pos": 385, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "authority", "after": "consulting the Royal Niger Company authorities", "start_char_pos": 404, "end_char_pos": 413, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "a British punitive force deposed", "after": "the British government dispatched an punitive force against", "start_char_pos": 432, "end_char_pos": 464, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "pillaged Benin City and took many riches including the Benin Bronzes back to Britain", "after": "which the force defeated and deposed, leading to the kingdom's eventual absorption into colonial Nigeria", "start_char_pos": 485, "end_char_pos": 569, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Benin Punitive Expedition of", "after": "punitive expedition which departed in", "start_char_pos": 663, "end_char_pos": 691, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "taken by the British and the kingdom of Benin overthrown", "after": "captured by the expedition and the Kingdom of Benin was overthrown; it was eventually absorbed into colonial Nigeria", "start_char_pos": 738, "end_char_pos": 794, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Government", "after": "government", "start_char_pos": 809, "end_char_pos": 819, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "killings", "after": "deaths", "start_char_pos": 998, "end_char_pos": 1006, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the Bida Battle (", "after": "skirmishes in the Bida Emirate in", "start_char_pos": 1031, "end_char_pos": 1048, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "D", "before": ")", "after": null, "start_char_pos": 1054, "end_char_pos": 1055, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "crown", "after": "Crown", "start_char_pos": 1319, "end_char_pos": 1324, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "British", "after": "Britons", "start_char_pos": 1803, "end_char_pos": 1810, "major_intent": "style", "raw_intents": ["style", "clarity", "fluency"]}], "sents_char_pos": [0, 108, 194, 298, 367, 706, 796, 1104, 1247, 1326, 1470, 1615, 1698, 1731], "domain": "wiki"}
{"doc_id": "53199994", "revision_depth": 2, "before_revision": "James Robert Phillips (1863 - 1897) was the deputy commissioner and consul for the Niger Coast Protectorate. He is remembered for his part in the events that led to the Benin Expedition of 1897. In 1897, Phillips set out to depose the Oba of Benin, although his reasons for doing so remain unclear. He and his party was ambushed and slaughtered as they approached Benin City, with Philips being among the casualties. Though Philips had acted without consulting the Royal Niger Company authorities, after his death the British government dispatched an punitive force against the Benin monarchy, which the force defeated and deposed, leading to the kingdom's eventual absorption into colonial Nigeria. Life James Phillips was the eldest son of Rev. Thompson Phillips, vicar of Ivegill and later Archdeacon of Furness in the Diocese of Carlisle, and Eliza, daughter of General James Wallace Sleigh. The Benin massacre Boisragon and Locke, the two British men to survive the ambush Captain Phillips statue, Museum of Black Civilisations, Dakar On 17 December 1896, Phillips set off from Old Calabar in the yacht Ivy on the expedition to Benin. He also sent a message through Chief Dogho (Itsekiri Chief) to Ovonramwen, the 36th Oba of Benin, that he was on his way to visit. Although they had given word of their intended visit, they were later informed that their arrival must be delayed, because no foreigner could enter the city while rituals were being conducted;Benin Diplomatic Handbook, p. 21. however, the travellers continued on their expedition. His true intentions were not clear to Oba Ovonramwen, who replied asking him to wait for some days whilst he and his Council of Chiefs investigated claims that the whitemen are bringing war to Benin. He also informed Phillips that he was currently engaged in 'worshipping his father's head' in the Ague Festival. Phillips failed to heed the Oba's advice, as well as pleas from his Itsekiri advisors, and on 4 January 1897 he pressed ahead with his plan.", "after_revision": "James Robert Phillips (1863 - 1897) was the deputy commissioner and consul for the Niger Coast Protectorate. He is remembered for his part in the events that led to the Benin Expedition of 1897. In 1897, Phillips set out to depose the Oba of Benin, although his reasons for doing so remain unclear. He and his party was ambushed and slaughtered as they approached Benin City, with Phillips being among the casualties. Though Phillips had acted without consulting the Royal Niger Company authorities, after his death the British government dispatched an punitive expedition against the Benin monarchy, which the force defeated and deposed, leading to the kingdom's eventual absorption into colonial Nigeria. Life James Phillips was the eldest son of Reverend Thompson Phillips, vicar of Ivegill and later Archdeacon of Furness in the Diocese of Carlisle, and Eliza, daughter of General James Wallace Sleigh. The Benin massacre Boisragon and Locke, the two British men to survive the ambush On 17 December 1896, Phillips set off from Old Calabar in the yacht Ivy on the expedition to Benin. He also sent a message through Chief Dogho (Itsekiri Chief) to Ovonramwen, the 36th Oba of Benin, that he was on his way to visit. Although they had given word of their intended visit, they were later informed that their arrival must be delayed, because no foreigner could enter the city while rituals were being conducted;Benin Diplomatic Handbook, p. 21. however, the travellers continued on their expedition. His true intentions were not clear to Oba Ovonramwen, who replied asking him to wait for some days whilst he and his Council of Chiefs investigated claims that the whitemen are bringing war to Benin. He also informed Phillips that he was currently engaged in 'worshipping his father's head' in the Ague Festival. Phillips failed to heed the Oba's advice, as well as pleas from his Itsekiri advisors, and on 4 January 1897 he pressed ahead with his plan.", "edit_actions": [{"type": "R", "before": "Philips", "after": "Phillips", "start_char_pos": 381, "end_char_pos": 388, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Philips", "after": "Phillips", "start_char_pos": 424, "end_char_pos": 431, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "force", "after": "expedition", "start_char_pos": 560, "end_char_pos": 565, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Rev.", "after": "Reverend", "start_char_pos": 742, "end_char_pos": 746, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "D", "before": "Captain Phillips statue, Museum of Black Civilisations, Dakar", "after": null, "start_char_pos": 978, "end_char_pos": 1039, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 108, 194, 298, 416, 699, 895, 1139, 1270, 1551, 1751, 1864], "domain": "wiki"}
{"doc_id": "53635253", "revision_depth": 1, "before_revision": "Following the outbreak of the influenza epidemic in 1918 which devastated her area, Nkwenkwe believed that she was spared for a divine purpose. And in undertaking her work as first a seer, then a diviner, and ultimately a prophet, her life was changed dramatically. She began having visions and telling locals that had God told her the epidemic was punishment for people's sins and that her mission was to reform society. However, in the wake of the Bulhoek massacre where a 183 African Israelites were killed at Bulhoek in 1921, white attitudes to any large-scale black gatherings became increasingly paranoid. Officials reported that farm workers around Fort Beaufort (near Fort Hare) had been enraptured by her message and were reluctant to return to work. The government became paranoid of any independent black thinkers. Nkwenkwe  great-grandson Mzimkhulu Bungu believes that some established mission churches, worried about her growing following, had also complained about her activities. Nkwenkwe's sermons also encouraged unity among educated and \"red\" Xhosa people, something that was in conflict with the colonial system's divide and conquer mechanism. According to Edgar, R and Sapire, H, she was further accused of encouraging Africans to boycott white churches  . She was now seen as subversive, [ and was] arrested, and jailed in 1922. Nkwenkwe continued to preach despite the terms of her release, and she was re-arrested in April 1923. The arrest of Nkwenkwe enraged her supporters. In a show of solidarity, hundreds of her supporters gathered, ready to engage the authorities in the event that she was charged. Mindful of the disorder that could be unleashed by a possible court appearance, the authorities committed Nontetha to Fort Beaufort Mental Hospital. Her followers would walk as far as 80 km to the town to consult with her. In 1924, Nkwenkwe was transferred her to Weskoppies asylum in Pretoria . After two years of no correspondence about her condition, Nkwenkwe's supporters , decided to visit their leader in Pretoria. On November 23 1926, her follower began a 1000 km walk which lasted for 55 day walk from Eastern Cape to Pretoria. When they met her on January 18 1927, the movement had grown as some people joined them along the way. A second pilgrimage was cut short and loaded onto trains after crossing the Orange River at Aliwal North without passes  1930.", "after_revision": "Following the outbreak of the influenza epidemic in 1918 which devastated her area, Nkwenkwe believed that she was spared for a divine purpose. And in undertaking her work as first a seer, then a diviner, and ultimately a prophet, her life was changed dramatically. She began having visions and telling locals that God had told her the epidemic was punishment for people's sins and that her mission was to reform society. However, in the wake of the Bulhoek massacre where a 183 African Israelites were killed at Bulhoek in 1921, white attitudes to any large-scale black gatherings became increasingly paranoid. Officials reported that farm workers around Fort Beaufort (near Fort Hare) had been enraptured by her message and were reluctant to return to work. The government became suspicious of any independent black thinkers. Nkwenkwe 's great-grandson Mzimkhulu Bungu believes that some established mission churches, worried about her growing following, had also complained about her activities. Nkwenkwe's sermons also encouraged unity among educated and \"red\" Xhosa people, something that was in conflict with the colonial system's divide and conquer mechanism. According to Edgar, R and Sapire, H, she was further accused of encouraging Africans to boycott white churches ' . She was now seen as subversive, [ was] arrested, and jailed in 1922. Nkwenkwe continued to preach despite the terms of her release, and she was re-arrested in April 1923. The arrest of Nkwenkwe enraged her supporters. In a show of solidarity, hundreds of her supporters gathered, ready to engage the authorities in the event that she was charged. Mindful of the disorder that could be unleashed by a possible court appearance, the authorities committed Nontetha to Fort Beaufort Mental Hospital. Her followers would walk as far as 80 km to the town to consult with her. In 1924, Nkwenkwe was transferred to the Weskoppies Mental Hospital in Pretoria, the government's prime psychiatric observation institution . After two years of no correspondence about her condition, Nkwenkwe's supporters decided to visit their leader in Pretoria. On November 23 1926, her followers began a 1000 km walk which lasted for 55 days, walking from Eastern Cape to Pretoria. When they met her on January 18 1927, the movement had grown as some people joined them along the way. A second pilgrimage was cut short and the marchers loaded back onto trains after crossing the Orange River at Aliwal North without passes in 1930.", "edit_actions": [{"type": "R", "before": "had God", "after": "God had", "start_char_pos": 315, "end_char_pos": 322, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "paranoid", "after": "suspicious", "start_char_pos": 782, "end_char_pos": 790, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "'s", "start_char_pos": 835, "end_char_pos": 835, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "'", "start_char_pos": 1274, "end_char_pos": 1274, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "D", "before": "and", "after": null, "start_char_pos": 1311, "end_char_pos": 1314, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "R", "before": "her to Weskoppies asylum in Pretoria", "after": "to the Weskoppies Mental Hospital in Pretoria, the government's prime psychiatric observation institution", "start_char_pos": 1885, "end_char_pos": 1921, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 2004, "end_char_pos": 2005, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "follower", "after": "followers", "start_char_pos": 2074, "end_char_pos": 2082, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "day walk", "after": "days, walking", "start_char_pos": 2124, "end_char_pos": 2132, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": "loaded", "after": "the marchers loaded back", "start_char_pos": 2305, "end_char_pos": 2311, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "in", "start_char_pos": 2387, "end_char_pos": 2387, "major_intent": "fluency", "raw_intents": ["fluency", "coherence", "fluency"]}], "sents_char_pos": [0, 143, 265, 421, 611, 759, 825, 994, 1162, 1276, 1349, 1451, 1498, 1627, 1776, 1850, 1923, 2048, 2163, 2266], "domain": "wiki"}
{"doc_id": "54114327", "revision_depth": 2, "before_revision": "Comparison with American flag 205px|The Grand Union Flag, 1775 The flag of the East India Company is considered to have inspired the 1775 Grand Union Flag, the first flag of the United States, as the two flags were of the same design. This connection is attributed to numerous sources. Benjamin Franklin of Pennsylvania once gave a speech endorsing the adoption of the Company's flag by the United States as their national flag. He said to George Washington of Virginia, \"While the field of your flag must be new in the details of its design, it need not be entirely new in its elements. There is already in use a flag, I refer to the flag of the East India Company.\" This was a way of symbolising American loyalty to the Crown as well as the United States' aspirations to be self-governing, as was the East India Company. Some colonists also felt that the Company could be a powerful ally in the American War of Independence, as they shared similar aims and grievances against Crown tax policies. Colonists therefore flew the Company's flag, to endorse the Company. However, the theory that the Grand Union Flag was a direct descendant of the flag of the East India Company has been criticised as lacking written evidence. On the other hand, the resemblance is obvious, and a number of the Founding Fathers of the United States were aware of the East India Company 's activities and of their free administration of India under Company rule.", "after_revision": "Comparison with American flag 205px|The Grand Union Flag, 1775 The flag of the East India Company may or may not have have inspired the 1775 Grand Union Flag, the first flag of the United States, as the two flags were of the same design. This connection is attributed to numerous sources. Benjamin Franklin of Pennsylvania once gave a speech endorsing the adoption of the Company's flag by the United States as their national flag. He said to George Washington of Virginia, \"While the field of your flag must be new in the details of its design, it need not be entirely new in its elements. There is already in use a flag, I refer to the flag of the East India Company.\" This was a way of symbolising American loyalty to the Crown as well as the United States' aspirations to be self-governing, as was the East India Company. Some colonists also felt that the Company could be a powerful ally in the American War of Independence, as they shared similar aims and grievances against Crown tax policies. Colonists therefore flew the Company's flag, to endorse the Company. However, the theory that the Grand Union Flag was a direct descendant of the flag of the East India Company has been criticised as lacking written evidence. The British East India Company \u2019s flag ranged from nine to thirteen red and white stripes and was usually only flown when it was sailing in the Indian Ocean. The Continental Army flew the flag until 1777.", "edit_actions": [{"type": "R", "before": "is considered to have", "after": "may or may not have have", "start_char_pos": 98, "end_char_pos": 119, "major_intent": "style", "raw_intents": ["style", "style", "style"]}, {"type": "R", "before": "On the other hand, the resemblance is obvious, and a number of the Founding Fathers of the United States were aware of the", "after": "The British", "start_char_pos": 1224, "end_char_pos": 1346, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "'s activities and of their free administration of India under Company rule.", "after": "\u2019s flag ranged from nine to thirteen red and white stripes and was usually only flown when it was sailing in the Indian Ocean. The Continental Army flew the flag until 1777.", "start_char_pos": 1366, "end_char_pos": 1441, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 234, 285, 428, 587, 667, 822, 997, 1066, 1223], "domain": "wiki"}
{"doc_id": "54114327", "revision_depth": 3, "before_revision": "Comparison with American flag 205px|The Grand Union Flag, 1775 The flag of the East India Company may or may not have have inspired the 1775 Grand Union Flag, the first flag of the United States, as the two flags were of the same design. This connection is attributed to numerous sources. Benjamin Franklin of Pennsylvania once gave a speech endorsing the adoption of the Company's flag by the United States as their national flag. He said to George Washington of Virginia, \"While the field of your flag must be new in the details of its design, it need not be entirely new in its elements. There is already in use a flag, I refer to the flag of the East India Company.\" This was a way of symbolising American loyalty to the Crown as well as the United States' aspirations to be self-governing, as was the East India Company. Some colonists also felt that the Company could be a powerful ally in the American War of Independence, as they shared similar aims and grievances against Crown tax policies. Colonists therefore flew the Company's flag, to endorse the Company. However, the theory that the Grand Union Flag was a direct descendant of the flag of the East India Company has been criticised as lacking written evidence. The British East India Company \u2019s flag ranged from nine to thirteen red and white stripes and was usually only flown when it was sailing in the Indian Ocean. The Continental Army flew the flag until 1777.", "after_revision": "Comparison with American flag 205px|The Grand Union Flag, 1775 The flag of the East India Company is considered to have inspired the 1775 Grand Union Flag, the first flag of the United States, as the two flags were of the same design. This connection is attributed to numerous sources. Benjamin Franklin of Pennsylvania once gave a speech endorsing the adoption of the Company's flag by the United States as their national flag. He said to George Washington of Virginia, \"While the field of your flag must be new in the details of its design, it need not be entirely new in its elements. There is already in use a flag, I refer to the flag of the East India Company.\" This was a way of symbolising American loyalty to the Crown as well as the United States' aspirations to be self-governing, as was the East India Company. Some colonists also felt that the Company could be a powerful ally in the American War of Independence, as they shared similar aims and grievances against Crown tax policies. Colonists therefore flew the Company's flag, to endorse the Company. However, the theory that the Grand Union Flag was a direct descendant of the flag of the East India Company has been criticised as lacking written evidence. On the other hand, the resemblance is obvious, and a number of the Founding Fathers of the United States were aware of the East India Company 's activities and of their free administration of India under Company rule.", "edit_actions": [{"type": "R", "before": "may or may not have have", "after": "is considered to have", "start_char_pos": 98, "end_char_pos": 122, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "The British", "after": "On the other hand, the resemblance is obvious, and a number of the Founding Fathers of the United States were aware of the", "start_char_pos": 1227, "end_char_pos": 1238, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "\u2019s flag ranged from nine to thirteen red and white stripes and was usually only flown when it was sailing in the Indian Ocean. The Continental Army flew the flag until 1777.", "after": "'s activities and of their free administration of India under Company rule.", "start_char_pos": 1258, "end_char_pos": 1431, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 237, 288, 431, 590, 670, 825, 1000, 1069, 1226, 1384], "domain": "wiki"}
{"doc_id": "54247804", "revision_depth": 1, "before_revision": "Fashion activism is the practice of the cultural capital of fashion as a medium for social and environmental change. The term was coined by Cline Semaan, co-founder of the Slow Factory Foundation . As with other forms of activism, the aim is to promote, impede, direct, or intervene into social arrangements of dress to lay claim to a certain political agenda as well as influence systemic change within the fashion industry . It merges popular styles of dress, from clothing and shoes, to headwear and accessories, with efforts to implement social and political change beyond the designated channels of influence offered by the local political system, such as voting. Fashion activism can be used as a form of protest, whether expressing dissent or support. Slow factory Beyond the use amongst designers promoting sustainability and user engagement, the term has also more recently been coined by designer Cline Semaan in her work with the NYC-based fashion lab Slow Factory. In this context, items in her collections have been used as means of fashion activism; a 'Dignity Key' necklace with which people can show support for displaced Middle Eastern refugees, a 'Banned' scarf showing the universal impact of President Trump's Muslim ban, and a '1st Amendment Flight Jacket' collaboration with ACLU, featuring the First Amendment text is written in Arabic, standing up to the rise in islamophobia in the United States and hate crimes against American Muslims.", "after_revision": "Fashion activism is the practice of using fashion as a means of social change, sometimes through ethical consumerism, but often beyond the act of purchasing symbolic fashionable objects . As with other forms of activism, the aim is to promote, impede, direct, or intervene into social arrangements of dress to lay claim to a certain political agenda . It merges popular styles of dress, from clothing and shoes, to headwear and accessories, with efforts to implement social and political change beyond the designated channels of influence offered by the local political system, such as voting. Fashion activism can be used as a form of protest, whether expressing dissent or support. Slow factory Beyond the use amongst designers promoting sustainability and user engagement, the term has also more recently been used by designer Cline Semaan in her work with the NYC-based fashion lab Slow Factory. In this context, items in her collections have been used as means of fashion activism; a 'Dignity Key' necklace with which people can show support for displaced Middle Eastern refugees, a 'Banned' scarf showing the universal impact of President Trump's Muslim ban, and a '1st Amendment Flight Jacket' collaboration with ACLU, featuring the First Amendment text is written in Arabic, standing up to the rise in islamophobia in the United States and hate crimes against American Muslims.", "edit_actions": [{"type": "R", "before": "the cultural capital of", "after": "using", "start_char_pos": 36, "end_char_pos": 59, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "medium for social and environmental change. The term was coined by Cline Semaan, co-founder of the Slow Factory Foundation", "after": "means of social change, sometimes through ethical consumerism, but often beyond the act of purchasing symbolic fashionable objects", "start_char_pos": 73, "end_char_pos": 195, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "D", "before": "as well as influence systemic change within the fashion industry", "after": null, "start_char_pos": 360, "end_char_pos": 424, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "coined", "after": "used", "start_char_pos": 888, "end_char_pos": 894, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 116, 197, 426, 668, 758, 976, 1063], "domain": "wiki"}
{"doc_id": "55002506", "revision_depth": 1, "before_revision": "Deepak Mathur (born 8 April 1952) is an Indian molecular and atomic physicist and , until recently, a distinguished professor at the Tata Institute of Fundamental Research. He is currently the J C Bose National Fellow at the Department of Atomic and Molecular Physics at Manipal University . Known for his research on molecular and biological physics, Mathur is an elected fellow of the Indian Academy of Sciences, Indian National Science Academy and The World Academy of Sciences. The Council of Scientific and Industrial Research, the apex agency of the Government of India for scientific research, awarded him the Shanti Swarup Bhatnagar Prize for Science and Technology, one of the highest Indian science awards, for his contributions to physical sciences in 1991.", "after_revision": "Deepak Mathur (born 8 April 1952) is an Indian molecular and atomic physicist and was a Distinguished Professor at the Tata Institute of Fundamental Research. He has been the J C Bose National Fellow at the Department of Atomic and Molecular Physics at Manipal Academy of Higher Education and Founding Director of the UM-DAE Centre for Excellence in Basic Science at the University of Mumbai . Known for his research on molecular and biological physics, Mathur is an elected fellow of the Indian Academy of Sciences, Indian National Science Academy and The World Academy of Sciences. The Council of Scientific and Industrial Research, the apex agency of the Government of India for scientific research, awarded him the Shanti Swarup Bhatnagar Prize for Science and Technology, one of the highest Indian science awards, for his contributions to physical sciences in 1991.", "edit_actions": [{"type": "R", "before": ", until recently, a distinguished professor", "after": "was a Distinguished Professor", "start_char_pos": 82, "end_char_pos": 125, "major_intent": "style", "raw_intents": ["style", "others", "clarity"]}, {"type": "R", "before": "is currently", "after": "has been", "start_char_pos": 176, "end_char_pos": 188, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": "University", "after": "Academy of Higher Education and Founding Director of the UM-DAE Centre for Excellence in Basic Science at the University of Mumbai", "start_char_pos": 279, "end_char_pos": 289, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 172, 481], "domain": "wiki"}
{"doc_id": "5525848", "revision_depth": 1, "before_revision": "Development and dissolution Invested with considerable powers, Santa Cruz endeavoured to establish in Peru the same type of authoritarian order he had imposed in Bolivia. He issued a civil code, a penal code, a trade regulation, a customs regulation and URLanized tax collectionprocedures allowing an increase in state revenues while restraining expenditures. However, the Confederation generated resistances among several groups in both countries , which resented the dilution of national identities , and also among neighbouring countries. An important number of Peruvian politicians opposed to the idea of the Confederation fled to Chile, where they received support and this led to the War of the Confederation . Chile declared war on December 28, 1836, and Argentina followed suit on May 9, 1837. The Chilean military expedition against Santa Cruz, led by Admiral Manuel Blanco Encalada  failed and had to submit to the signature of the Treaty of Paucarpata, on November 17, 1837. The Chilean government URLanized a second expedition, which defeated the Supreme Protector at the Battle of Yungay on January 20, 1839 , and forced the dissolution of the Confederation. When Agustn Gamarra took office as the new president of Peru on August 25, 1839, he officially declared the dissolution of the Confederation and of the merging of the Northern and Southern Peruvian Republics into a single state to be called again Peru and separate from Bolivia.", "after_revision": "Development and dissolution Invested with considerable powers, Santa Cruz endeavoured to establish in Peru the same type of authoritarian order he had imposed in Bolivia. He issued a civil code, a penal code, a trade regulation, a customs regulation and procedures for URLanizing tax collection, which allowed an increase in state revenues and reduced expenditures. However, the Confederation generated resistances among several groups in both countries that resented the dilution of national identities and by neighbouring countries. Many Peruvian politicians opposed to the idea of the Confederation fled to Chile, where they received support and this led to the War of the Confederation Chile declared war on December 28, 1836, and Argentina followed suit on May 9, 1837. The Chilean military expedition against Santa Cruz, led by Admiral Manuel Blanco Encalada , failed and had to submit to the signature of the Treaty of Paucarpata, on November 17, 1837. The Chilean government URLanized a second expedition, which defeated the Supreme Protector at the Battle of Yungay on January 20, 1839 and forced the dissolution of the Confederation. When Agustn Gamarra took office as the new president of Peru on August 25, 1839, he officially declared the dissolution of the Confederation and of the merging of the Northern and Southern Peruvian Republics into a single state , again called Peru, and separate from Bolivia.", "edit_actions": [{"type": "R", "before": "URLanized tax collectionprocedures allowing", "after": "procedures for URLanizing tax collection, which allowed", "start_char_pos": 254, "end_char_pos": 297, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "while restraining", "after": "and reduced", "start_char_pos": 328, "end_char_pos": 345, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": ", which", "after": "that", "start_char_pos": 448, "end_char_pos": 455, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": ", and also among", "after": "and by", "start_char_pos": 501, "end_char_pos": 517, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "An important number of", "after": "Many", "start_char_pos": 542, "end_char_pos": 564, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 715, "end_char_pos": 716, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 892, "end_char_pos": 892, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ",", "after": null, "start_char_pos": 1121, "end_char_pos": 1122, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "to be called again Peru", "after": ", again called Peru,", "start_char_pos": 1400, "end_char_pos": 1423, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 170, 359, 541, 716, 801, 985, 1171], "domain": "wiki"}
{"doc_id": "55503854", "revision_depth": 1, "before_revision": "Relation to traditional epistemology Traditional epistemology and Bayesian epistemology are both forms of epistemology, but they differ in various respects, for example, concerning their methodology, their interpretation of belief , the role justification or confirmation plays in them and some of their research interests. Traditional epistemology focuses on topics such as the analysis of the nature of knowledge, usually in terms of justified true beliefs , the sources of knowledge , like perception or testimony, the structure of a body of knowledge, for example in the form of foundationalismor coherentism, and the problem of philosophical skepticism or the question of whether knowledge is possible at all . These inquiries are usually based on epistemic intuitions and regard beliefs as either present or absent . Bayesian epistemology, on the other hand, works by formalizing concepts and problems, which are often vague in the traditional approach. It thereby focuses more on mathematical intuitions and promises a higher degree of precision. It sees belief as a continuous phenomenon that comes in various degrees, so-called credences. Some Bayesians have even suggested that the regular notion of belief should be abandoned. But there are also proposals to connect the two, for example, the Lockean thesis, which defines belief as credence above a certain threshold. Justification plays a central role in traditional epistemology while Bayesians have focused on the related notions of confirmation and disconfirmation through evidence. The notion of evidence is important for both approaches but only the traditional approach has been interested in studying the sources of evidence, like perception and memory. Bayesianism, on the other hand, has focused on the role of evidence for rationality: how someone's credence should be adjusted upon receiving new evidence. There is an analogy between the Bayesian norms of rationality in terms of probabilistic laws and the traditional norms of rationality in terms of deductive consistency. Certain traditional problems, like the topic of skepticism about our knowledge of the external world, are difficult to express in Bayesian terms.", "after_revision": "Relation to traditional epistemology Traditional epistemology and Bayesian epistemology are both forms of epistemology, but they differ in various respects, for example, concerning their methodology, the propositional attitudes they study , the role justification or confirmation plays in them and some of their research interests. Bayesian epistemology focuses on degrees of belief, also often called credences. On the other hand, Traditional epistemology focuses on knowledge, analyzed by Plato as being justified true beliefs . They then seek to understand what counts as knowledge and how to evaluate one's beliefs (see foundationalism, coherentism, and the problem of philosophical skepticism or the question of whether knowledge is possible at all ) . Bayesian epistemology, on the other hand, works by formalizing concepts and problems, which are often vague in the traditional approach. It thereby focuses more on mathematical intuitions and promises a higher degree of precision. It sees belief as a continuous phenomenon that comes in various degrees, so-called credences. Some Bayesians have even suggested that the regular notion of belief should be abandoned. But there are also proposals to connect the two, for example, the Lockean thesis, which defines belief as credence above a certain threshold. Justification plays a central role in traditional epistemology while Bayesians have focused on the related notions of confirmation and disconfirmation through evidence. The notion of evidence is important for both approaches but only the traditional approach has been interested in studying the sources of evidence, like perception and memory. Bayesianism, on the other hand, has focused on the role of evidence for rationality: how someone's credence should be adjusted upon receiving new evidence. There is an analogy between the Bayesian norms of rationality in terms of probabilistic laws and the traditional norms of rationality in terms of deductive consistency. Certain traditional problems, like the topic of skepticism about our knowledge of the external world, are difficult to express in Bayesian terms.", "edit_actions": [{"type": "R", "before": "their interpretation of belief", "after": "the propositional attitudes they study", "start_char_pos": 200, "end_char_pos": 230, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "Traditional", "after": "Bayesian", "start_char_pos": 324, "end_char_pos": 335, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "topics such as the analysis of the nature of knowledge, usually in terms of", "after": "degrees of belief, also often called credences. On the other hand, Traditional epistemology focuses on knowledge, analyzed by Plato as being", "start_char_pos": 360, "end_char_pos": 435, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", the sources of knowledge , like perception or testimony, the structure of a body of knowledge, for example in the form of foundationalismor", "after": ". They then seek to understand what counts as knowledge and how to evaluate one's beliefs (see foundationalism,", "start_char_pos": 459, "end_char_pos": 600, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ". These inquiries are usually based on epistemic intuitions and regard beliefs as either present or absent", "after": ")", "start_char_pos": 714, "end_char_pos": 820, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 323, 715, 959, 1053, 1147, 1237, 1379, 1548, 1723, 1879, 2048], "domain": "wiki"}
{"doc_id": "55503854", "revision_depth": 2, "before_revision": "Relation to traditional epistemology Traditional epistemology and Bayesian epistemology are both forms of epistemology, but they differ in various respects, for example, concerning their methodology, the propositional attitudes they study , the role justification or confirmation plays in them and some of their research interests. Bayesian epistemology focuses on degrees of belief, also often called credences. On the other hand, Traditional epistemology focuses on knowledge, analyzed by Plato as being justified true beliefs . They then seek to understand what counts as knowledgeand how to evaluate one's beliefs (see foundationalism, coherentism, and the problem of philosophical skepticism or the question of whether knowledge is possible at all ) . Bayesian epistemology, on the other hand, works by formalizing concepts and problems, which are often vague in the traditional approach. It thereby focuses more on mathematical intuitions and promises a higher degree of precision. It sees belief as a continuous phenomenon that comes in various degrees, so-called credences. Some Bayesians have even suggested that the regular notion of belief should be abandoned. But there are also proposals to connect the two, for example, the Lockean thesis, which defines belief as credence above a certain threshold. Justification plays a central role in traditional epistemology while Bayesians have focused on the related notions of confirmation and disconfirmation through evidence. The notion of evidence is important for both approaches but only the traditional approach has been interested in studying the sources of evidence, like perception and memory. Bayesianism, on the other hand, has focused on the role of evidence for rationality: how someone's credence should be adjusted upon receiving new evidence. There is an analogy between the Bayesian norms of rationality in terms of probabilistic laws and the traditional norms of rationality in terms of deductive consistency. Certain traditional problems, like the topic of skepticism about our knowledge of the external world, are difficult to express in Bayesian terms.", "after_revision": "Relation to traditional epistemology Traditional epistemology and Bayesian epistemology are both forms of epistemology, but they differ in various respects, for example, concerning their methodology, their interpretation of belief , the role justification or confirmation plays in them and some of their research interests. Traditional epistemology focuses on topics such as the analysis of the nature of knowledge, usually in terms of justified true beliefs , the sources of knowledge, like perception or testimony, the structure of a body of knowledge, for example in the form of foundationalism or coherentism, and the problem of philosophical skepticism or the question of whether knowledge is possible at all . These inquiries are usually based on epistemic intuitions and regard beliefs as either present or absent . Bayesian epistemology, on the other hand, works by formalizing concepts and problems, which are often vague in the traditional approach. It thereby focuses more on mathematical intuitions and promises a higher degree of precision. It sees belief as a continuous phenomenon that comes in various degrees, so-called credences. Some Bayesians have even suggested that the regular notion of belief should be abandoned. But there are also proposals to connect the two, for example, the Lockean thesis, which defines belief as credence above a certain threshold. Justification plays a central role in traditional epistemology while Bayesians have focused on the related notions of confirmation and disconfirmation through evidence. The notion of evidence is important for both approaches but only the traditional approach has been interested in studying the sources of evidence, like perception and memory. Bayesianism, on the other hand, has focused on the role of evidence for rationality: how someone's credence should be adjusted upon receiving new evidence. There is an analogy between the Bayesian norms of rationality in terms of probabilistic laws and the traditional norms of rationality in terms of deductive consistency. Certain traditional problems, like the topic of skepticism about our knowledge of the external world, are difficult to express in Bayesian terms.", "edit_actions": [{"type": "R", "before": "the propositional attitudes they study", "after": "their interpretation of belief", "start_char_pos": 200, "end_char_pos": 238, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Bayesian", "after": "Traditional", "start_char_pos": 332, "end_char_pos": 340, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "degrees of belief, also often called credences. On the other hand, Traditional epistemology focuses on knowledge, analyzed by Plato as being", "after": "topics such as the analysis of the nature of knowledge, usually in terms of", "start_char_pos": 365, "end_char_pos": 505, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": ". They then seek to understand what counts as knowledgeand how to evaluate one's beliefs (see foundationalism,", "after": ", the sources of knowledge, like perception or testimony, the structure of a body of knowledge, for example in the form of foundationalism or", "start_char_pos": 529, "end_char_pos": 639, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "R", "before": ")", "after": ". These inquiries are usually based on epistemic intuitions and regard beliefs as either present or absent", "start_char_pos": 753, "end_char_pos": 754, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 331, 412, 530, 893, 987, 1081, 1171, 1313, 1482, 1657, 1813, 1982], "domain": "wiki"}
{"doc_id": "55964497", "revision_depth": 1, "before_revision": "Muweilah () is an archaeological site in Sharjah, United Arab Emirates , in the suburb of Al Jurainah near Sharjah University City. A large, fortified settlement thought to have been occupied during the Iron Age II period (1,100-600BC), the site has been explored by archaeologists since the discovery of pottery shards by a local resident led to a French survey of the area in 1989 and archaeological work by an Australian expedition in 1994 It has yielded the oldest known example of writing found to date in the UAE, a pottery shard with an inscription, thought to be Sabean, with the letters 'bml'. Foundation Carbon dating artefacts found at Muweilah puts the settlement's original date of establishment at between 850 and 800 BC and  it enjoyed a brief heyday before being destroyed in a fire around 600BC. Constructed in the main from interlocked mud bricks and mud/stone brick walls, the walled settlement itself surrounds a large walled enclosure with seven buildings, thought to have provided living quarters as well as an administrative centre. This central building contained at least twenty columns and has been a rich trove for archaeologists, with extensive finds of painted and spouted vessels, iron weapons and hundreds of bronze pieces. Analysis of bioarchaeological remains has uncovered important information on subsistence strategies at Muweilah, largely linked to the role played by the domestication of the camel, with marine resources supplementing subsistence based on animal husbandry. The settlement at Muweilah is thought to have flourished through trade with inland settlements such as Masafi that made use of falaj irrigation to build agricultural resources.", "after_revision": "Muweilah () is an archaeological site in Sharjah, United Arab Emirates . it is located in what is now the suburb of Al Jurainah near Sharjah University City. A large, fortified settlement thought to have been occupied during the Iron Age II period (1,100-600BC), the site has been explored by archaeologists since the discovery of pottery shards by a local resident led to a French survey of the area in 1989 and archaeological work by an Australian expedition in 1994 It has yielded the oldest known example of writing found to date in the UAE, a pottery shard with an inscription, thought to be Sabean, with the letters 'bml'. Foundation Carbon dated artefacts found at Muweilah put the settlement's original date of establishment at between 850 and 800 BC and archaeological analysis has shown it enjoyed a brief heyday before being destroyed in a fire around 600BC. Constructed in the main from interlocked mud bricks and mud/stone brick walls, the walled settlement itself surrounds a large walled enclosure with seven buildings, thought to have provided living quarters as well as an administrative centre. This central building contained at least twenty columns and has been a rich trove for archaeologists, with extensive finds of painted and spouted vessels, iron weapons and hundreds of bronze pieces. Bio-archaeological analysis of remains has uncovered important information on subsistence strategies at Muweilah, largely linked to the role played by the domestication of the camel, with marine resources supplementing subsistence based on animal husbandry. The settlement at Muweilah is thought to have flourished through trade with inland settlements such as Masafi that made use of falaj irrigation to build agricultural resources.", "edit_actions": [{"type": "R", "before": ", in", "after": ". it is located in what is now", "start_char_pos": 71, "end_char_pos": 75, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "R", "before": "dating", "after": "dated", "start_char_pos": 621, "end_char_pos": 627, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "puts", "after": "put", "start_char_pos": 656, "end_char_pos": 660, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "archaeological analysis has shown", "start_char_pos": 739, "end_char_pos": 739, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Analysis of bioarchaeological", "after": "Bio-archaeological analysis of", "start_char_pos": 1255, "end_char_pos": 1284, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 131, 602, 812, 1055, 1254, 1511], "domain": "wiki"}
{"doc_id": "560813", "revision_depth": 1, "before_revision": "In its purest form, Civil libertarianism is a political philosophy supporting civil liberties, or which emphasize individual rights and personal freedoms over authorities - such as governments , a corporation and/or social peer pressures . Civil libertarianism is not limited to an individual ideology, rather, it is a philosophy based on a strict interpretation of the Constitution's civil liberties and the many civil rightslaws enacted since . In the libertarian movement The primary concern of most Civil Libertarians is the government's reach over the individual. In theory, the civil libertarian seeks to reduce the role of government to a minimum - preferring private businesses to fill the many social needs for most Americans. Under the control of elected officials, the government is only relied upon for regulation, without excessively interfering in the lives of its citizens. The Civil Libertarian sees the government's role as critical - supporting free enterprise and the privatization of government programs wherever improvements are possible. One of the central pillars of Civil Libertarianism is free speech. In many cases Civil Libertarians oppose bans on hate speech and obscenity. Although they may or may not personally condone behaviors associated with these issues, civil libertarians hold that the advantages of unfettered public discourse outweigh all disadvantages. With the advent of personal computers, the Internet, email, cell phones and other information technology advances a subset of civil libertarianism has arisen that focuses on protecting individuals' digital rights and privacy . Just as acting Democrats and Republicans hold varying degrees of left and right-leaning beliefs; few Civil Libertarians, if any, hold all of the views defined by this philosophy .", "after_revision": " Civil libertarianism is a strain of political thought that supports civil liberties, or which emphasizes the supremacy of individual rights and personal freedoms over and against any kind of authority (such as a state , a corporation , social norms imposed through peer pressure and so on) . Civil libertarianism is not a complete ideology\u2014 rather, it is a collection of views on the specific issues of civil liberties and civil rights . In the libertarian movement In the domain of libertarian philosophy, the primary concern of the civil libertarian is the relationship of the government to the individual. In theory, the civil libertarian seeks to restrict this relationship to an absolute minimum in which the state can function and provide basic services and securities without excessively interfering in the lives of its citizens. One key cause of civil libertarianism is upholding free speech. Specifically, civil libertarians oppose bans on hate speech and obscenity. Although they may or may not personally condone behaviors associated with these issues, civil libertarians hold that the advantages of unfettered public discourse outweigh all disadvantages. With the advent of personal computers, the Internet, email, cell phones and other information technology advances a subset of civil libertarianism has arisen that focuses on protecting individuals' digital rights and privacy .", "edit_actions": [{"type": "D", "before": "In its purest form,", "after": null, "start_char_pos": 0, "end_char_pos": 19, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "political philosophy supporting", "after": "strain of political thought that supports", "start_char_pos": 46, "end_char_pos": 77, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "emphasize", "after": "emphasizes the supremacy of", "start_char_pos": 104, "end_char_pos": 113, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "authorities - such as governments", "after": "and against any kind of authority (such as a state", "start_char_pos": 159, "end_char_pos": 192, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "and/or social peer pressures", "after": ", social norms imposed through peer pressure and so on)", "start_char_pos": 209, "end_char_pos": 237, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "limited to an individual ideology,", "after": "a complete ideology\u2014", "start_char_pos": 268, "end_char_pos": 302, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "philosophy based on a strict interpretation of the Constitution's", "after": "collection of views on the specific issues of", "start_char_pos": 319, "end_char_pos": 384, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "the many civil rightslaws enacted since", "after": "civil rights", "start_char_pos": 405, "end_char_pos": 444, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "The", "after": "In the domain of libertarian philosophy, the", "start_char_pos": 475, "end_char_pos": 478, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "most Civil Libertarians is the government's reach over", "after": "the civil libertarian is the relationship of the government to", "start_char_pos": 498, "end_char_pos": 552, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": "reduce the role of government to a minimum - preferring private businesses to fill the many social needs for most Americans. Under the control of elected officials, the government is only relied upon for regulation,", "after": "restrict this relationship to an absolute minimum in which the state can function and provide basic services and securities", "start_char_pos": 611, "end_char_pos": 826, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "The Civil Libertarian sees the government's role as critical - supporting free enterprise and the privatization of government programs wherever improvements are possible. One of the central pillars of Civil Libertarianism is", "after": "One key cause of civil libertarianism is upholding", "start_char_pos": 889, "end_char_pos": 1113, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "In many cases Civil Libertarians", "after": "Specifically, civil libertarians", "start_char_pos": 1127, "end_char_pos": 1159, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 1618, "end_char_pos": 1619, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "D", "before": "Just as acting Democrats and Republicans hold varying degrees of left and right-leaning beliefs; few Civil Libertarians, if any, hold all of the views defined by this philosophy", "after": null, "start_char_pos": 1620, "end_char_pos": 1797, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 446, 568, 735, 888, 1059, 1126, 1201, 1392, 1716], "domain": "wiki"}
{"doc_id": "56153035", "revision_depth": 1, "before_revision": "upTrujillo in 1952 In May 1961, the ruler of the Dominican Republic, Rafael Trujillo was murdered with weapons supplied by the United States Central Intelligence Agency (CIA). An internal CIA memorandum states that a 1973 Office of Inspector General investigation into the murder disclosed \"quite extensive Agency involvement with the plotters.\" The CIA described its role in \"changing\" the government of the Dominican Republic as a 'success' in that it assisted in moving the Dominican Republic from a totalitarian dictatorship to a Western-style democracy.\"CIA \"Family Jewels\" Memo, 1973 (see page 434) Family Jewels (Central Intelligence Agency)  Juan Bosch, an earlier recipient of CIA funding , was elected president of the Dominican Republic in 1962, and was deposed in 1963.", "after_revision": "upTrujillo in 1952 In May 1961, the ruler of the Dominican Republic, right-wing dictator Rafael Trujillo, was murdered with weapons supplied by the United States Central Intelligence Agency (CIA). An internal CIA memorandum states that a 1973 Office of Inspector General investigation into the murder disclosed \"quite extensive Agency involvement with the plotters.\" The CIA described its role in \"changing\" the government of the Dominican Republic as a 'success' in that it assisted in moving the Dominican Republic from a totalitarian dictatorship to a Western-style democracy.\"CIA \"Family Jewels\" Memo, 1973 (see page 434) Family Jewels (Central Intelligence Agency) Socialist Juan Bosch, whose propoganda and institute for political training had recieved some CIA funding via the J.M. Kaplan Fund , was elected president of the Dominican Republic in its first free elections, in December 1962. Bosch was deposed by a right-wing coup in September 1963, and President Johnson sent troops to prevent him from being restored to power.", "edit_actions": [{"type": "R", "before": "Rafael Trujillo", "after": "right-wing dictator Rafael Trujillo,", "start_char_pos": 69, "end_char_pos": 84, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Socialist", "start_char_pos": 649, "end_char_pos": 649, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "an earlier recipient of CIA funding", "after": "whose propoganda and institute for political training had recieved some CIA funding via the J.M. Kaplan Fund", "start_char_pos": 662, "end_char_pos": 697, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "1962, and was deposed in 1963.", "after": "its first free elections, in December 1962. Bosch was deposed by a right-wing coup in September 1963, and President Johnson sent troops to prevent him from being restored to power.", "start_char_pos": 751, "end_char_pos": 781, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 175, 345, 559], "domain": "wiki"}
{"doc_id": "5623", "revision_depth": 1, "before_revision": "The Alter Strom, in the sea resort of Warnemnde, Germany. The Royal Canal in Ireland. Small boat canals such as the Basingstoke Canal fuelled the industrial revolution in much of Europe and the United States. Bridge on the Naviglio Grande, in the town of Cassinetta di Lugagnano, in Italy Canal in Broek in Waterland, Netherlands. Canal in Venice. Canals are waterways channels, or artificial waterways, for water conveyance, or to service water transport vehicles. They may also help with irrigation. It can be thought of as an artificial version of a river. Canals carry free surface flow under atmospheric pressure  . In most cases, the engineered works will have a series of dams and locks that create reservoirs of low speed current flow. These reservoirs are referred to as slack water levels, often just called levels. A canal is also known as a navigation when it parallels a river and shares part of its waters and drainage basin, and leverages its resources by building dams and locks to increase and lengthen its stretches of slack water levels while staying in its valley. In contrast, a canal cuts across a drainage divide atop a ridge, generally requiring an external water source above the highest elevation.  Many canals have been built at elevations towering over valleys and other water ways crossing far below .", "after_revision": "The Alter Strom, in the sea resort of Warnemnde, Germany. The Royal Canal in Ireland. Small boat canals such as the Basingstoke Canal fuelled the industrial revolution in much of Europe and the United States. Bridge on the Naviglio Grande, in the town of Cassinetta di Lugagnano, in Italy Canal in Broek in Waterland, Netherlands. Canal in Venice. Canals are waterway channels, or artificial waterways, for water conveyance, or for servicing water transport vehicles. They carry free surface flow under atmospheric pressure , and can be thought of as artificial rivers . In most cases, a canal has a series of dams and locks that create reservoirs of low speed current flow. These reservoirs are referred to as slack water levels, often just called levels. A canal can be called a navigation canal when it parallels a river and shares part of its waters and drainage basin, and leverages its resources by building dams and locks to increase and lengthen its stretches of slack water levels while staying in its valley. A canal can cut across a drainage divide atop a ridge, generally requiring an external water source above the highest elevation. The best-known example of such a canal is the Panama Canal. Many canals have been built at elevations , above valleys and other waterways .", "edit_actions": [{"type": "R", "before": "Canals are waterways", "after": "Canals are waterway", "start_char_pos": 348, "end_char_pos": 368, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "to service", "after": "for servicing", "start_char_pos": 429, "end_char_pos": 439, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "D", "before": "may also help with irrigation. It can be thought of as an artificial version of a river. Canals", "after": null, "start_char_pos": 471, "end_char_pos": 566, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": ", and can be thought of as artificial rivers", "start_char_pos": 618, "end_char_pos": 618, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the engineered works will have a", "after": "a canal has a", "start_char_pos": 636, "end_char_pos": 668, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "is also known as a navigation", "after": "can be called a navigation canal", "start_char_pos": 834, "end_char_pos": 863, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "In contrast, a canal cuts", "after": "A canal can cut", "start_char_pos": 1085, "end_char_pos": 1110, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "The best-known example of such a canal is the Panama Canal.", "start_char_pos": 1224, "end_char_pos": 1224, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "towering over", "after": ", above", "start_char_pos": 1267, "end_char_pos": 1280, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "water ways crossing far below", "after": "waterways", "start_char_pos": 1299, "end_char_pos": 1328, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 57, 85, 208, 330, 347, 465, 501, 559, 620, 743, 825, 1084, 1223]}
{"doc_id": "563109", "revision_depth": 1, "before_revision": "2000 onwards Sabri Brothers performed at Musica Sacra International Mktoberdorf Event in the year 2000 at Germany. Maqbool Ahmed Sabri   & Mehmood Ghaznavi Sabri Leading The Sabri Brothers In Moscow, 2001 Mehmood Ghaznavi Sabri performing in Moscow, 2001 In 2001, Kamal Ahmed Sabri  second eldest of the Sabri brothers , died after suffering from a second heart attack, after the death of Kamal Sabri, instruments such as Swarmandal and Flexatone came to end in the ensemble . Later the same year, Sabri Brothers performed live at in DOM at On The Carpet Oriental Culture Festival on 17 November 2001 in Moscow which was later released in 2003 as album Live in Moscow Diwani Maqbool Ahmed Sabri along with Mehmood Ghaznavi Sabri toured India in 2004 and recorded an album titled Rabb E Akbar. which was his last official release. In 2005, Sabri Brothers performed at Virsast Fesival held in Dehradun, India. The same year, Maqbool Ahmed Sabri was invited to perform in different cities of Madhya Pardesh, India and was awarded the Tansen Samman Award by Government of Madhya Pradesh. In 2006, Sabri Brothers performed at Ajmer Sharif and they performed in Pune, Maharashtra during the same year. the same year , he re created and re released his old hits like Khwaja Ki Deewani, Mera Koi Nahi Hai Tere Siwa, Mann Bole Mann, and Hum Panjatani Hai which  proved to be a big hit  . Maqbool Ahmed Sabri did various international and national tours till his death. He performed in Hyderabad India in 2008.", "after_revision": "2000 onwards Sabri Brothers performed at Musica Sacra International Mktoberdorf Event concerts during the year 2000 in Germany. Maqbool Ahmed Sabri along with Mehmood Ghaznavi Sabri did several tours of India and recorded various Musial albums which were released by Tips Music. Maqbool Ahmed Sabri & Mehmood Ghaznavi Sabri Leading The Sabri Brothers In Moscow, 2001 Mehmood Ghaznavi Sabri performing in Moscow, 2001 In 2001, Kamal Ahmed Sabri who was the second eldest of the Sabri Brothers , died after suffering from a second heart attack, after the death of Kamal Sabri, instruments such as Swarmandal (Zither) and Chidya Tarang (Flexatone) came to end in Sabri Brothers group . Later the same year, Sabri Brothers performed live in DOM at On The Carpet Oriental Culture Festival on 17 November 2001 in Moscow which was later released in 2003 as album Live in Moscow Diwani In 2005, Sabri Brothers performed at Virsast Fesival held in Dehradun, India. The same year, Maqbool Ahmed Sabri was invited to perform in different cities of Madhya Pardesh, India and was awarded the Tansen Samman Award by Government of Madhya Pradesh. In 2006, Sabri Brothers performed at Ajmer Sharif and they performed in Pune, Maharashtra . The same year , Maqbool Ahmed Sabri re created and re released his old blockbuster super hits including Khwaja Ki Deewani, Mera Koi Nahi Hai Tere Siwa, Mann Bole Mann, and Hum Panjatani Hai which were released by SB Studios in Karachi and proved to be a big hit worldwide . Maqbool Ahmed Sabri did various tours and Qawwali concerts till his death. He performed in Hyderabad India in 2008.", "edit_actions": [{"type": "R", "before": "in", "after": "concerts during", "start_char_pos": 86, "end_char_pos": 88, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "at", "after": "in", "start_char_pos": 103, "end_char_pos": 105, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "along with Mehmood Ghaznavi Sabri did several tours of India and recorded various Musial albums which were released by Tips Music.", "start_char_pos": 135, "end_char_pos": 135, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Maqbool Ahmed Sabri", "start_char_pos": 136, "end_char_pos": 136, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "who was the", "start_char_pos": 282, "end_char_pos": 282, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "brothers", "after": "Brothers", "start_char_pos": 310, "end_char_pos": 318, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "and Flexatone", "after": "(Zither) and Chidya Tarang (Flexatone)", "start_char_pos": 433, "end_char_pos": 446, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "the ensemble", "after": "Sabri Brothers group", "start_char_pos": 462, "end_char_pos": 474, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "D", "before": "at", "after": null, "start_char_pos": 528, "end_char_pos": 530, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "Maqbool Ahmed Sabri along with Mehmood Ghaznavi Sabri toured India in 2004 and recorded an album titled Rabb E Akbar. which was his last official release.", "after": null, "start_char_pos": 675, "end_char_pos": 829, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "during the same year. the", "after": ". The", "start_char_pos": 1174, "end_char_pos": 1199, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "he", "after": "Maqbool Ahmed Sabri", "start_char_pos": 1212, "end_char_pos": 1214, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": "hits like", "after": "blockbuster super hits including", "start_char_pos": 1250, "end_char_pos": 1259, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "were released by SB Studios in Karachi and", "start_char_pos": 1352, "end_char_pos": 1352, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "worldwide", "start_char_pos": 1376, "end_char_pos": 1376, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "international and national tours", "after": "tours and Qawwali concerts", "start_char_pos": 1411, "end_char_pos": 1443, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 114, 199, 249, 263, 320, 369, 829, 907, 1083, 1459], "domain": "wiki"}
{"doc_id": "564429", "revision_depth": 1, "before_revision": "The Wlastkwewiyik, or Maliseet (,Erickson, Vincent O. (1978). \"Maliseet-Passamaquoddy.\" In Northeast, ed. Bruce G. Trigger. Vol. 15 of Handbook of North American Indians, ed. William C. Sturtevant. Washington, D.C.: Smithsonian Institution, pg. 123. also spelled Malecite), are an Algonquian-speaking First Nation of the Wabanaki Confederacy. They are the indigenous people of the Wlastk River (Saint John River) valley and its tributaries , and their territory extends across the current borders of New Brunswick and Quebec in Canada, and parts of Maine in the United States. The Houlton Band of Maliseet Indians, based on the Meduxnekeag River in the Maine portion of their traditional homeland, are since 19 July 1776, the first foreign Treaty allies with the United States of America and a federally recognized tribe of Maliseet people by the United States . Today Maliseet people have also migrated to other parts of the world. Maliseets are forest, river and coastal people within their 20,000, 000 acre, 200 mile wide, and 600 mile long Saint John river watershed homeland . Name The people call themselves Wlastkwewiyik after the Wlastkw River at the heart of their expansive territory in what is now New Brunswick and Maine. Wlastkw means \"Beautiful River.\" English colonists later named it the Saint John River. Wlastkwewiyik means \"People of the Beautiful River,\" in Maliseet.LeSourd, Philip, ed. 2007. Tales from Maliseet Country: The Maliseet Texts of Karl V. Teeter, Lincoln, Nebraska: University of Nebraska Press, p. 17, fnote 4 The Maliseet (Malecite) have long been associated with the Saint John  . Their territory still extends as far as the Saint Lawrence River. Their lands and resources are bounded on the east by the Mikmaq people, on the west by the Penobscot people , and on the south by the Passamaquoddy people , who also still speak related Algonquian languages. Malesse'jik was a Mikmaq word believed to mean \"He speaks slowly,\" or differently  by which the Mikmaq people contrasted the other tribe's language to their own. The meaning of the word today is unknown but  is commonly mistranslated to \"he speaks badly, lazy, or broken\"  Erickson 1978, pg. 135 This term is the exonym by which the Mikmaq people referred to this group when speaking to early Europeans. The Europeans met the Mikmaq people before the Wlastkwewiyik, and adopted Malesse'jik as Malcite in French for the people, not understanding that it was not their true name. The later English colonists anglicized this term as Maliseet, in another transliteration of sound  . 17th century At the time of European encounter, the Wlastkwewiyik were living in walled villages and practicing horticulture (corn, beans, squash and tobacco). In addition to  growing crops, the women gathered and processed fruits, berries, nuts and natural produce. The men contributed by fishing and hunting  . Written accounts in the early 17th century, such as those of Samuel de Champlain and Marc LesCarbot, refer to a large Malecite village at the mouth of the Saint John River. Later in the century, sources indicate their headquarters had shifted upriver to Meductic, on the middle reaches of the Saint John River. The French explorers were the first to establish a fur trade with the Wlastkwewiyik, which became important in their territory. Some European goods were desired because they were useful to Wlastkwewiyik subsistence and culture. The French Jesuits also established missions  where some Wlastkwewiyik converted to Catholicism. With years of colonialism, many learned the French language. The French called them Malcite, a transliteration of the Mi'kmaq name for the people. Local histories depict many encounters with the Iroquois, five powerful nations based south and east of the Great Lakes, and the Innu  . Contact with European fisher-traders in the early 17th century and with specialized fur traders developed into a stable relationship which lasted for nearly 100 years. Despite devastating population losses to European infectious diseases, to which they had no immunity, these Atlantic First Nations held on to their traditional coastal or river locations for hunting, fishing and gathering , and were concentrated along river valleys for trapping.", "after_revision": "The Wlastkwewiyik, or Maliseet (,Erickson, Vincent O. (1978). \"Maliseet-Passamaquoddy.\" In Northeast, ed. Bruce G. Trigger. Vol. 15 of Handbook of North American Indians, ed. William C. Sturtevant. Washington, D.C.: Smithsonian Institution, pg. 123. also spelled Malecite), are an Algonquian-speaking First Nation of the Wabanaki Confederacy. They are the indigenous people of the Wlastk River (Saint John River) valley and its tributaries . Their territory extends across the current borders of New Brunswick and Quebec in Canada, and parts of Maine in the United States. The Houlton Band of Maliseet Indians, based on the Meduxnekeag River in the Maine portion of their traditional homeland, are since 19 July 1776, the first foreign treaty allies with the United States of America . They are a federally recognized tribe of Maliseet people . Today Maliseet people have also migrated to other parts of the world. The Maliseet have occupied areas of forest, river and coastal areas within their 20,000, 000-acre, 200-mile wide, and 600-mile long homeland in the Saint John river watershed . Name The people call themselves Wlastkwewiyik after the Wlastkw River at the heart of their expansive territory in what is now New Brunswick and Maine. Wlastkw means \"Beautiful River.\" English colonists later named it the Saint John River. Wlastkwewiyik means \"People of the Beautiful River,\" in Maliseet.LeSourd, Philip, ed. 2007. Tales from Maliseet Country: The Maliseet Texts of Karl V. Teeter, Lincoln, Nebraska: University of Nebraska Press, p. 17, fnote 4 The Maliseet (Malecite) have long been associated with the Saint John River . Their territory still extends as far as the Saint Lawrence River. Their lands and resources are bounded on the east by the Mikmaq people, on the west by the Penobscot , and on the south by the Passamaquoddy , who also still speak related Algonquian languages. Malesse'jik was a Mikmaq word believed to mean \"He speaks slowly,\" or differently , by which the Mikmaq people contrasted the other tribe's language to their own. The meaning of the word today is unknown but it is commonly mistranslated to \"he speaks badly, lazy, or broken\" . Erickson 1978, pg. 135 This term is the exonym by which the Mikmaq people referred to this group when speaking to early Europeans. The Europeans met the Mikmaq people before the Wlastkwewiyik, and transliterated Malesse'jik to Malcite in French for this people, not understanding that it was not their true name. The later English colonists anglicized this term as Maliseet, in another transliteration of sound in their own language . 17th century At the time of European encounter, the Wlastkwewiyik were living in walled villages and practicing horticulture (corn, beans, squash and tobacco). In addition to cultivating and growing crops, the women gathered and processed fruits, berries, nuts and natural produce. The men contributed by fishing and hunting , and the women cooked these finds . Written accounts in the early 17th century, such as those of Samuel de Champlain and Marc LesCarbot, refer to a large Malecite village at the mouth of the Saint John River. Later in the century, sources indicate their headquarters had shifted upriver to Meductic, on the middle reaches of the Saint John River. The French explorers were the first to establish a fur trade with the Wlastkwewiyik, which became important in their territory. Some European goods were desired because they were useful to Wlastkwewiyik subsistence and culture. The French Jesuits also established missions , where some Wlastkwewiyik converted to Catholicism. After years of colonialism, many learned the French language. The French called them Malcite, a transliteration of the Mi'kmaq name for the people. Local histories depict many encounters with the Iroquois, five powerful nations based south and east of the Great Lakes, and the Innu located to the north . Contact with European fisher-traders in the early 17th century and with specialized fur traders developed into a stable relationship which lasted for nearly 100 years. Despite devastating population losses to European infectious diseases, to which they had no immunity, these Atlantic First Nations held on to their traditional coastal or river locations for hunting, fishing and gathering . They lived along river valleys for trapping.", "edit_actions": [{"type": "R", "before": ", and their", "after": ". Their", "start_char_pos": 440, "end_char_pos": 451, "major_intent": "coherence", "raw_intents": ["others", "coherence", "coherence"]}, {"type": "R", "before": "Treaty", "after": "treaty", "start_char_pos": 740, "end_char_pos": 746, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and", "after": ". They are", "start_char_pos": 788, "end_char_pos": 791, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "by the United States", "after": null, "start_char_pos": 840, "end_char_pos": 860, "major_intent": "clarity", "raw_intents": ["others", "clarity", "clarity"]}, {"type": "R", "before": "Maliseets are", "after": "The Maliseet have occupied areas of", "start_char_pos": 933, "end_char_pos": 946, "major_intent": "others", "raw_intents": ["others", "fluency", "clarity"]}, {"type": "R", "before": "people", "after": "areas", "start_char_pos": 973, "end_char_pos": 979, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "000 acre, 200 mile", "after": "000-acre, 200-mile", "start_char_pos": 1001, "end_char_pos": 1019, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "R", "before": "600 mile long", "after": "600-mile long homeland in the", "start_char_pos": 1030, "end_char_pos": 1043, "major_intent": "others", "raw_intents": ["others", "fluency", "style"]}, {"type": "D", "before": "homeland", "after": null, "start_char_pos": 1071, "end_char_pos": 1079, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "A", "before": null, "after": "River", "start_char_pos": 1615, "end_char_pos": 1615, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "people", "after": null, "start_char_pos": 1785, "end_char_pos": 1791, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "people", "after": null, "start_char_pos": 1832, "end_char_pos": 1838, "major_intent": "others", "raw_intents": ["others", "style", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1974, "end_char_pos": 1974, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "it", "start_char_pos": 2099, "end_char_pos": 2099, "major_intent": "fluency", "raw_intents": ["coherence", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ".", "start_char_pos": 2164, "end_char_pos": 2164, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "adopted", "after": "transliterated", "start_char_pos": 2362, "end_char_pos": 2369, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "as", "after": "to", "start_char_pos": 2382, "end_char_pos": 2384, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "R", "before": "the", "after": "this", "start_char_pos": 2407, "end_char_pos": 2410, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "in their own language", "start_char_pos": 2568, "end_char_pos": 2568, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "A", "before": null, "after": "cultivating and", "start_char_pos": 2746, "end_char_pos": 2746, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ", and the women cooked these finds", "start_char_pos": 2881, "end_char_pos": 2881, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 3468, "end_char_pos": 3468, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "With", "after": "After", "start_char_pos": 3520, "end_char_pos": 3524, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "located to the north", "start_char_pos": 3801, "end_char_pos": 3801, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", and were concentrated", "after": ". They lived", "start_char_pos": 4194, "end_char_pos": 4217, "major_intent": "coherence", "raw_intents": ["others", "coherence", "coherence"]}], "sents_char_pos": [0, 61, 87, 105, 123, 174, 197, 342, 576, 862, 932, 1233, 1321, 1387, 1413, 1472, 1617, 1683, 1891, 2053, 2295, 2469, 2730, 2837, 2883, 3056, 3194, 3322, 3422, 3519, 3580, 3666, 3971], "domain": "wiki"}
{"doc_id": "5742250", "revision_depth": 1, "before_revision": "Rauf Lala (Urdu: ) (born 1970 in Larkana, Sindh, Pakistan) is a popular Pakistani comedian, actor, writer, and producer. He has been working for over three decades in Pakistani media, And is well known for his comedic talent and stage work. Lala also have been working in Indiaand has won The Great Indian Laughter Challenge - Season 2. Career Stage play He started his career in 1985  as a comedian in One day Theaterin Karachialong with his teacher Razak Rajo. His first commercial show was Susral Bara Janjaal  in 1985 with Umer Sharif Javed Sheikh and others. He has worked  with Umer Sharif and Moin Akhtar on many stage shows such as Bhudda Ghar Per Hai, and  Bakra Qistoon Pay in 1989 which still is a popular and well-known Pakistani stage show internationally. Lala came on several comedy reality shows and gained popularity on television  as well as movies in India.  Filmography Year Title Role Place 1983 Manila K Jaanbaz Supporting Role Lollywood 1990 Hum Toh Chale Susraal Guy From America Lollywood 1993 Duniya Dil Walo Ki Saud's Friend Lollywood 2009 Runway Tulip Joshi's Friend Bollywood 2010 Bhavnao Ko Samjho Film Director Bollywood Television Lala has worked in various serials, commercials , television filmsand shows for almost three decades including: Ek Raat Ek Kahani Funny Family Super Karara Comedy Kings Comedy Champions The Great Indian Family Drama In 2006, he was invited in an Indian reality show by channel Star One, to participate with other fellow Pakistani's and Indian actors. His special episode was when Bipasha Basu was there in the final episode of The Great Indian Laughter Challenge - Season 2 which he won and where he was awarded the title \"Comedy Ka Shahenshah\". He is the first Pakistani to win that major Indian contest. | |", "after_revision": "Rauf Lala (Urdu: ) (born 1970 in Larkana, Sindh, Pakistan) is a Pakistani comedian, actor, writer, and producer. Working in Pakistani media for over three decades , he is well known for his comedic talent and stage work. Lala has also worked in India, winning season two of The Great Indian Laughter Challenge . Career Rauf Lala began his stage career in 1985 , as a comedian in One Day Theater, Karachi, alongside his teacher, Razak Rajo. His first commercial show was Susral Bara Janjaal also in 1985 , which featured Umer Sharif and Javed Sheikh. He has collaborated with Umer Sharif and Moin Akhtar on many stage shows , including Bhudda Ghar Per Hai, and in 1989, Bakra Qistoon Pay , which remains a popular and well-known Pakistani stage show internationally. Lala has appeared on several comedy reality shows , gaining popularity on television , as well as appearing in movies in Pakistan and India. In 2006, he was invited to appear in season two of The Great Indian Laughter Challenge, an Indian reality television show on channel Star One, which he won. He was awarded the title \"Comedy Ka Shahenshah\", becoming the first Pakistani to win the major Indian contest. He appeared alongside actress Bipasha Basu in the final episode. Filmography Year Title Role Place 1983 Manila K Jaanbaz Supporting Role Lollywood 1990 Hum Toh Chale Susraal Guy From America Lollywood 1993 Duniya Dil Walo Ki Saud's Friend Lollywood 2009 Runway Tulip Joshi's Friend Bollywood 2010 Bhavnao Ko Samjho Film Director Bollywood Television Lala has worked in various serials, commercials and television films, including: Ek Raat Ek Kahani Funny Family Super Karara Comedy Kings Comedy Champions The Great Indian Family Drama ", "edit_actions": [{"type": "D", "before": "popular", "after": null, "start_char_pos": 64, "end_char_pos": 71, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "He has been working", "after": "Working in Pakistani media", "start_char_pos": 121, "end_char_pos": 140, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "R", "before": "in Pakistani media, And", "after": ", he", "start_char_pos": 164, "end_char_pos": 187, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "also have been working in Indiaand has won", "after": "has also worked in India, winning season two of", "start_char_pos": 246, "end_char_pos": 288, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": "- Season 2.", "after": ".", "start_char_pos": 325, "end_char_pos": 336, "major_intent": "coherence", "raw_intents": ["coherence", "others", "coherence"]}, {"type": "R", "before": "Stage play He started his", "after": "Rauf Lala began his stage", "start_char_pos": 344, "end_char_pos": 369, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 385, "end_char_pos": 385, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "day Theaterin Karachialong with his teacher", "after": "Day Theater, Karachi, alongside his teacher,", "start_char_pos": 407, "end_char_pos": 450, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "also", "start_char_pos": 513, "end_char_pos": 513, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "R", "before": "with Umer Sharif Javed Sheikh and others. He has worked", "after": ", which featured Umer Sharif and Javed Sheikh.", "start_char_pos": 522, "end_char_pos": 577, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "He has collaborated", "start_char_pos": 578, "end_char_pos": 578, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "such as", "after": ", including", "start_char_pos": 632, "end_char_pos": 639, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "A", "before": null, "after": "in 1989,", "start_char_pos": 665, "end_char_pos": 665, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in 1989 which still is", "after": ", which remains", "start_char_pos": 684, "end_char_pos": 706, "major_intent": "coherence", "raw_intents": ["coherence", "others", "clarity"]}, {"type": "R", "before": "came", "after": "has appeared", "start_char_pos": 775, "end_char_pos": 779, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "R", "before": "and gained", "after": ", gaining", "start_char_pos": 812, "end_char_pos": 822, "major_intent": "coherence", "raw_intents": ["coherence", "others", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 848, "end_char_pos": 848, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "movies in", "after": "appearing in movies in Pakistan and", "start_char_pos": 860, "end_char_pos": 869, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "In 2006, he was invited to appear in season two of The Great Indian Laughter Challenge, an Indian reality television show on channel Star One, which he won. He was awarded the title \"Comedy Ka Shahenshah\", becoming the first Pakistani to win the major Indian contest. He appeared alongside actress Bipasha Basu in the final episode.", "start_char_pos": 877, "end_char_pos": 877, "major_intent": "coherence", "raw_intents": ["coherence", "meaning-changed", "coherence"]}, {"type": "R", "before": ", television filmsand shows for almost three decades", "after": "and television films,", "start_char_pos": 1211, "end_char_pos": 1263, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "In 2006, he was invited in an Indian reality show by channel Star One, to participate with other fellow Pakistani's and Indian actors. His special episode was when Bipasha Basu was there in the final episode of The Great Indian Laughter Challenge - Season 2 which he won and where he was awarded the title \"Comedy Ka Shahenshah\". He is the first Pakistani to win that major Indian contest. | |", "after": null, "start_char_pos": 1379, "end_char_pos": 1772, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}], "sents_char_pos": [0, 120, 240, 462, 563, 769, 876, 1274, 1513, 1708], "domain": "wiki"}
{"doc_id": "5742518", "revision_depth": 2, "before_revision": "Wissenschaft is the German language term for scholarship, research, higher education, and academia. The word translates exactly into many other languages, e.g. vetenskap in Swedish or nauka in Polish, but there is no exact translation in modern English. The common translation as \"science \" is misleading, because Wissenschaft equally includes humanities  , and sciences and humanities are mutually exclusive categories in modern English. Wissenschaft includes humanities like art or religion at the same level as sciences like chemistry or psychology. Wissenschaft incorporates scientific and non-scientific inquiry, learning, knowledge, scholarship, and does not necessarily imply empirical research. History Before Immanuel Kant published his Critique of Judgment in 1790, the \"sch\u00f6ne Wissenschaft\" (roughly, \"fine sciences\") were highly regarded. The \" sch\u00f6ne Wissenschaft\" included poetry, rhetoric, and other subjects that were meant to promote an understanding of truth, beauty, and goodness. Kant argued that aesthetic judgments were not an area of systematic knowledge, and therefore were outside the realm of Wissenschaft.", "after_revision": "Wissenschaft is a German language term that embraces scholarship, research, higher education, and academia. Wissenschaft translates exactly into many other languages, e.g. vetenskap in Swedish or nauka in Polish, but there is no exact translation in modern English. The common translation to science is misleading, because Wissenschaft equally includes humanities (Geisteswissenschaft) , and sciences and humanities are mutually exclusive categories in modern English. Wissenschaft includes humanities like history, anthropology, or arts (study of literature, visual arts, or music) at the same level as sciences like chemistry or psychology. Wissenschaft incorporates scientific and non-scientific inquiry, learning, knowledge, scholarship, and does not necessarily imply empirical research. History Before Immanuel Kant published his Critique of Judgment in 1790, \"sch\u00f6ne Wissenschaft\" was highly regarded. \" Sch\u00f6ne Wissenschaft\" included poetry, rhetoric, and other subjects that were meant to promote an understanding of truth, beauty, and goodness. Kant argued that aesthetic judgments were not an area of systematic knowledge, and therefore were outside the realm of Wissenschaft.", "edit_actions": [{"type": "R", "before": "the", "after": "a", "start_char_pos": 16, "end_char_pos": 19, "major_intent": "clarity", "raw_intents": ["clarity", "others", "fluency"]}, {"type": "R", "before": "for", "after": "that embraces", "start_char_pos": 41, "end_char_pos": 44, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "The word", "after": "Wissenschaft", "start_char_pos": 100, "end_char_pos": 108, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "as \"science \"", "after": "to science", "start_char_pos": 277, "end_char_pos": 290, "major_intent": "fluency", "raw_intents": ["fluency", "others", "clarity"]}, {"type": "A", "before": null, "after": "(Geisteswissenschaft)", "start_char_pos": 355, "end_char_pos": 355, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "art or religion", "after": "history, anthropology, or arts (study of literature, visual arts, or music)", "start_char_pos": 477, "end_char_pos": 492, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "the", "after": null, "start_char_pos": 776, "end_char_pos": 779, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "(roughly, \"fine sciences\") were", "after": "was", "start_char_pos": 802, "end_char_pos": 833, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "D", "before": "The", "after": null, "start_char_pos": 851, "end_char_pos": 854, "major_intent": "coherence", "raw_intents": ["coherence", "others", "fluency"]}, {"type": "R", "before": "sch\u00f6ne", "after": "Sch\u00f6ne", "start_char_pos": 857, "end_char_pos": 863, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 99, 253, 438, 552, 702, 850, 999], "domain": "wiki"}
{"doc_id": "58229645", "revision_depth": 1, "before_revision": "The celebration took place for the first time on 1 October 1991. The day is known to increase awareness of the factors and issues that affect older adults, such as age deterioration and the study of problems that affect their lives, elder abuse. It is also a day to recognize and acknowledge the contributions of older people to society. Purpose The main purpose of National Senior Citizen's Day is to raise awareness of the condition of older people and to support them throughout the aging process. The day is also recognized for the well-being of the elderly to recognize and appreciate their accomplishments and their participation in society in order to promote through their skills and knowledge. The reason to celebrate this day is to thank the elders for everything they have done for their children and to acknowledge them. They sacrifice their entire lives to nurture their relationships. Such a selfless service that his whole life has received from his family deserves something important. That's why National Senior Citizens' Day is celebrated. When they dedicate a special day of the year to the elderly, they give their families the opportunity to see their love and commitment. National Senior Citizens' Day is an important moment to remind the government of its commitment to work towards goals, to help seniors and, among other things, to end hunger and poverty for the elderly and people of all ages.", "after_revision": "The celebration took place for the first time on 1 October 1991. The day is intended to increase awareness of the factors and issues that affect older adults, such as health deterioration and elder abuse. It is also a day to recognize and acknowledge the contributions of older people to society. Purpose The main purpose of National Senior Citizen's Day is to raise awareness of the condition of older people and to support them throughout the aging process. The day is also recognized for the well-being of the elderly to recognize and appreciate their accomplishments and their participation in society in order to promote their skills and knowledge. The reason to celebrate this day is to thank the elders for everything they have done for their children and to acknowledge them. By dedicating a special day of the year to the elderly, they give their families the opportunity to see their love and commitment. National Senior Citizens' Day is an important moment to remind governments of their commitment to work towards goals, to help seniors and, among other things, to end hunger and poverty for the elderly and people of all ages.", "edit_actions": [{"type": "R", "before": "known", "after": "intended", "start_char_pos": 76, "end_char_pos": 81, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "R", "before": "age deterioration and the study of problems that affect their lives,", "after": "health deterioration and", "start_char_pos": 164, "end_char_pos": 232, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "through", "after": null, "start_char_pos": 667, "end_char_pos": 674, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "They sacrifice their entire lives to nurture their relationships. Such a selfless service that his whole life has received from his family deserves something important. That's why National Senior Citizens' Day is celebrated. When they dedicate a", "after": "By dedicating a", "start_char_pos": 833, "end_char_pos": 1078, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "the government of its", "after": "governments of their", "start_char_pos": 1257, "end_char_pos": 1278, "major_intent": "fluency", "raw_intents": ["fluency", "others", "style"]}], "sents_char_pos": [0, 64, 245, 337, 500, 702, 832, 898, 1001, 1057, 1193], "domain": "wiki"}
{"doc_id": "59117588", "revision_depth": 1, "before_revision": "Focus of celebrations by year Event logo on a digital screen in Riga  Each year of the anniversary has a keyword in recognition of events that took place 100 years ago, which were significant to the foundation of the country. Dignitaries laying wreaths at the Freedom Monument on 18 November 2018 As per  events on its actual centennial anniversary on 18 November 2018, many high-level government officials, dignitaries, military leaders, and public figures laid wreaths and flowers at the Freedom Monument in the capital. A musical concert was held at the Latvian National Theatre, where Latvia declared its sovereignty in 1918. A special meeting of the Saeima was alsoheld at the theatre. Military Parade The Italian Bersaglieri marching in the military parade . The annual military parade along the 11. novembra krastmala (November 11th Embankment) in Riga marked the centennial anniversary' s main events . It was the largest in history with around 1,700 soldiers from the Latvian National Armed Forces participating in the march past. It was composed of personnel of the Land Forces, Navy, Air Force, National Guard, Border Guard, Military Police, State Police, State Fire and Rescue Service and military academies. The parade was led by Latvian President Raimonds Vjonis, Chief of the National Defence Forces Leonds Kalni and Naval Forces commander Ingus Vizulis. Estonian President Kersti Kaljulaid and Armed Forces chief Riho Terras, Lithuanian President Dalia Grybauskait, Finnish President Sauli Niinist and Icelandic President Guni Th. Jhannesson joined in attending the military parade and the celebrations. The NATO countries and other military partners of Latvia that were represented at the parade were Italy, Estonia, Lithuania, Poland, Romania, the Czech Republic, Finland, Albania, Croatia, the Netherlands, Norway, Slovakia, Spain Slovenia, Britain, Denmark, Sweden, Germany, Canada and the United States.26th Latvian Song and Dance Festival closing concert \"On the Starry Path\" . Latvian Song and Dance Festival", "after_revision": "Focus of celebrations by year Event logo on a digital screen in Riga on Brvbas iela Each year of the anniversary has a keyword in recognition of events that took place 100 years ago, which were significant to the foundation of the country. Dignitaries laying wreaths at the Freedom Monument on 18 November 2018 As per the events on its actual centennial anniversary on 18 November 2018, many high-level government officials, dignitaries, military leaders, and public figures laid wreaths and flowers at the Freedom Monument in the capital. A musical concert was held at the Latvian National Theatre, where Latvia declared its sovereignty in 1918. A special meeting of the Saeima was alsoheld at the theatre. Military Parade The Italian Bersaglieri marching in the military parade The annual military parade along the 11. novembra krastmala (November 11th Embankment) in Riga marked one of the main events of the centennial anniversary' . It was the largest military parade in Latvian history with around 1,700 soldiers from the Latvian National Armed Forces participating in the march past. It was composed of personnel of the Land Forces, Navy, Air Force, National Guard, Border Guard, Military Police, State Police, State Fire and Rescue Service and military academies. The parade was led by Latvian President Raimonds Vjonis, Chief of the National Defence Forces Leonds Kalni and Naval Forces commander Ingus Vizulis. Estonian President Kersti Kaljulaid and Armed Forces chief Riho Terras, Lithuanian President Dalia Grybauskait, Finnish President Sauli Niinist and Icelandic President Guni Th. Jhannesson joined in attending the military parade and the celebrations. The NATO countries and other military partners of Latvia that were represented at the parade were Italy, Estonia, Lithuania, Poland, Romania, the Czech Republic, Finland, Albania, Croatia, the Netherlands, Norway, Slovakia, Spain Slovenia, Britain, Denmark, Sweden, Germany, Canada and the United States.26th Latvian Song and Dance Festival closing concert \"On the Starry Path\" Latvian Song and Dance Festival", "edit_actions": [{"type": "A", "before": null, "after": "on Brvbas iela", "start_char_pos": 69, "end_char_pos": 69, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 304, "end_char_pos": 304, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 763, "end_char_pos": 764, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "others"]}, {"type": "R", "before": "the", "after": "one of the main events of the", "start_char_pos": 867, "end_char_pos": 870, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "s main events", "after": null, "start_char_pos": 895, "end_char_pos": 908, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "in", "after": "military parade in Latvian", "start_char_pos": 930, "end_char_pos": 932, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "D", "before": ".", "after": null, "start_char_pos": 1998, "end_char_pos": 1999, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 225, 522, 629, 690, 764, 910, 1039, 1220, 1369, 1546, 1619, 1924], "domain": "wiki"}
{"doc_id": "59170248", "revision_depth": 1, "before_revision": "To some degree the term 'Iron Age' is misapplied, as little evidence exists for any indigenous iron-work outside the finds at Muwailah , themselves thought to be imports, and even the extensive evidence of smelting throughout the Iron Age found at Saruq Al Hadid is dominated by copper and tin production. Iron Age II A pot, discovered in the Iron Age building at Bidaa bint Saud and on display at the Al Ain National Museum. It is thought to be an incense burner. Extensive evidence of Iron Age II settlement has been found throughout the UAE, particularly at Muwailah, Al Thuqeibah, Bidaa bint Saud, as well as Rumailah and Qattara in Al Ain. The development of increasingly complex irrigation ditches and waterways, falaj (plural aflaj) took place during this time and finds at Bidaa bint Saud and Thuqeibah date back to the Iron Age II period pre-dating finds of qanat waterways in Iran. Early finds of aflaj, particularly those around the desert city of Al Ain, have been cited as the earliest evidence of the construction of these waterways. It is thought nearby Bidaa bint Saud became an important site during the Iron Age, both as a caravan stop and as a settled community of farmers that used the falaj irrigation system there. Two of these irrigation passages have been partly excavated at Bidaa bint Saud, with a number of sections remaining in reasonable condition. In one of the excavations, a number of sandstone-lined shaft holes were discovered, as well as a stepped underground access point and a large open cistern. Evidence of formerly irrigated land has also been found at the site. Iron age aflaj Recent finds of pottery in Thuqeibah and Madam have further linked the development of early aflaj ( or qanat ) water systems there to an Iron Age II date, further substantiating the attribution of the innovation of these water systems to a southeastern Arabian origin based on the extensive archaeological work of Dr Wasim Takriti around the area of Al Ain. The 2002 publication of a paper by Tikriti, The south-east Arabian origin of the falaj system, provided the first counterpoint to the long-accepted narrative, that the Qanat originated in Persia and was identified as such by accounts of the campaigns of the Assyrian King, Sargon II, in 714 BCE. Tikriti cites this and also accounts by the Greek second and third century historian Polybius as being the basis for academic attribution of the technology to Persia. He notes academics such as JC Wilkinson (1977) adopting an Iranian origin for the technology under the influence of Sargon's annals and Polybius, but points out at least seven Iron Age aflaj (plural for falaj, the word used to denote waterways of this type in the United Arab Emirates) recently discovered in the Al Ain area of the UAE have been reliably carbon dated back to the beginning of the first millennium BCE. Additional to finds of Iron Age aflaj in Al Ain, Tikrit pointed to excavations in Al Madam, Sharjah, by the French archaeological team working there, as well as by a German team working in Maysar, in Oman. Tikriti is at pains to point out that, despite long-standing efforts since the 19th century to excavate qanat systems in Iran, no evidence has been found for any such qanat there dated earlier than the 5th century BCE. He concludes that the technology originated in South East Arabia and was likely taken to Persia, likely by the Sasanian conquest of the Oman peninsular. Others have followed Tikriti's lead. In 2016, Rmy Boucharlat in his paper Qant and Falaj: Polycentric and Multi-Period Innovations Iran and the United Arab Emirates as Case Studies, asserted that the attribution of the technology to Iranians in the early first millennium BCE is a position that cannot longer be maintained . He asserts that the carbon dating of alfaj in Oman and the UAE to the ninth century BCE by Cleuziou and evidence for such an early date provided by Tikriti are definitive. Additionally, Boucharlat maintains that no known Iranian qanat can be dated to the pre-Islamic period. Post Iron Age The period from 3000 BCE has been dubbed both the Mleiha and the Late Pre-Islamic period, and follows on from the dissolution of Darius III's empire. Although the era has been called Hellenistic, Alexander the Great's conquests went no further than Persia and he left Arabia untouched. However, Macedonian coinage unearthed at Ed-Dur dates back to Alexander the Great and contemporary Greek manuscripts have given the exports from Ed-Dur as 'pearls, purple dye, clothing, wine, gold and slaves, and a great quantity of dates'.", "after_revision": "To some degree the term 'Iron Age' is misapplied, as little evidence exists for any indigenous iron-work outside the finds at Muweilah , themselves thought to be imports, and even the extensive evidence of smelting throughout the Iron Age found at Saruq Al Hadid is dominated by copper and tin production. Iron Age II A pot, discovered in the Iron Age building at Bidaa bint Saud and on display at the Al Ain National Museum. It is thought to be an incense burner. Extensive evidence of Iron Age II settlement has been found throughout the UAE, particularly at Muweilah, Thuqeibah, Bidaa bint Saud, as well as Rumailah and Qattara in Al Ain. The development of increasingly complex irrigation ditches and waterways, falaj (plural aflaj) took place during this time and finds at Bidaa bint Saud and Thuqeibah date back to the Iron Age II period pre-dating finds of qanat waterways in Iran. Early finds of aflaj, particularly those around the desert city of Al Ain, have been cited as the earliest evidence of the construction of these waterways. It is thought nearby Bidaa bint Saud became an important site during the Iron Age, both as a caravan stop and as a settled community of farmers that used the falaj irrigation system there. Two of these irrigation passages have been partly excavated at Bidaa bint Saud, with a number of sections remaining in reasonable condition. In one of the excavations, a number of sandstone-lined shaft holes were discovered, as well as a stepped underground access point and a large open cistern. Evidence of formerly irrigated land has also been found at the site. Iron age aflaj Recent finds of pottery in Thuqeibah and Madam have further linked the development of early aflaj ( plural for falaj, the word used to denote waterways of this type in the United Arab Emirates ) water systems there to an Iron Age II date, further substantiating the attribution of the innovation of these water systems to a southeastern Arabian origin based on the extensive archaeological work of Dr Wasim Takriti around the area of Al Ain. The 2002 publication of a paper by Tikriti, The south-east Arabian origin of the falaj system, provided the first counterpoint to the long-accepted narrative, that the Qanat originated in Persia and was identified as such by accounts of the campaigns of the Assyrian King, Sargon II, in 714 BCE. Tikriti cites this and also accounts by the Greek second and third century historian Polybius as being the basis for academic attribution of the technology to Persia. He notes academics such as JC Wilkinson (1977) adopting an Iranian origin for the technology under the influence of Sargon's annals and Polybius, but points out at least seven Iron Age aflaj recently discovered in the Al Ain area of the UAE have been reliably carbon dated back to the beginning of the first millennium BCE. Additional to finds of Iron Age aflaj in Al Ain, Tikrit pointed to excavations in Al Madam, Sharjah, by the French archaeological team working there, as well as by a German team working in Maysar, in Oman. Tikriti is at pains to point out that, despite long-standing efforts since the 19th century to excavate qanat systems in Iran, no evidence has been found for any such qanat there dated earlier than the 5th century BCE. He concludes that the technology originated in South East Arabia and was likely taken to Persia, likely by the Sasanian conquest of the Oman peninsular. Others have followed Tikriti's lead. In 2016, Rmy Boucharlat in his paper Qant and Falaj: Polycentric and Multi-Period Innovations Iran and the United Arab Emirates as Case Studies, asserted that the attribution of the technology to Iranians in the early first millennium BCE is a position that cannot longer be maintained , and that the carbon dating of aflaj in Oman and the UAE to the ninth century BCE by Cleuziou and evidence for such an early date provided by Tikriti are definitive. Additionally, Boucharlat maintains that no known Iranian qanat can be dated to the pre-Islamic period. Post Iron Age The period from 3000 BCE has been dubbed both the Mleiha and the Late Pre-Islamic period, and follows on from the dissolution of Darius III's empire. Although the era has been called Hellenistic, Alexander the Great's conquests went no further than Persia and he left Arabia untouched. However, Macedonian coinage unearthed at Ed-Dur dates back to Alexander the Great . Contemporary Greek manuscripts have given the exports from Ed-Dur as 'pearls, purple dye, clothing, wine, gold and slaves, and a great quantity of dates'.", "edit_actions": [{"type": "R", "before": "Muwailah", "after": "Muweilah", "start_char_pos": 126, "end_char_pos": 134, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Muwailah, Al", "after": "Muweilah,", "start_char_pos": 561, "end_char_pos": 573, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "or qanat", "after": "plural for falaj, the word used to denote waterways of this type in the United Arab Emirates", "start_char_pos": 1718, "end_char_pos": 1726, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "(plural for falaj, the word used to denote waterways of this type in the United Arab Emirates)", "after": null, "start_char_pos": 2630, "end_char_pos": 2724, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": ". He asserts", "after": ", and", "start_char_pos": 3759, "end_char_pos": 3771, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "R", "before": "alfaj", "after": "aflaj", "start_char_pos": 3798, "end_char_pos": 3803, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "and contemporary", "after": ". Contemporary", "start_char_pos": 4418, "end_char_pos": 4434, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 305, 425, 464, 644, 891, 1047, 1236, 1377, 1533, 1602, 1975, 2271, 2438, 2857, 3063, 3282, 3435, 3472, 3760, 3932, 4035, 4199, 4335], "domain": "wiki"}
{"doc_id": "60043236", "revision_depth": 1, "before_revision": "Brig. Pritam Singh is an Indian Army Officer, born in Dina Village in Ferozpur, Punjab, India. He fought in the battle in Singapore in 1942. After the war, he promoted to Lt. Colonel. In 1947, he fought against Pakistan in Poonch. Army Life and Struggle Operation Easy. Poonch link-up 1 November 1948 - 26 November 1948 As a young officer, Brig Pritam Singh fought  the battle of Singapore in 1942 and he wounded badly. Opponent Army prisoned Pritam Singh. Somehow he escaped from the prison of the army campand after, six months he reached , Manipur, India. Later, he was awarded the coveted Military Cross for his great show of bravery. Lt Col Pritam Singh was promoted to the rank of Brigadier in December 1948. Poonch withstood a siege by the Pakistanis from November 1947 till relieved by an Indian offensive, Operation Easy on 20 November 1948. The besieged garrison , commanded by Brig. Pritam Singh , was maintained by air supply. Military operations ended with Poonch town and the eastern part of Poonch district in Indian hands and western Poonch in Pakistani hands. In 1951, he faced a court-martial.", "after_revision": "Brig. Pritam Singh is an Indian Army Officer, born in Dina Village in Ferozpur, Punjab, India. He fought in the Battle of Singapore in 1942. After the war, he promoted to Lt. Colonel. In 1947, he fought against Pakistan in Poonch. Army career Operation Easy. Poonch link-up 1 November 1948 - 26 November 1948 As a young officer, Pritam Singh fought in the battle of Singapore in 1942 and was wounded badly. He was imprisoned by the enemy, but escaped from the army camp, and after six months reached Manipur, India. Later, he was awarded the coveted Military Cross for his bravery. Lt Col Pritam Singh was promoted to the rank of Brigadier in December 1948. Poonch withstood a siege by the Pakistan Army from November 1947 till relieved by an Indian offensive, Operation Easy on 20 November 1948. The besieged garrison Singh commanded was maintained by air supply. Military operations ended with Poonch town and the eastern part of Poonch district in Indian hands and western Poonch in Pakistani hands. In 1951, Singh faced a court-martial.", "edit_actions": [{"type": "R", "before": "battle in", "after": "Battle of", "start_char_pos": 112, "end_char_pos": 121, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "R", "before": "Life and Struggle", "after": "career", "start_char_pos": 236, "end_char_pos": 253, "major_intent": "others", "raw_intents": ["others", "others", "clarity"]}, {"type": "D", "before": "Brig", "after": null, "start_char_pos": 340, "end_char_pos": 344, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "A", "before": null, "after": "in", "start_char_pos": 365, "end_char_pos": 365, "major_intent": "others", "raw_intents": ["others", "others", "fluency"]}, {"type": "R", "before": "he", "after": "was", "start_char_pos": 402, "end_char_pos": 404, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "R", "before": "Opponent Army prisoned Pritam Singh. Somehow he", "after": "He was imprisoned by the enemy, but", "start_char_pos": 420, "end_char_pos": 467, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "prison of the army campand after, six months he reached ,", "after": "army camp, and after six months reached", "start_char_pos": 485, "end_char_pos": 542, "major_intent": "others", "raw_intents": ["others", "others", "clarity"]}, {"type": "D", "before": "great show of", "after": null, "start_char_pos": 616, "end_char_pos": 629, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Pakistanis", "after": "Pakistan Army", "start_char_pos": 747, "end_char_pos": 757, "major_intent": "clarity", "raw_intents": ["clarity", "meaning-changed", "clarity"]}, {"type": "R", "before": ", commanded by Brig. Pritam Singh ,", "after": "Singh commanded", "start_char_pos": 873, "end_char_pos": 908, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "he", "after": "Singh", "start_char_pos": 1086, "end_char_pos": 1088, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 94, 140, 174, 230, 269, 419, 456, 558, 638, 714, 850, 893, 938, 1076], "domain": "wiki"}
{"doc_id": "6069126", "revision_depth": 1, "before_revision": "Time perception is usually categorized under the following three distinct ranges, because different ranges of durations are processed in different parts of the brain . There are many theories and computational models for time perception mechanisms in the brain. William J. Friedman (1993) contrasted two theories for a sense of time: Another theory involves the brain's subconscious tallying of \"pulses\" during a specific interval, forming a biological stopwatch. This theory proposes that the brain can run multiple biological stopwatches at one time depending on the type of task one is involved in. The location of these pulses and what these pulses actually consist of is unclear. This model is only a metaphor and does not stand up in terms of brain physiology or anatomy .", "after_revision": "Time perception is typically categorized in three distinct ranges, because different ranges of duration are processed in different areas of the brain : There are many theories and computational models for time perception mechanisms in the brain. William J. Friedman (1993) contrasted two theories of the sense of time: Another hypothesis involves the brain's subconscious tallying of \"pulses\" during a specific interval, forming a biological stopwatch. This theory proposes that the brain can run multiple biological stopwatches independently depending on the type of tasks being tracked. The source and nature of the pulses is unclear. They are as yet a metaphor whose correspondence to brain anatomy or physiology is unknown .", "edit_actions": [{"type": "R", "before": "usually categorized under the following", "after": "typically categorized in", "start_char_pos": 19, "end_char_pos": 58, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "durations", "after": "duration", "start_char_pos": 110, "end_char_pos": 119, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "parts", "after": "areas", "start_char_pos": 147, "end_char_pos": 152, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": ".", "after": ":", "start_char_pos": 166, "end_char_pos": 167, "major_intent": "others", "raw_intents": ["others", "coherence", "others"]}, {"type": "R", "before": "for a", "after": "of the", "start_char_pos": 313, "end_char_pos": 318, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "theory", "after": "hypothesis", "start_char_pos": 342, "end_char_pos": 348, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "at one time", "after": "independently", "start_char_pos": 540, "end_char_pos": 551, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "task one is involved in. The location of these pulses and what these pulses actually consist of", "after": "tasks being tracked. The source and nature of the pulses", "start_char_pos": 577, "end_char_pos": 672, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "This model is only a metaphor and does not stand up in terms of brain physiology or anatomy", "after": "They are as yet a metaphor whose correspondence to brain anatomy or physiology is unknown", "start_char_pos": 685, "end_char_pos": 776, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 167, 261, 463, 601, 684], "domain": "wiki"}
{"doc_id": "60834637", "revision_depth": 1, "before_revision": "Principles and Foundation During the Warring States period, Han Feizi would propose the establishment of the first all-encompassing autocratic monarchy for the future of the state. During the same time period, Shang Yang from the state of Qin would enact political reforms into practice. The imperial system would  be constituted by the time of the establishment of the Qin, as well as the system of three lords and nine ministers, and the system of prefectures and counties. Weights, measures, currency, and writing were unified. Books and scholars were burned and buried as the ideological control strengthened . Officials were  to act as teachers of the law  .", "after_revision": "Principles and Foundation During the Warring States period, Han Feizi would propose the establishment of the first all-encompassing autocratic monarchy for the future of the state. During the same time period, Shang Yang from the state of Qin would enact political reforms into practice. The imperial system would eventually be constituted by the time of the establishment of the Qin, which would introduce the system of Three Lords and Nine Ministers, and foster the system of prefectures and counties. Units of measurements, currency, and writing would be standardized and books and scholars of the previous regime were burned and buried ; vital to ideological integrity . Officials were then to act as faculties of the law from this point forward .", "edit_actions": [{"type": "A", "before": null, "after": "eventually", "start_char_pos": 314, "end_char_pos": 314, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "as well as", "after": "which would introduce", "start_char_pos": 375, "end_char_pos": 385, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "R", "before": "three lords and nine ministers, and", "after": "Three Lords and Nine Ministers, and foster", "start_char_pos": 400, "end_char_pos": 435, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "Weights, measures,", "after": "Units of measurements,", "start_char_pos": 476, "end_char_pos": 494, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "were unified. Books and scholars", "after": "would be standardized and books and scholars of the previous regime", "start_char_pos": 517, "end_char_pos": 549, "major_intent": "style", "raw_intents": ["style", "clarity", "coherence"]}, {"type": "R", "before": "as the ideological control strengthened", "after": "; vital to ideological integrity", "start_char_pos": 573, "end_char_pos": 612, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "A", "before": null, "after": "then", "start_char_pos": 630, "end_char_pos": 630, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "R", "before": "teachers", "after": "faculties", "start_char_pos": 641, "end_char_pos": 649, "major_intent": "style", "raw_intents": ["style", "style", "clarity"]}, {"type": "A", "before": null, "after": "from this point forward", "start_char_pos": 661, "end_char_pos": 661, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 180, 287, 475, 530, 614], "domain": "wiki"}
{"doc_id": "60834637", "revision_depth": 2, "before_revision": "Evolution of the system of selecting officials Imperial examination paper of Ming dynasty in 1598 AD The emperor receives a candidate during the Palace Examination. The selection of official standards by family background gradually developed to the selection of talent , while the selection method by selection gradually developed to the form of public examination. The selecting mechanism based on talent later became institutionalised and much more rigorous . + Examination and degree hierarchy Degree Ranks Exam Times held Child student (Tongsheng) County/Prefectural Annual (February/April) Student member (Shengyuan) Granary student (1st class)Expanded student (2nd class)Attached student (3rd class) College Triennial (twice) Recommended man (Juren) Top escorted examinee (1st rank) Provincial Triennial Tribute scholar (Gongshi) Top conference examinee (1st rank) Metropolitan Triennial Advanced scholar (Jinshi) Top thesis author (1st rank)Eyes positioned alongside (2nd rank)Flower snatcher (3rd rank) Palace Triennial", "after_revision": "Evolution of Official Selection Imperial examination paper of Ming dynasty in 1598 AD The emperor receives a candidate during the Palace Examination. The standard of Official Selection by familial history gradually developed to emphasize talent instead. This method would eventually progress to form the standard of public examinations, to which this mechanism of cultivating talent would be institutionalized, making examinations a notoriously rigorous process to accomplish . + Examination and degree hierarchy Degree Ranks Exam Times held Child student (Tongsheng) County/Prefectural Annual (February/April) Student member (Shengyuan) Granary student (1st class)Expanded student (2nd class)Attached student (3rd class) College Triennial (twice) Recommended man (Juren) Top escorted examinee (1st rank) Provincial Triennial Tribute scholar (Gongshi) Top conference examinee (1st rank) Metropolitan Triennial Advanced scholar (Jinshi) Top thesis author (1st rank)Eyes positioned alongside (2nd rank)Flower snatcher (3rd rank) Palace Triennial", "edit_actions": [{"type": "R", "before": "the system of selecting officials", "after": "Official Selection", "start_char_pos": 13, "end_char_pos": 46, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "selection of official standards by family background", "after": "standard of Official Selection by familial history", "start_char_pos": 169, "end_char_pos": 221, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "the selection of talent , while the selection method by selection gradually developed to the form of public examination. The selecting mechanism based on talent later became institutionalised and much more rigorous", "after": "emphasize talent instead. This method would eventually progress to form the standard of public examinations, to which this mechanism of cultivating talent would be institutionalized, making examinations a notoriously rigorous process to accomplish", "start_char_pos": 245, "end_char_pos": 459, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}], "sents_char_pos": [0, 164, 365, 551], "domain": "wiki"}
{"doc_id": "6327601", "revision_depth": 1, "before_revision": "He is known for using his position as a marja for political use in the Persian Constitutional Revolution (19051911), where he was one of the main clerical supporters of the revolution. He believed that \"constitutional form of government\" would be the best possible choice in the absence of Imam and regarded the \"Persian revolution\" a Jihad (\"holy war\") in which all Muslims had to participate. He died \"suddenly\" and \"mysteriously\", when he aimed to leave Iraq for Iran in order against support constitutionalists' resistance to the Anglo-Russian invasion in 1911.", "after_revision": "He is known for using his position as a marja for a poignant political leadership in the Persian Constitutional Revolution (19051911), where he was one of the main clerical supporters of the revolution. He believed that \"constitutional form of government\" would be the best possible choice in the absence of Imam and regarded the constitutional revolution\" a Jihad (\"holy war\") in which all Muslims had to participate. He died \"suddenly\" and \"mysteriously\", when he aimed to leave Iraq for Iran in order to support constitutionalists' resistance to the Anglo-Russian invasion in 1911.", "edit_actions": [{"type": "R", "before": "political use", "after": "a poignant political leadership", "start_char_pos": 50, "end_char_pos": 63, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "\"Persian", "after": "constitutional", "start_char_pos": 312, "end_char_pos": 320, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "against", "after": "to", "start_char_pos": 480, "end_char_pos": 487, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}], "sents_char_pos": [0, 184, 394], "domain": "wiki"}
{"doc_id": "63624111", "revision_depth": 1, "before_revision": "The dwarf planet Pluto has an unusual set of climate zones, due to its atypical axial configuration. Five climate zones are assigned on the dwarf planet: tropics, arctic, tropical arctic, diurnal, and polar. These climate zones are delineated based on astronomically defined boundaries or sub-solar latitudes, which are not associated with the atmospheric circulations on the dwarf planet. Charon, the largest moon of Pluto, is tidally locked with it, and thus has the same climate zone structure as Pluto itself. 318x318px|Haze with multiple layers in the atmosphere of Pluto. Part of the plain Sputnik Planitia with nearby mountains is seen below. Photo by New Horizons, taken 15 min after the closest approach to Pluto. Pluto is an icy body, the most prominent object in the solar system's Kuiper belt. Its surface is primarily composed of methane (CH4), nitrogen (N2), and carbon monoxide (CO) volatile ices in various spatial abundances and distribution. Though Pluto is small compared to typical planets, it has an atmosphere, though much thinner than Earth's. Consisting of multiple layers of haze, its atmosphere is composed mainly of nitrogen (N2) with trace amount of methane (CH4) and carbon monoxide (CO). Long-term climate cycles of planetary bodies (e.g., Earth) are associated with the obliquity variation of the parent body, orbital eccentricitycycle, and axial precession cycles . However, in the case of Pluto, the orbital eccentricity cycle and axial precession cycles have less influence in the climate cycles of the dwarf planet than the obliquity variation. Consequently, the climate zones of Pluto were determined based on the obliquity variation only.", "after_revision": "The dwarf planet Pluto has an unusual set of climate zones, due to its atypical axial configuration. Five climate zones are assigned on the dwarf planet: tropics, arctic, tropical arctic, diurnal, and polar. These climate zones are delineated based on astronomically defined boundaries or sub-solar latitudes, which are not associated with the atmospheric circulations on the dwarf planet. Charon, the largest moon of Pluto, is tidally locked with it, and thus has the same climate zone structure as Pluto itself. 318x318px|Haze with multiple layers in the atmosphere of Pluto. Part of the plain Sputnik Planitia with nearby mountains is seen below. Photo by New Horizons, taken 15 min after the closest approach to Pluto. Pluto is an icy body, the most prominent object in the solar system's Kuiper belt. Its surface is primarily composed of methane (CH4), nitrogen (N2), and carbon monoxide (CO) volatile ices in various spatial abundances and distribution. Though Pluto is small compared to typical planets, it has an atmosphere, though much thinner than Earth's. Containing multiple layers of haze, its atmosphere is composed mainly of nitrogen (N2) with trace amount of methane (CH4) and carbon monoxide (CO). Long-term climate cycles of planetary bodies (e.g., Earth) are associated with axial precession and variations in the obliquity and orbital eccentricity . However, in the case of Pluto, the orbital eccentricity and axial precession cycles have less influence on the climate cycles of the dwarf planet than the obliquity variation. Consequently, the climate zones of Pluto were determined based on the obliquity variation only.", "edit_actions": [{"type": "R", "before": "Consisting of", "after": "Containing", "start_char_pos": 1067, "end_char_pos": 1080, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "the obliquity variation of the parent body, orbital eccentricitycycle, and axial precession cycles", "after": "axial precession and variations in the obliquity and orbital eccentricity", "start_char_pos": 1297, "end_char_pos": 1395, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "D", "before": "cycle", "after": null, "start_char_pos": 1454, "end_char_pos": 1459, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "in", "after": "on", "start_char_pos": 1508, "end_char_pos": 1510, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 100, 207, 389, 577, 649, 722, 805, 959, 1066, 1217, 1397, 1579], "domain": "wiki"}
{"doc_id": "63624111", "revision_depth": 2, "before_revision": "Tropical arctic The tropical arctic zone covers the overlapping regions of the tropical and arctic zones. This unique climatic zone is only possible on objects that have an obliquity range between 45 and 135. Since the obliquity of Pluto varies between a minimum of 103 and a maximum of 127, it therefore has a tropical arctic climate zone. This high obliquity implies that most of Pluto's surface experiences both tropical and arctic climates over the 2.8 million year obliquity cycle. The tropical arctic zone includes the region that overlaps between tropical and arctic zones extending between 13N to 77N and 13S to 77S. At this latitudinal range the zones cover 75\\% of the total surface area. However, like the oscillation of tropical (between 53 and 77 latitude) and arctic (between 13 and 37 latitude) zones, the tropical arctic zone also expands and contrasts over the course of the obliquity cycle. Only the latitude range between 37 to 57 in each hemisphere remains stable over the period, and therefore, those bands are called the permanent tropical arctic zones on Pluto (which covers 20\\% of the total area). Polar Pluto's polar zone comprises those areas where the Sun never reaches the zenith at any time of the orbital period of the dwarf planet over its entire obliquity cycle. This region always experiences arctic climate, never tropical climate, and it consistently experiences the longest duration of arctic winter and summer during each and every orbital rotation of Pluto. The polar zone has a radius of 13 from each pole. The zone spans from 77N to 90N and from 77S to 90S, encompassing an area total of 3\\% of the dwarf planet. The regions between 77 to 90 latitudes of each hemisphere are called the permanent range of the polar climate zone . However, the maximum extent of the polar zone can spans up to from 53N to 90N and from 53S to 90S with an area total of 20\\% of the body. At present day, the polar zones extend between 60 and 90 of both the northern and southern hemispheres (covering an area of 13\\% of the planet). Pluto's biggest moon Charon also has a polar zone consisting of a polar cap.", "after_revision": "Tropical arctic The tropical arctic zone covers the overlapping regions of the tropical and arctic zones. This unique climatic zone is only possible on objects that have an obliquity range between 45 and 135. Since the obliquity of Pluto varies between a minimum of 103 and a maximum of 127, it therefore has a tropical arctic climate zone. This high obliquity implies that most of Pluto's surface experiences both tropical and arctic climates over the 2.8 million year obliquity cycle. The tropical arctic zone includes the region that overlaps between tropical and arctic zones extending between 13N to 77N and 13S to 77S. At this latitudinal range the zones cover 75\\% of the total surface area. However, like the oscillation of tropical (between 53 and 77 latitude) and arctic (between 13 and 37 latitude) zones, the tropical arctic zone also expands and contrasts over the course of the obliquity cycle. Only the latitude range between 37 to 53 in each hemisphere remains stable over the period, and therefore, those bands are called the permanent tropical arctic zones on Pluto (which covers 20\\% of the total area). Polar Pluto's polar zone comprises those areas where the Sun never reaches the zenith at any time of the orbital period of the dwarf planet over its entire obliquity cycle. This region always experiences arctic climate, never tropical climate, and it consistently experiences the longest duration of arctic winter and summer during each and every orbital rotation of Pluto. The polar zone has a radius of 13 from each pole. The zone spans from 77N to 90N and from 77S to 90S, encompassing an area total of 3\\% of the dwarf planet. The regions between 77 to 90 latitudes of each hemisphere are called the permanent polar climate zones . However, the maximum extent of the polar zones can span from 53N to 90N and from 53S to 90S with an area total of 20\\% of the body. At present day, the polar zones extend between 60 and 90 of both the northern and southern hemispheres (covering an area of 13\\% of the planet). Pluto's biggest moon Charon also has a polar zone consisting of a polar cap.", "edit_actions": [{"type": "R", "before": "57", "after": "53", "start_char_pos": 947, "end_char_pos": 949, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "range of the polar climate zone", "after": "polar climate zones", "start_char_pos": 1737, "end_char_pos": 1768, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "zone can spans up to", "after": "zones can span", "start_char_pos": 1812, "end_char_pos": 1832, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}], "sents_char_pos": [0, 105, 208, 340, 486, 624, 698, 908, 1122, 1295, 1496, 1546, 1653, 1770, 1908, 2053], "domain": "wiki"}
{"doc_id": "63712897", "revision_depth": 1, "before_revision": "In 2011, Mercado accepted an offer to complete a PhD on Criminology at the Philippine College of Criminology , with his brother Marlon Lim . After being designated as the Vice President of Research, Extension, Planning and Development of the Polytechnic University of the Philippines, he attended an Executive Development Program held by the Commission on Higher Education (CHED) with other university administrators to further his knowledge on higher education management. Career Polytechnic University of the Philippines ( 2005present ) In 2004, Mercado first sat as the Acting Director of PUP Ragay. Consequently, in 2006, he was appointed as the director of PUP Sto. Tomas Batangas and then designated as the Dean of the College of Science. At the same time, he was also the Assistant Director of Center for Data and Statistical Analysis. Three years later, he became the Executive Director for Branches and Campuses. Mercado then sat as the Vice President for Branches and Campuses for another three years. In late 2015, Mercado was then appointed by Emanuel de Guzman, the university's President, to be the Vice President for Research, Extension, Planning and Development. To pursue personal endeavours, Mercado resigned from office in mid-2018. However, he still serves as a faculty member and professor in College of Science and teaches basic statistics, quantitative business analysis, management science statistics and other higher statistics subjects. He also manages several working groups and committees such as the Federasyon ng mga Alumni Associations sa PUP, Inc (FEDAAPI), Unyon ng mga guro sa PUP, etc .", "after_revision": "In 2011, Mercado accepted an offer to complete a PhD on Criminology at the Philippine College of Criminology . After being designated as the Vice President of Research, Extension, Planning and Development of the Polytechnic University of the Philippines, he attended an Executive Development Program held by the Commission on Higher Education (CHED) with other university administrators to further his knowledge on higher education management. Career Polytechnic University of the Philippines ( 20052018 ) In 2004, Mercado first sat as the Acting Director of PUP Ragay. Consequently, in 2006, he was appointed as the director of PUP Sto. Tomas Batangas and then designated as the Dean of the College of Science. At the same time, he was also the Assistant Director of Center for Data and Statistical Analysis. Three years later, he became the Executive Director for Branches and Campuses. Mercado then sat as the Vice President for Branches and Campuses for another three years. In late 2015, Mercado was then appointed by Emanuel de Guzman, the university's President, to be the Vice President for Research, Extension, Planning and Development. To pursue personal endeavours, Mercado resigned from office in mid-2018. He still serves as an expert to the university, specializing in Applied Statistics .", "edit_actions": [{"type": "D", "before": ", with his brother Marlon Lim", "after": null, "start_char_pos": 109, "end_char_pos": 138, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "2005present", "after": "20052018", "start_char_pos": 525, "end_char_pos": 536, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "However, he", "after": "He", "start_char_pos": 1252, "end_char_pos": 1263, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "a faculty member and professor in College of Science and teaches basic statistics, quantitative business analysis, management science statistics and other higher statistics subjects. He also manages several working groups and committees such as the Federasyon ng mga Alumni Associations sa PUP, Inc (FEDAAPI), Unyon ng mga guro sa PUP, etc", "after": "an expert to the university, specializing in Applied Statistics", "start_char_pos": 1280, "end_char_pos": 1619, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 140, 473, 602, 670, 744, 842, 921, 1011, 1178, 1251, 1462], "domain": "wiki"}
{"doc_id": "63849334", "revision_depth": 1, "before_revision": "Campaign against Armenia Till the summer of 34 BCE  Antony waited in Alexandria till he made his attack in early September. The Armenian troops attempted to put up resistance but were no match for the Romans. On September 18th Antonys troops entered the capital and captured Artavasdes II . He was then brought to Alexandria and was kept imprisoned by Cleoptatra till he was executed in 31 BCE. Rome maintained control over Armenia till the Final war of the Roman Republic. In 20 BCE  Augustus negotiated a peace treaty between Rome and Parthia  which led to Armenia becoming a buffer state between the two .", "after_revision": "Campaign Until the summer of 34 BCE , Antony waited in Alexandria , and he made his attack in early September. The Armenian troops attempted to put up resistance but were no match for the Romans. On September 18, Antony's troops entered the capital and captured Artavasdes II , who was then brought to Alexandria and was kept imprisoned by Cleoptatra until he was executed in 31 BCE. Rome maintained control over Armenia till the Final war of the Roman Republic. In 20 BCE , Augustus negotiated a peace treaty between Rome and Parthia , which led to Armenia becoming a buffer state between them .", "edit_actions": [{"type": "R", "before": "against Armenia Till", "after": "Until", "start_char_pos": 9, "end_char_pos": 29, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 51, "end_char_pos": 51, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "till", "after": ", and", "start_char_pos": 80, "end_char_pos": 84, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "18th Antonys", "after": "18, Antony's", "start_char_pos": 222, "end_char_pos": 234, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "R", "before": ". He", "after": ", who", "start_char_pos": 289, "end_char_pos": 293, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "till", "after": "until", "start_char_pos": 363, "end_char_pos": 367, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 484, "end_char_pos": 484, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 545, "end_char_pos": 545, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "the two", "after": "them", "start_char_pos": 599, "end_char_pos": 606, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 123, 208, 394, 473], "domain": "wiki"}
{"doc_id": "66550186", "revision_depth": 1, "before_revision": "Farzi refers to an outer garment of the Mughal court. It was a short coat with half sleeves and fur collars, opened in front similar to Jama. Farzi was essentially fastened with a decorated piece of cloth, i.e., Katzeb around the waist area with loosely hanging ends. Farzi was a winter's garment. Mughal emperors and courtiers were wearing it. Farzi was one of the costumes given in Khilat ( awards to the Mughal nobles and other courtiers ) . The coat was very much famous in the 17th century among the royals.", "after_revision": "Farzi refers to an outer garment of the Mughal court. It was a coat with short sleeves and fur collars, opened in front . The length was shorter than Jama. Farzi was a winter's garment. Mughal emperors and courtiers were wearing it over the Jama, fastened with a decorated piece of cloth, i.e., Katzeb around the waist area with loosely hanging ends. Farzi was one of the costumes given in Khilat ( rewards of robes) to the Mughal nobles and other courtiers . The coat was very much famous in the 17th century among the royals.", "edit_actions": [{"type": "R", "before": "short coat with half", "after": "coat with short", "start_char_pos": 63, "end_char_pos": 83, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "similar to", "after": ". The length was shorter than", "start_char_pos": 125, "end_char_pos": 135, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "meaning-changed"]}, {"type": "R", "before": "essentially", "after": "a winter's garment. Mughal emperors and courtiers were wearing it over the Jama,", "start_char_pos": 152, "end_char_pos": 163, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "D", "before": "a winter's garment. Mughal emperors and courtiers were wearing it. Farzi was", "after": null, "start_char_pos": 278, "end_char_pos": 354, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "awards", "after": "rewards of robes)", "start_char_pos": 393, "end_char_pos": 399, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "D", "before": ")", "after": null, "start_char_pos": 441, "end_char_pos": 442, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 53, 141, 267, 297, 344, 444], "domain": "wiki"}
{"doc_id": "66915446", "revision_depth": 1, "before_revision": "Abduljabbar Nasuru Kabara (born 7 February 1970) is an Islamic cleric and a scholar based in Kano, Nigeria.  He is the son of Nasuru Kabara, a formal Qadiriyya leader in Africa.", "after_revision": "Abduljabbar Nasuru Kabara (born 7 February 1970) is a controversial Islamic cleric and a scholar based in Kano, Nigeria. Who was accused of blasphemy towards prophet Muhammad, He is the son of Nasuru Kabara, the formal Qadiriyya leader of West Africa.", "edit_actions": [{"type": "R", "before": "an", "after": "a controversial", "start_char_pos": 52, "end_char_pos": 54, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "style"]}, {"type": "A", "before": null, "after": "Who was accused of blasphemy towards prophet Muhammad,", "start_char_pos": 108, "end_char_pos": 108, "major_intent": "meaning-changed", "raw_intents": ["coherence", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "a", "after": "the", "start_char_pos": 141, "end_char_pos": 142, "major_intent": "fluency", "raw_intents": ["style", "fluency", "fluency"]}, {"type": "R", "before": "in", "after": "of West", "start_char_pos": 167, "end_char_pos": 169, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}], "sents_char_pos": [0, 107], "domain": "wiki"}
{"doc_id": "6704145", "revision_depth": 1, "before_revision": "As a crew member, it is to be noticed that a Wiper may undergo a condition known as burnout syndrome . This is where a seafarer may be overcome by increased anxiety and stress by means of poor leadership, long hours, and high-risk applications. The implications of work-related stress stemming from seafaring operations can be largely alleviated through clear communication of crew and leadership and occasionally moderation through a human resources department. For this reason, it is of medical importance that a wiper finds employment where the safety and well being of workers is held to a high standard. Seafarers and labor unions, necessitated from this occurrence, are closely related in this nature.", "after_revision": "A Wiper may undergo burnout syndrome and be overcome by increased anxiety and stress due to leadership, long hours, and high-risk applications. Work-related stress stemming from seafaring operations can be largely alleviated through clear communication between crew and leadership and occasionally moderation through a human resources department. For this reason, it is of medical importance that a wiper finds employment where the safety and well being of workers is held to a high standard. Seafarers and labor unions, necessitated from this occurrence, are closely related in this nature.", "edit_actions": [{"type": "R", "before": "As a crew member, it is to be noticed that a", "after": "A", "start_char_pos": 0, "end_char_pos": 44, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "a condition known as burnout syndrome . This is where a seafarer may", "after": "burnout syndrome and", "start_char_pos": 63, "end_char_pos": 131, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "by means of poor", "after": "due to", "start_char_pos": 176, "end_char_pos": 192, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "The implications of work-related", "after": "Work-related", "start_char_pos": 245, "end_char_pos": 277, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "of", "after": "between", "start_char_pos": 374, "end_char_pos": 376, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 102, 244, 462, 608], "domain": "wiki"}
{"doc_id": "67329715", "revision_depth": 1, "before_revision": "Empress Gi was one of those Kongnyo women, and when she became empress of Yuan, she started to recruit many Goryeo court maids in the palace. The late Yuan dynasty poet Zhang Xu wrote a few poems about the popularity of Goryeo-style, which was dubbed as Goryeoyang (). At the end of the Yuan dynasty, the Goryeoyang clothing is described as a short, banryong (, square collar) banbi (, a short sleeve upper garment) which passed beneath the waist. However, so far, the modern interpretation on the appearance of Mongol royal women's clothing influenced by Goryeoyang is based on authors' suggestions. Moreover, the possibility that goryeoyang  continued to influence  the clothing of the Ming dynasty after the fall of the Yuan dynastyis also based on speculations and is a topic which needs to investigated further.", "after_revision": "Empress Gi was one of those Kongnyo women, and when she became empress of Yuan, she started to recruit many Goryeo court maids in the palace. The late Yuan dynasty poet Zhang Xu wrote a few poems about the popularity of Goryeo-style, which was dubbed as Goryeoyang (). At the end of the Yuan dynasty, the Goryeoyang clothing is described as a short, banryong (, square collar) banbi (, a short sleeve upper garment) which passed beneath the waist. However, so far, the modern interpretation on the appearance of Mongol royal women's clothing influenced by Goryeoyang is based on authors' suggestions. According to Hyunhee Park: \"Like the Mongolian style, it is possible that this Kory style Kory yang continued to influence some Chinese in the Ming period after the Ming dynasty replaced the Yuan dynasty, a topic to investigate further.\"", "edit_actions": [{"type": "R", "before": "Moreover, the possibility that goryeoyang", "after": "According to Hyunhee Park: \"Like the Mongolian style, it is possible that this Kory style", "start_char_pos": 601, "end_char_pos": 642, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Kory yang", "start_char_pos": 643, "end_char_pos": 643, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "some Chinese in the Ming period after", "start_char_pos": 667, "end_char_pos": 667, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "clothing of the", "after": null, "start_char_pos": 672, "end_char_pos": 687, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "after the fall of the Yuan dynastyis also based on speculations and is a topic which needs to investigated further.", "after": "replaced the Yuan dynasty, a topic to investigate further.\"", "start_char_pos": 701, "end_char_pos": 816, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 141, 268, 447, 600], "domain": "wiki"}
{"doc_id": "6795857", "revision_depth": 1, "before_revision": "The Shanghai massacre of 12 April 1927, the April 12 Purge or the April 12 Incident as it is commonly known in China, was the violent suppression of Communist Party of China URLanizations  in Shanghai by the military forces of Generalissimo Chiang Kai-shek and conservative factions in the Kuomintang (Nationalist Party, or KMT). Following the incident, conservative KMT elements carried out a full-scale purge of Communists in all areas under their control, and even more violent suppression occurred in Guangzhou and Changsha.Wilbur, Nationalist Revolution 114 The purge led to an open split between left and right wing factions in the KMT, with Chiang Kai-shek establishing himself as the leader of the right wing faction based in Nanjing, in opposition to the original left-wing KMT government based in Wuhan led by Wang Jingwei. By 15 July 1927, the Wuhan regime had also expelled the Communists in its ranks, effectively ending the KMT's four-year alliance with the Comintern and the Chinese Communist Party . For the remainder of 1927, the CPC launched several uprisings in an attempt to regain their previous power, marking the beginning of the Chinese Civil War. With the failure of the Guangzhou Uprising however, the Communist Party was largely diminished in their influence , unable to launch another major urban offensive.Wilbur, Nationalist Revolution 170.", "after_revision": "The Shanghai massacre of 12 April 1927, the April 12 Purge or the April 12 Incident as it is commonly known in China, was the violent suppression of Communist Party of China URLanizations and leftist elements in Shanghai by forces supporting General Chiang Kai-shek and conservative factions in the Kuomintang (Nationalist Party, or KMT). Following the incident, conservative KMT elements carried out a full-scale purge of Communists in all areas under their control, and violent suppression occurred in Guangzhou and Changsha.Wilbur, Nationalist Revolution 114 The purge led to an open split between left and right wing factions in the KMT, with Chiang Kai-shek establishing himself as the leader of the right wing faction based in Nanjing, in opposition to the original left-wing KMT government based in Wuhan led by Wang Jingwei. By 15 July 1927, the Wuhan regime had also expelled the Communists in its ranks, effectively ending the First United Front, a working alliance of the KMT and the CPC under the tutelage of Comintern agents . For the remainder of 1927, the CPC launched uprisings to regain power, marking the beginning of the Chinese Civil War. With the failure of the Guangzhou Uprising however, the Communist power was largely diminished , unable to launch another major urban offensive.Wilbur, Nationalist Revolution 170.", "edit_actions": [{"type": "A", "before": null, "after": "and leftist elements", "start_char_pos": 188, "end_char_pos": 188, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "the military forces of Generalissimo", "after": "forces supporting General", "start_char_pos": 204, "end_char_pos": 240, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "even more", "after": null, "start_char_pos": 463, "end_char_pos": 472, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "KMT's four-year alliance with the Comintern and the Chinese Communist Party", "after": "First United Front, a working alliance of the KMT and the CPC under the tutelage of Comintern agents", "start_char_pos": 938, "end_char_pos": 1013, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "several uprisings in an attempt to regain their previous", "after": "uprisings to regain", "start_char_pos": 1060, "end_char_pos": 1116, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "Party", "after": "power", "start_char_pos": 1238, "end_char_pos": 1243, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "in their influence", "after": null, "start_char_pos": 1267, "end_char_pos": 1285, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 329, 528, 833, 1015, 1171, 1335], "domain": "wiki"}
{"doc_id": "68248886", "revision_depth": 1, "before_revision": "Fidelis Thaliath (1929 2008), popularly known as Doctoramma (Mother Doctor), was an Indian nun, gynecologist and social reformer who worked among the destitute people in Delhi. She was declared Servant of God by the Holy See in 2021. Early years Fidelis Thaliath was born  in 1929 in Puthanpally, a village in Ernakulam district of the south Indian state of Kerala to Joseph Thaliath and Mariakkutty . She joined the  Sisters of the Destitute movement in 1952 and took the oath of celibacy in 1956. In 1964, she moved to Chicago where she joined the Loyola University to obtain a medical degree in Gynecology. Subsequently, she returned to India to start her career as a medical professional in Holy Angels' Nursing Home in Delhi but returned to Chicago in 1973 to complete her master's degree in 1977. Social career Returning to India in 1977, Thaliath joined the efforts of Varghese Payyappilly Palakkappilly, the founder of the Sisters of the Destitute movement, to found the Jivodya Hospital, which was built on a plot of land in Ashok Vihar in Delhi, bought with a gift of 1000 she had received for purchasing a car and served there in the Department of Gynecology. During her time in Delhi, she was involved in several social activities, focusing on helping the destitute people and founded two centres, one to house the destitute women in Vikaspuri and the other, a home for the handicapped children in Ghaziabad.  Thaliath died on January 17, 2008, at the age of 79. Loyola University Chicago, her alma mater, have since instituted a scholarship, Class Of 1960 Sister Fidelis Thaliath, MD, Medical Student Scholarship Fund, in her honor. Canonization Thaliath, who was popularly known as Doctoramma (Mother Doctor), is reported to have performed several miracles which are being considered and verified by Roman Catholic Church and her social and religious contributions were considered by the Church while deciding upon initiation of her canonization process. As the first step towards canonization, she was declared Servant of God by the Roman Catholic Church on 14 July 2021.", "after_revision": "Fidelis Thaliath (1929 2008), popularly known as Doctoramma (Mother Doctor), was an Indian nun, gynecologist and social reformer who worked among the destitute people in Delhi. She was declared Servant of God in 2021. Early years Fidelis Thaliath was born as Kochuthresia in 1929 in Puthanpally, a village in Ernakulam district of the south Indian state of Kerala to Joseph Thaliath and Mariamkkutty . She joined the religious congregation, Sisters of the Destitute in 1952 and took the perpetual vow of chastity in 1956. In 1964, she moved to Chicago where she joined the Loyola University to obtain a medical degree in Gynecology. Subsequently, she returned to India to start her career as a medical professional in Holy Angels' Nursing Home in Delhi but returned to Chicago in 1973 to complete her master's degree in 1977. Social career Returning to India in 1977, she founded the Jivodya Hospital, which was built on a plot of land in Ashok Vihar in Delhi, bought with a gift of 1000 she had received for purchasing a car and served there in the Department of Gynecology. During her time in Delhi, she was involved in several social activities, focusing on helping the destitute people and founded two centres, one to house the poor women in Vikaspuri and the other, a home for the handicapped children in Ghaziabad. Spirituality Hers was a life set aside for the underprivileged and the needy. Nourished by the sacrament of eucharist and rosary in her hands, she walked through the streets of Delhi, caring for the sick and the abandoned. She not only took care of them and provided medicines, but also prayed for them. Her selfless service brought meaning to the lives of so many poor patients, especially the lepers. Thaliath died on January 17, 2008, at the age of 79. Loyola University Chicago, her alma mater, have since instituted a scholarship, Class Of 1960 Sister Fidelis Thaliath, MD, Medical Student Scholarship Fund, in her honor. Cause of Canonization Thaliath is reported to have performed several miracles which are being considered and verified by Roman Catholic Church and her social and religious contributions were considered by the Church while deciding upon initiation of her canonization process. As the first step towards canonization, she was declared Servant of God by the Roman Catholic Church on 14 July 2021 in the archdiocese of Faridabad by the Archbishop Mar Kuriakose Bharanikulangara.", "edit_actions": [{"type": "D", "before": "by the Holy See", "after": null, "start_char_pos": 209, "end_char_pos": 224, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "as Kochuthresia", "start_char_pos": 272, "end_char_pos": 272, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Mariakkutty", "after": "Mariamkkutty", "start_char_pos": 388, "end_char_pos": 399, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "A", "before": null, "after": "religious congregation,", "start_char_pos": 417, "end_char_pos": 417, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "movement", "after": null, "start_char_pos": 443, "end_char_pos": 451, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "oath of celibacy", "after": "perpetual vow of chastity", "start_char_pos": 473, "end_char_pos": 489, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Thaliath joined the efforts of Varghese Payyappilly Palakkappilly, the founder of the Sisters of the Destitute movement, to found the", "after": "she founded the", "start_char_pos": 845, "end_char_pos": 978, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "R", "before": "destitute", "after": "poor", "start_char_pos": 1327, "end_char_pos": 1336, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Spirituality Hers was a life set aside for the underprivileged and the needy. Nourished by the sacrament of eucharist and rosary in her hands, she walked through the streets of Delhi, caring for the sick and the abandoned. She not only took care of them and provided medicines, but also prayed for them. Her selfless service brought meaning to the lives of so many poor patients, especially the lepers.", "start_char_pos": 1421, "end_char_pos": 1421, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Canonization Thaliath, who was popularly known as Doctoramma (Mother Doctor),", "after": "Cause of Canonization Thaliath", "start_char_pos": 1646, "end_char_pos": 1723, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "2021.", "after": "2021 in the archdiocese of Faridabad by the Archbishop Mar Kuriakose Bharanikulangara.", "start_char_pos": 2081, "end_char_pos": 2086, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "fluency"]}], "sents_char_pos": [0, 176, 233, 401, 498, 609, 802, 1170, 1420, 1474, 1645, 1968], "domain": "wiki"}
{"doc_id": "6853425", "revision_depth": 1, "before_revision": "URLe Lilanga (1934 27 June 2005) was a Tanzanian painter and sculptor, active from the late 1970s and until the early years of the 21st century. He belonged to the Makonde people and lived most of his life in Dar es Salaam, the largest city in Tanzania. The main gate of the Nyumba ya Sanaa cultural centre in Dar es Salaam has decorations by Lilanga URLe Lilanga (right) and fellow painter Patrick Francis Imanjama outside Nyumba ya Sanaa In 1970, Lilanga decided to move to Dar es Salaam, where there were greater opportunities for selling sculptures and practising contemporary art. In 1971, he found his first employment, thanks to his uncle, Augustino Malaba, an already well-known sculptor, who would be his future collaborator. He worked as a night guard at the House of Art (Nyumba ya Sanaa), a  centre for the development of African art and craftsmanship in Dar es Salaam. Lilanga's talents were soon recognized by Jean Pruitt, an American aid worker promoting Tanzanian art. Lilanga welcomed  other artists like Robino Ntila, Augustino Malaba and Patrick Francis Imanjama to join Nyumba ya Sanaa . He also began to create paintings on batik cloth, on goatskin and on metal sheets for the finishing of railings and gates of the centre.  In 1977, he made his first journey outside Africa, travelling to New York, where he had a show at the Maryknoll Sisters Ossining Center . He stayed for a brief time in Manhattan, selling prints made on paper or cardboard, standing on street corners. Artistic development & success Shortly after beginning grammar school, Lilanga had his first contact with sculpture ( roots, softwood and, later, hard ebony), working in the Makonde tradition. He dedicated himself almost exclusively to this technique from 1961 until 1972. He showed his first works to Europeans who worked in the refugee camps during Mozambique's war of independence. Following their advice, in 1970 Lilanga decided to move to Dar es Salaam, where there were greater opportunities for selling sculptures. A lot of URLe Lilanga's lifetime art works can be found in leading international collections like The Contemporary African Art Collection (CAAC) of Jean Pigozzi or the Hamburg Mawingu Collection HMC of German collector Peter-Andreas Kamphausen. Directly after the death of the artist, the HMC: URLe Lilanga Collection presented the first systematically and thematically complete catalogue of Lilanga's work. In addition, the book explores the traditional roots of East African Makonde art, as well as four decades of Lilanga's artistic development with different materials and techniques, including sculpture, paintings, etchings, drawings and metal works. Currently the HMC publish the URLe Lilanga News as an online blog with information about the artist, his art works, exhibitions and other interesting background material. Also in 2005, the African Collection series (Skira Editore, Milan) published a well-illustrated book on Lilanga's work with useful information.", "after_revision": "URLe Lilanga (1934 27 June 2005) was a Tanzanian painter and sculptor, active from the late 1970s and until the early 21st century. He belonged to the Makonde people and lived most of his life in Dar es Salaam, the largest city in Tanzania. The main gate of the Nyumba ya Sanaa cultural centre in Dar es Salaam with sculptures by Lilanga URLe Lilanga (right) and fellow painter Patrick Francis Imanjama outside Nyumba ya Sanaa In 1970, Lilanga decided to move to Dar es Salaam, where there were greater opportunities for selling sculptures and practising contemporary art. In 1971, he found his first employment, thanks to his uncle, Augustino Malaba, an already well-known sculptor, who would be his future collaborator. He started as a night guard at the House of Art (Nyumba ya Sanaa), a cultural centre for the development of African art and craftsmanship in Dar es Salaam. In 1982, he created wooden doors with clourful shetani figures for this cultural centre. After the Nyumba ya Sanaa was destroyed, these doors were bought and restored by a German collector and are documented in the book URLe Lilanga: The Doors of Nyumba ya Sanaa. As a modern artist in his own right, he joined other artists like Robino Ntila, Augustino Malaba and Patrick Francis Imanjama . He also began to create paintings on batik cloth, on goatskin and on metal sheets for the finishing of railings and gates of the centre. Lilanga's talents were recognized by Maryknoll Sister Jean Pruitt, an American aid worker promoting Tanzanian art. In 1977, he made his first journey outside Africa, travelling to New York, where he had a show at the Maryknoll Sisters ' Ossining Center in New York State . He stayed for a brief time in Manhattan, selling prints made on paper or cardboard, standing on street corners. Artistic development & success Shortly after beginning grammar school, Lilanga had his first contact with wooden sculpture, made from roots, softwood and, later, ebony, and working in the Makonde tradition. He dedicated himself almost exclusively to this technique from 1961 until 1972. He showed his first works to Europeans who worked in the refugee camps during Mozambique's war of independence. Following their advice, in 1970 Lilanga decided to move to Dar es Salaam, where there were greater opportunities for selling sculptures. URLe Lilanga's art works can be found in leading international collections like The Contemporary African Art Collection (CAAC) of Jean Pigozzi or the Hamburg Mawingu Collection (HMC) of German collector Peter-Andreas Kamphausen. Shortly after the death of the artist, the URLe Lilanga Collection presented the first systematically and thematically complete catalogue of Lilanga's work. In addition, the book explores the traditional roots of East African Makonde art, as well as four decades of Lilanga's artistic development with different materials and techniques, including sculpture, paintings, etchings, drawings and metal works. Currently the HMC publish the URLe Lilanga News as an online blog with information about the artist, his art works, exhibitions and other interesting background material. Also in 2005, the African Collection series (Skira Editore, Milan) published a well-illustrated book on Lilanga's work with useful information.", "edit_actions": [{"type": "D", "before": "years of the", "after": null, "start_char_pos": 118, "end_char_pos": 130, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "has decorations", "after": "with sculptures", "start_char_pos": 324, "end_char_pos": 339, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "worked", "after": "started", "start_char_pos": 738, "end_char_pos": 744, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "cultural", "start_char_pos": 803, "end_char_pos": 803, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "Lilanga's talents were soon recognized by Jean Pruitt, an American aid worker promoting Tanzanian art. Lilanga welcomed", "after": "In 1982, he created wooden doors with clourful shetani figures for this cultural centre. After the Nyumba ya Sanaa was destroyed, these doors were bought and restored by a German collector and are documented in the book URLe Lilanga: The Doors of Nyumba ya Sanaa.", "start_char_pos": 882, "end_char_pos": 1001, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "As a modern artist in his own right, he joined", "start_char_pos": 1002, "end_char_pos": 1002, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "D", "before": "to join Nyumba ya Sanaa", "after": null, "start_char_pos": 1082, "end_char_pos": 1105, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Lilanga's talents were recognized by Maryknoll Sister Jean Pruitt, an American aid worker promoting Tanzanian art.", "start_char_pos": 1245, "end_char_pos": 1245, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Ossining Center", "after": "' Ossining Center in New York State", "start_char_pos": 1366, "end_char_pos": 1381, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "sculpture (", "after": "wooden sculpture, made from", "start_char_pos": 1602, "end_char_pos": 1613, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "others"]}, {"type": "R", "before": "hard ebony),", "after": "ebony, and", "start_char_pos": 1642, "end_char_pos": 1654, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "D", "before": "A lot of", "after": null, "start_char_pos": 2018, "end_char_pos": 2026, "major_intent": "coherence", "raw_intents": ["clarity", "coherence", "coherence"]}, {"type": "D", "before": "lifetime", "after": null, "start_char_pos": 2042, "end_char_pos": 2050, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "HMC", "after": "(HMC)", "start_char_pos": 2213, "end_char_pos": 2216, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}, {"type": "R", "before": "Directly", "after": "Shortly", "start_char_pos": 2263, "end_char_pos": 2271, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "clarity"]}, {"type": "D", "before": "HMC:", "after": null, "start_char_pos": 2307, "end_char_pos": 2311, "major_intent": "fluency", "raw_intents": ["others", "fluency", "fluency"]}], "sents_char_pos": [0, 144, 253, 585, 734, 881, 984, 1107, 1244, 1383, 1495, 1688, 1768, 1880, 2017, 2262, 2425, 2674, 2845], "domain": "wiki"}
{"doc_id": "6857524", "revision_depth": 1, "before_revision": "Salimuzzaman Siddiqui, ( ; 19 October 1897 14 April 1994) was a URLanic chemist specialising in natural products, and a professor of the chemistry at the University of Karachi. Siddiqui studied philosophy at Aligarh Muslim University and later studied chemistry at Frankfurt University, where he received his PhD in 1927. On return to British India, he worked at the Tibbia College Delhi and the Indian Council for Scientific and Industrial Research. He later moved to Pakistan and worked in the Pakistan Council of Scientific and Industrial Research and went on to establish the Pakistan National Science Council and was appointed as its first chairman in 1961. In the same year he became the Fellow of the Royal Society. He later also co-founded  Pakistan Academy of Sciences, and after retirement from the government, he founded the Hussain Ebrahim Jamal Research Institute of Chemistry. Siddiqui is credited for pioneering the isolation of unique chemical compounds from the Neem (Azadirachta indica), Rauvolfia, and various other flora. As the founder director of H.E.J. Research Institute of Chemistry, he revolutionised the research on pharmacology of various domestic plants found in South Asia to extract novel chemical substances of medicinal importance.Akhtar (1996), pp. 400417 During his career, Siddiqui published more than 300 research papers and obtained 40 patents mainly from the field of natural product chemistry. In addition to his scientific talents, Siddiqui was also an avid painter, a poet, and a great connoisseur of Western music. His paintings were exhibited in the United States, Germany, India, and Pakistan.", "after_revision": "Salimuzzaman Siddiqui, ( ; 19 October 1897 14 April 1994) was a URLanic chemist specialising in natural products, and a professor of chemistry at the University of Karachi. Siddiqui studied philosophy at Aligarh Muslim University and later studied chemistry at Frankfurt University, where he received his PhD in 1927. On return to British India, he worked at the Tibbia College Delhi and the Indian Council for Scientific and Industrial Research. He later moved to Pakistan and worked in the Pakistan Council of Scientific and Industrial Research . He went on to establish the Pakistan National Science Council and was appointed as its first chairman in 1961. In the same year he became a Fellow of the Royal Society. He later co-founded the Pakistan Academy of Sciences, and after retirement from the government, he founded the Hussain Ebrahim Jamal Research Institute of Chemistry. Siddiqui is credited for pioneering the isolation of unique chemical compounds from the Neem (Azadirachta indica), Rauvolfia, and various other flora. As the founder director of H.E.J. Research Institute of Chemistry, he revolutionised research of the pharmacology of various domestic plants found in South Asia to extract novel chemical substances of medicinal importance.Akhtar (1996), pp. 400417 During his career, Siddiqui published more than 300 research papers and obtained 40 patents mainly from the field of natural product chemistry. In addition to his scientific talents, Siddiqui was also an avid painter, poet, and a great connoisseur of Western music. His paintings were exhibited in the United States, Germany, India, and Pakistan.", "edit_actions": [{"type": "D", "before": "the", "after": null, "start_char_pos": 133, "end_char_pos": 136, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "and", "after": ". He", "start_char_pos": 551, "end_char_pos": 554, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "R", "before": "the", "after": "a", "start_char_pos": 690, "end_char_pos": 693, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "D", "before": "also", "after": null, "start_char_pos": 732, "end_char_pos": 736, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}, {"type": "A", "before": null, "after": "the", "start_char_pos": 748, "end_char_pos": 748, "major_intent": "fluency", "raw_intents": ["fluency", "others", "fluency"]}, {"type": "R", "before": "the research on", "after": "research of the", "start_char_pos": 1127, "end_char_pos": 1142, "major_intent": "clarity", "raw_intents": ["clarity", "others", "style"]}, {"type": "D", "before": "a", "after": null, "start_char_pos": 1508, "end_char_pos": 1509, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}], "sents_char_pos": [0, 26, 176, 321, 450, 662, 722, 890, 1041, 1264, 1433, 1557], "domain": "wiki"}
{"doc_id": "6970425", "revision_depth": 1, "before_revision": "In matrilineal succession (also known as Marumakathayam), practiced in Kerala by the Nair nobility and royal families, a man's wealth and title is inherited by his sister's children, and his own children receive their inheritance from their own maternal uncles.  The Maharajah of Travancore is therefore succeeded by his sister's son, and his own son receives a courtesy title but has no place in the line of succession. Since Indian Independence and the passing of several acts such as the Hindu Succession Act (1956), this form of inheritance is no longer recognised by law. Regardless, the pretender to the Travancore throne is still determined by matrilinear succession. The Akans of Ghana and the Ivory Coast, West Africa have similar matrilineal succession and as such Otumfour Osei-Tutu II, Asantehene inherited the Golden Stool  through his mother (the Asantehemaa) Nana Afia Kobi Serwaa Ampem II.", "after_revision": "Some cultures pass honours down through the female line. A man's wealth and title are inherited by his sister's children, and his children receive their inheritance from their maternal uncles. In Kerala, southern India, this custom is known as Marumakkathayam. It is practiced by the Nair nobility and royal families. The Maharajah of Travancore is therefore succeeded by his sister's son, and his own son receives a courtesy title but has no place in the line of succession. Since Indian Independence and the passing of several acts such as the Hindu Succession Act (1956), this form of inheritance is no longer recognised by law. Regardless, the pretender to the Travancore throne is still determined by matrilinear succession. The Akans of Ghana and the Ivory Coast, West Africa have similar matrilineal succession and as such Otumfour Osei-Tutu II, Asantehene inherited the Golden Stool (the throne) through his mother (the Asantehemaa) Nana Afia Kobi Serwaa Ampem II.", "edit_actions": [{"type": "R", "before": "In matrilineal succession (also known as Marumakathayam), practiced in Kerala by the Nair nobility and royal families, a", "after": "Some cultures pass honours down through the female line. A", "start_char_pos": 0, "end_char_pos": 120, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": "is", "after": "are", "start_char_pos": 144, "end_char_pos": 146, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "coherence"]}, {"type": "D", "before": "own", "after": null, "start_char_pos": 191, "end_char_pos": 194, "major_intent": "clarity", "raw_intents": ["clarity", "fluency", "clarity"]}, {"type": "D", "before": "own", "after": null, "start_char_pos": 241, "end_char_pos": 244, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "In Kerala, southern India, this custom is known as Marumakkathayam. It is practiced by the Nair nobility and royal families.", "start_char_pos": 262, "end_char_pos": 262, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "meaning-changed"]}, {"type": "A", "before": null, "after": "(the throne)", "start_char_pos": 836, "end_char_pos": 836, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 261, 420, 576, 674], "domain": "wiki"}
{"doc_id": "6970425", "revision_depth": 2, "before_revision": "Lateral succession Lateral or fraternal system of succession mandates principles of seniority among members of a dynasty or dynastic clan, with a purpose of election a best qualified candidate for the leadership. The leaders are elected as being the most mature elders of the clan, already in possession of military power and competence. Fraternal succession is preferred to ensure that mature leaders are in charge, removing a need for regents. The lateral system of succession may or may not exclude male descendants in the female line from succession. In practice, when no male heir is mature enough, a female heir is usually determined \"pragmatically\", by proximity to the last monarch, like Boariks of the Caucasian Huns or Tamiris of Massagetes in Middle Asia  . The lateral monarch is generally elected after the leadership throne becomes vacant. In the early years of the Mongol empire, the death of the ruling monarchs, Genghis Khan and gedei Khan, immediately stopped the Mongol western campaigns because of the upcoming elections. A drawback of the lateral succession is that, while ensuring a most competent leadership for the moment, the system inherently created derelict prince lines not eligible for succession. Any scion of an eligible heir that did not live long enough to ascend to the throne was cast aside as not eligible, creating a pool of discontented pretenders called Tegin in Turkic and Izgoi in Rus dynastic lines. The unsettled pool of derelict princes would eventually bring havoc to the succession order and dismemberment to the state.", "after_revision": "Lateral succession Lateral or fraternal system of succession mandates principles of seniority among members of a dynasty or dynastic clan, with a purpose of election a best qualified candidate for the leadership. The leaders are elected as being the most mature elders of the clan, already in possession of military power and competence. Fraternal succession is preferred to ensure that mature leaders are in charge, removing a need for regents. The lateral system of succession may or may not exclude male descendants in the female line from succession. In practice, when no male heir is mature enough, a female heir is usually determined \"pragmatically\", by proximity to the last monarch, like Boariks of the Caucasian Huns or Tamiris of Massagetes in Middle Asia were selected . The lateral monarch is generally elected after the leadership throne becomes vacant. In the early years of the Mongol empire, the death of the ruling monarchs, Genghis Khan and gedei Khan, immediately stopped the Mongols' western campaigns because of the upcoming elections. A drawback of the lateral succession is that, while ensuring a most competent leadership for the moment, the system inherently created derelict princely lines not eligible for succession. Any scion of an eligible heir who did not live long enough to ascend to the throne was cast aside as not eligible, creating a pool of discontented pretenders called Tegin in Turkic and Izgoi in Rus dynastic lines. The unsettled pool of derelict princes would eventually bring havoc to the succession order and dismemberment to the state.", "edit_actions": [{"type": "A", "before": null, "after": "were selected", "start_char_pos": 766, "end_char_pos": 766, "major_intent": "clarity", "raw_intents": ["style", "clarity", "clarity"]}, {"type": "R", "before": "Mongol", "after": "Mongols'", "start_char_pos": 982, "end_char_pos": 988, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "prince", "after": "princely", "start_char_pos": 1186, "end_char_pos": 1192, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "that", "after": "who", "start_char_pos": 1258, "end_char_pos": 1262, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 212, 337, 445, 554, 768, 853, 1041, 1227, 1442], "domain": "wiki"}
{"doc_id": "7205879", "revision_depth": 1, "before_revision": "250px|Ziwa ruins, enclosure view from a distance. Ziwa, or Ziwa ruinsformerly Van Niekerk ruins after Major Pompey Van Niekerk who was second in command of the Afrikaner Corps is  the remains of a vast late Iron Age agricultural settlement that has been dated to the 15th century. The site is located in Nyanga District. Ziwa was declared a National Monument in 1946 and is currently under consideration for World Heritage listing. The site contains a large variety of stonework structures including stone terraces running along contours of hills and steep landscapes. Archaeological investigations have also engendered important aspects of pottery and rock art.", "after_revision": "250px|Ziwa ruins, enclosure view from a distance. Ziwaformerly Van Niekerk ruins after Major Pompey Van Niekerk who was second in command of the Afrikaner Corps is an archaeological site in Nyanga District, Zimbabwe, containing the remains of a vast late Iron Age agricultural settlement dated to the 15th century. Ziwa was declared a National Monument in 1946 and is currently under consideration for World Heritage listing. The site contains a large variety of stonework structures including stone terraces running along contours of hills and steep landscapes. Archaeological investigations have also uncovered important aspects of pottery and rock art.", "edit_actions": [{"type": "R", "before": "Ziwa, or Ziwa ruinsformerly", "after": "Ziwaformerly", "start_char_pos": 50, "end_char_pos": 77, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "style", "others"]}, {"type": "A", "before": null, "after": "an archaeological site in Nyanga District, Zimbabwe, containing", "start_char_pos": 179, "end_char_pos": 179, "major_intent": "coherence", "raw_intents": ["meaning-changed", "coherence", "coherence"]}, {"type": "D", "before": "that has been", "after": null, "start_char_pos": 240, "end_char_pos": 253, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "D", "before": "The site is located in Nyanga District.", "after": null, "start_char_pos": 281, "end_char_pos": 320, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "engendered", "after": "uncovered", "start_char_pos": 609, "end_char_pos": 619, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 49, 280, 320, 431, 568], "domain": "wiki"}
{"doc_id": "743441", "revision_depth": 1, "before_revision": " A Bouguer reduction is called simple or incomplete if the terrain is approximated by an infinite flat plate called the Bouguer plate. A refined or complete Bouguer reduction removes the effects of terrain  precisely. The difference between the two , the differential gravitational effect of the unevenness of the terrain , is called the terrain effect. It is always negative.", "after_revision": "Reduction A Bouguer reduction is called simple (or incomplete) if the terrain is approximated by an infinite flat plate called the Bouguer plate. A refined (or complete) Bouguer reduction removes the effects of terrain more precisely. The difference between the two is called the terrain effect and is due to the differential gravitational effect of the unevenness of the terrain ; it is always negative.", "edit_actions": [{"type": "A", "before": null, "after": "Reduction", "start_char_pos": 0, "end_char_pos": 0, "major_intent": "meaning-changed", "raw_intents": ["style", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "or incomplete", "after": "(or incomplete)", "start_char_pos": 38, "end_char_pos": 51, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "or complete", "after": "(or complete)", "start_char_pos": 145, "end_char_pos": 156, "major_intent": "fluency", "raw_intents": ["fluency", "clarity", "fluency"]}, {"type": "A", "before": null, "after": "more", "start_char_pos": 206, "end_char_pos": 206, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": ", the", "after": "is called the terrain effect and is due to the", "start_char_pos": 249, "end_char_pos": 254, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": ", is called the terrain effect. It", "after": "; it", "start_char_pos": 322, "end_char_pos": 356, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}], "sents_char_pos": [0, 134, 217, 353], "domain": "wiki"}
{"doc_id": "7451157", "revision_depth": 1, "before_revision": "Penang's only synagogue, located on 28, Nagore Road, was opened in 1929 but closed down in 1976 as the community could no longer fulfill the minyan, a quorum of ten or more adult Jews assembled for purposes of fulfilling a public religious obligation. Today, approximately 100 Jews who are refugees from Russia are said to reside in Malaysia. The original Penang Jewish community ceased to exist with the death of Mordecai (Mordy) David Mordecai on 15 July 2011.News Straits Times (Northern Edition) article on 18 July 2011 entitled: \"Uncle Mordy laid to rest\" The rest of the Penang Jews have either embraced Christianity or else have emigrated to other countries, especially with the rise of anti-semitic sentiments and anti-Israel policies pursued by the Malaysian government since the 1970s.", "after_revision": "Penang's only synagogue, located on 28, Nagore Road, was opened in 1929 but closed down in 1976 as the community could no longer fulfill the minyan, a quorum of ten or more adult Jews assembled for purposes of fulfilling a public religious obligation. In 2008, it was reported that approximately 100 Jewish refugees from Russia were residing in Malaysia. The original Penang Jewish community ceased to exist with the death of Mordecai (Mordy) David Mordecai on 15 July 2011.News Straits Times (Northern Edition) article on 18 July 2011 entitled: \"Uncle Mordy laid to rest\" The rest of the Penang Jews have either embraced Christianity or else have emigrated to other countries, especially with the rise of anti-semitic sentiments and anti-Israel policies pursued by the Malaysian government since the 1970s.", "edit_actions": [{"type": "R", "before": "Today,", "after": "In 2008, it was reported that", "start_char_pos": 252, "end_char_pos": 258, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Jews who are", "after": "Jewish", "start_char_pos": 277, "end_char_pos": 289, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "are said to reside", "after": "were residing", "start_char_pos": 311, "end_char_pos": 329, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}], "sents_char_pos": [0, 251, 342, 462], "domain": "wiki"}
{"doc_id": "7466623", "revision_depth": 1, "before_revision": "Economic importance Economic extraction of SMS deposits is in the theoretical stage, the greatest complication being the extreme water depths at which these deposits are forming. However, apparent vast areas of the peripheral areas of these black smoker zones contain a sulfide ooze which could, theoretically, be vacuumed up off the seafloor. Nautilus Minerals Inc. (Nautilus) is engaged in commercially exploring the ocean floor for copper, gold, silver and zinc seafloor massive sulphide (SMS) deposits, and mineral extraction from an SMS system. Nautilus' Solwara 1 Project located at 1600 metres water depth in the Bismarck Sea, Papua New Guinea, will be the world's first deep-sea mining project, with first production  expected in 2017. ", "after_revision": "Economic importance Economic extraction of SMS deposits is in the theoretical stage, the greatest complication being the extreme water depths at which these deposits are forming. However, apparent vast areas of the peripheral areas of these black smoker zones contain a sulfide ooze which could, theoretically, be vacuumed up off the seafloor. Nautilus Minerals Inc. (Nautilus) was engaged in commercially exploring the ocean floor for copper, gold, silver and zinc seafloor massive sulphide (SMS) deposits, and mineral extraction from an SMS system. Nautilus' Solwara 1 Project located at 1,600 metres water depth in the Bismarck Sea, Papua New Guinea, was an attempt at the world's first deep-sea mining project, with first production originally expected in 2017. However, the company went bankrupt in 2019 after failing to secure funding for the project.", "edit_actions": [{"type": "R", "before": "is", "after": "was", "start_char_pos": 378, "end_char_pos": 380, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "fluency"]}, {"type": "R", "before": "1600", "after": "1,600", "start_char_pos": 589, "end_char_pos": 593, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "will be", "after": "was an attempt at", "start_char_pos": 652, "end_char_pos": 659, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "originally", "start_char_pos": 725, "end_char_pos": 725, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "However, the company went bankrupt in 2019 after failing to secure funding for the project.", "start_char_pos": 744, "end_char_pos": 744, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 178, 343, 549], "domain": "wiki"}
{"doc_id": "7478519", "revision_depth": 1, "before_revision": "Books of remembrance at a crematorium. A Book of remembrance is a book commemorating those who have died, usually listing their names in date or alphabetical order. They are often compiled to commemorate war dead and others who have died on military operations , but may also commemorate those buried in a cemetery .(2019, April 3). \"History of the Books of Remembrance\". Veterans Affairs Canada. Retrieved August 15, 2019.  Books of Remembrance are often held by municipal authorities or churches and pages may be turned daily in memory of those on each page. (2019, February 14). \"Today's Page in the Books of Remembrance\". Veterans Affairs Canada. Retrieved August 15, 2019.", "after_revision": "Books of remembrance at a crematorium. A book of remembrance is a book commemorating those who have died, usually listing their names in date or alphabetical order. They are often compiled to commemorate war dead and others who have died on military operations .(2019, April 3). \"History of the Books of Remembrance\". Veterans Affairs Canada. Retrieved August 15, 2019. Another use is to commemorate people who have been cremated or buried at a location, as an alternative to grave markers.A guide to memorial services - www.funeralguide.co.uk Books of remembrance are often held by municipal authorities or churches and pages may be turned daily in memory of those on each page. (2019, February 14). \"Today's Page in the Books of Remembrance\". Veterans Affairs Canada. Retrieved August 15, 2019.", "edit_actions": [{"type": "R", "before": "Book", "after": "book", "start_char_pos": 41, "end_char_pos": 45, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ", but may also commemorate those buried in a cemetery", "after": null, "start_char_pos": 261, "end_char_pos": 314, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "A", "before": null, "after": "Another use is to commemorate people who have been cremated or buried at a location, as an alternative to grave markers.A guide to memorial services - www.funeralguide.co.uk", "start_char_pos": 424, "end_char_pos": 424, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "Remembrance", "after": "remembrance", "start_char_pos": 434, "end_char_pos": 445, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 38, 164, 332, 371, 396, 423, 560, 581, 625, 650], "domain": "wiki"}
{"doc_id": "773158", "revision_depth": 1, "before_revision": "At the 1972 U.S. Olympic trials in Eugene, Oregon, Jenner was in fifth place in the men's decathlon,  behind Steve Gough and Andrew Pettes with one event remaining  . Needing to make up a 19-second gap on Gough in the 1500 metres, Jenner qualified for the Olympic team by finishing in firstplace 22 seconds ahead of the others. This prompted the Eugene Register-Guard to ask: \"Who's Jenner?\" Following the Olympic Trials, Jenner finished in tenth place in the decathlon at the 1972 Summer Olympics in Munich . By watching Soviet Mykola Avilov win the event, Jenner was inspired to start an intense training regimen. \"For the first time, I knew what I wanted out of life and that was it, and this guy has it. I literally started training that night at midnight, running through the streets of Munich, Germany, training for the Games. I trained that day on through the 1976 Games, 68 hours a day, every day, 365 days a year.\" After graduating from Graceland, Jenner married girlfriend Chrystie Crownover and moved to San Jose, California. Crownover provided most of the family income as a flight attendant for United Airlines. Jenner trained during the day and sold insurance at night, earning a year. In the era before professional American athletes were allowed to compete in Olympic sports, this kind of training was unheard of. On the other hand, Soviet athletes were state sponsored, which gave them an advantage over amateur American athletes. During this period, Jenner trained at the San Jose City College (SJCC) and San Jose State University (SJSU) tracks. San Jose athletics centered on SJCC coach Bert Bonanno; at that time, the city was a hotbed for training and was called the \"Track Capital of the World\". Many other aspiring Olympic athletes also trained at San Jose; the list included Millard Hampton, Andre Phillips, John Powell, Mac Wilkins, and Al Feuerbach. Jenner's best events were on day two of the decathlon: hurdles, discus, pole vault, javelin  and 1500 meters.", "after_revision": "At the 1972 U.S. Olympic Trials at Hayward Field in Eugene, Oregon, Jenner was eleventh after the first day in the men's decathlon, but had climbed to fifth behind Steve Gough and Andrew Pettes with one event remaining on the Fourth of July . Needing to make up a 19-second gap on Gough in the 1500 metres, Jenner qualified for the Olympic team by finishing first, 22 seconds ahead of the others. This prompted the Eugene Register-Guard to ask: \"Who's Jenner?\" Following the trials, Jenner was tenth in the decathlon at the 1972 Summer Olympics in Munich , West Germany. Bywatching Soviet Mykola Avilov win the event, Jenner was inspired to start an intense training regimen. \"For the first time, I knew what I wanted out of life and that was it, and this guy has it. I literally started training that night at midnight, running through the streets of Munich, Germany, training for the Games. I trained that day on through the 1976 Games, 68 hours a day, every day, 365 days a year.\" After graduating from Graceland, Jenner married girlfriend Chrystie Crownover and moved to San Jose, California. Crownover provided most of the family income as a flight attendant for United Airlines. Jenner trained during the day and sold insurance at night, earning a year. In the era before professional American athletes were allowed to compete in Olympic sports, this kind of training was unheard of. On the other hand, Soviet athletes were state sponsored, which gave them an advantage over amateur American athletes. During this period, Jenner trained at the San Jose City College (SJCC) and San Jose State University (SJSU) tracks. San Jose athletics centered on SJCC coach Bert Bonanno; at that time, the city was a hotbed for training and was called the \"Track Capital of the World\". Many other aspiring Olympic athletes also trained at San Jose; the list included Millard Hampton, Andre Phillips, John Powell, Mac Wilkins, and Al Feuerbach. Jenner's best events were on day two of the decathlon: hurdles, discus, pole vault, javelin , and 1500 meters.", "edit_actions": [{"type": "R", "before": "trials", "after": "Trials at Hayward Field", "start_char_pos": 25, "end_char_pos": 31, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in fifth place in", "after": "eleventh after the first day in", "start_char_pos": 62, "end_char_pos": 79, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "but had climbed to fifth", "start_char_pos": 101, "end_char_pos": 101, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "on the Fourth of July", "start_char_pos": 164, "end_char_pos": 164, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "in firstplace", "after": "first,", "start_char_pos": 282, "end_char_pos": 295, "major_intent": "clarity", "raw_intents": ["fluency", "clarity", "clarity"]}, {"type": "R", "before": "Olympic Trials, Jenner finished in tenth place", "after": "trials, Jenner was tenth", "start_char_pos": 406, "end_char_pos": 452, "major_intent": "clarity", "raw_intents": ["clarity", "style", "clarity"]}, {"type": "R", "before": ". By watching", "after": ", West Germany. Bywatching", "start_char_pos": 508, "end_char_pos": 521, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 1968, "end_char_pos": 1968, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}], "sents_char_pos": [0, 166, 327, 391, 509, 615, 707, 832, 923, 1036, 1124, 1199, 1329, 1447, 1563, 1619, 1717, 1780, 1875], "domain": "wiki"}
{"doc_id": "7862242", "revision_depth": 2, "before_revision": "Black people have been 28\\% of those killed by police since 2013 despite being only 13\\% of the population. This number however does not include that black people are over represented in crimes , which means they have a higher chance of running into police, see Black vs white homicides", "after_revision": "Black people have been 28\\% of those killed by police since 2013 despite being only 13\\% of the population. This number however does not factor that black people are over represented in violent crime , which means they have a higher chance of negative encounters with the police, see Black vs white homicides", "edit_actions": [{"type": "R", "before": "include", "after": "factor", "start_char_pos": 137, "end_char_pos": 144, "major_intent": "clarity", "raw_intents": ["others", "clarity", "clarity"]}, {"type": "R", "before": "crimes", "after": "violent crime", "start_char_pos": 187, "end_char_pos": 193, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "running into", "after": "negative encounters with the", "start_char_pos": 237, "end_char_pos": 249, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}], "sents_char_pos": [0, 107], "domain": "wiki"}
{"doc_id": "804964", "revision_depth": 1, "before_revision": "Language Dehumanization and dehumanized perception can occur as a result of the language used to describe groups of people. Words such as migrant, immigrant, and expatriate are assigned to foreigners based on their social status and wealth, rather than ability, achievements, or political alignment. Expatriate is a word to describe the privileged, often light-skinned people newly residing in an area and has connotations that suggest ability, wealth, and trust. Meanwhile, the word immigrant is used to describe people coming to a new location to reside and infers a much less-desirable meaning. Further, \"immigrant\" is a word that can be paired with \"illegal\", which harbors a profoundly negative connotationto those projecting social cognition towards the other. The perpetual misuse of these words used to describe the other in the English language can alter the perception of a group in a derogatory way. \"Most of the time when we hear illegal immigrant%DIFDELCMD < ] %%% used, most of the time, the shorter version 'illegals' is being used as a noun, which implies that a human being is perpetually illegal. There is no other classification that I'm aware of where the individual is being rendered as unlawful as opposed to those individuals' actions.\" A series of language examinations found a direct relation between homophobic epithets and social cognitive distancing towards a group of homosexuals, a form of dehumanization. These epithets (e.g., faggot) were thought to function as dehumanizing labels because they tended to act as markers of deviance. One pair of studies found that subjects were more likely to associate malignant language with homosexuals, and that such language associations increased the physical distancing between the subject and the homosexual. This indicated that the malignant language could encourage dehumanization, cognitive and physical distancing in ways that other forms of malignant language do not.", "after_revision": "Language Dehumanization and dehumanized perception can occur as a result of the language used to describe groups of people. Words such as migrant, immigrant, and expatriate are assigned to foreigners based on their social status and wealth, rather than ability, achievements, or political alignment. Expatriate is a word to describe the privileged, often light-skinned people newly residing in an area and has connotations that suggest ability, wealth, and trust. Meanwhile, the word immigrant is used to describe people coming to a new location to reside and infers a much less-desirable meaning. The word \"immigrant\" is sometimes paired with \"illegal\", which harbors a profoundly derogatory connotation. Misuse of these terms - they are often used inaccurately - to describe the other , can alter the perception of a group %DIFDELCMD < ] %%% as a whole in a negative way. Ryan Eller, the executive director of the immigrant advocacy group, Define American, expressed the problem this way: A series of language examinations found a direct relation between homophobic epithets and social cognitive distancing towards a group of homosexuals, a form of dehumanization. These epithets (e.g., faggot) were thought to function as dehumanizing labels because they tended to act as markers of deviance. One pair of studies found that subjects were more likely to associate malignant language with homosexuals, and that such language associations increased the physical distancing between the subject and the homosexual. This indicated that the malignant language could encourage dehumanization, cognitive and physical distancing in ways that other forms of malignant language do not.", "edit_actions": [{"type": "R", "before": "Further,", "after": "The word", "start_char_pos": 598, "end_char_pos": 606, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "style"]}, {"type": "R", "before": "a word that can be", "after": "sometimes", "start_char_pos": 622, "end_char_pos": 640, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "meaning-changed"]}, {"type": "R", "before": "negative connotationto those projecting social cognition towards the other. The perpetual misuse of these words used", "after": "derogatory connotation. Misuse of these terms - they are often used inaccurately -", "start_char_pos": 691, "end_char_pos": 807, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "R", "before": "in the English language", "after": ",", "start_char_pos": 830, "end_char_pos": 853, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "D", "before": "in a derogatory way. \"Most of the time when we hear", "after": null, "start_char_pos": 890, "end_char_pos": 941, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "D", "before": "illegal immigrant", "after": null, "start_char_pos": 942, "end_char_pos": 959, "major_intent": "style", "raw_intents": ["coherence", "style", "style"]}, {"type": "R", "before": "used, most of the time, the shorter version 'illegals' is being used as a noun, which implies that a human being is perpetually illegal. There is no other classification that I'm aware of where the individual is being rendered as unlawful as opposed to those individuals' actions.\"", "after": "as a whole in a negative way. Ryan Eller, the executive director of the immigrant advocacy group, Define American, expressed the problem this way:", "start_char_pos": 978, "end_char_pos": 1259, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 123, 299, 463, 597, 766, 910, 1114, 1259, 1435, 1564, 1781]}
{"doc_id": "8150951", "revision_depth": 1, "before_revision": "Plant perception or biocommunication is the paranormal idea that plants are sentient, that they respond to humans in a manner that amounts to ESP, and that they experience pain and fear. The idea is not accepted by the scientific community, as plants lack nervous systems . Galston, Arthur W; Slayman, Clifford L. Plant Sensitivity and Sensation. In URLe Ogden Abell, Barry Singer. (1981). Science and the Paranormal: Probing the Existence of the Supernatural. Junction Books. pp. 40-55. Tittle, Peg. (2011). Critical Thinking: An Appeal to Reason. Routledge. p. 317. Jong, Tom de; Klinkhamer, Peter. (2005). Evolutionary Ecology of Plant Reproductive Strategies. Cambridge University Press. p. 2. \"Plants do not have a nervous system and certainly do not make conscious decisions about what to do next.\" Paranormal claims in regard to plant perception are considered to be pseudoscience by the scientific community.Audus, Leslie. (1974). Roots of Absurdity. New Scientist. 17 October. p. 207Galston, Arthur W; Slayman, Clifford L. (1979). The Not-So-Secret Life of Plants: In Which the Historical and Experimental Myths About Emotional Communication Between Animal and Vegetable Are Put to Rest. American Scientist 67 (3): 337-344. The idea is distinct from measured plant perception and chemical communication .", "after_revision": "Plant perception or biocommunication is the paranormal idea that plants are sentient, that they respond to humans in a manner that amounts to ESP, and that they experience a range of emotions or parapsychological states. Since plants lack nervous systems , Galston, Arthur W; Slayman, Clifford L. Plant Sensitivity and Sensation. In URLe Ogden Abell, Barry Singer. (1981). Science and the Paranormal: Probing the Existence of the Supernatural. Junction Books. pp. 40-55. Tittle, Peg. (2011). Critical Thinking: An Appeal to Reason. Routledge. p. 317. Jong, Tom de; Klinkhamer, Peter. (2005). Evolutionary Ecology of Plant Reproductive Strategies. Cambridge University Press. p. 2. \"Plants do not have a nervous system and certainly do not make conscious decisions about what to do next.\" paranormal claims regarding plant perception are considered pseudoscience by the scientific community.Audus, Leslie. (1974). Roots of Absurdity. New Scientist. 17 October. p. 207Galston, Arthur W; Slayman, Clifford L. (1979). The Not-So-Secret Life of Plants: In Which the Historical and Experimental Myths About Emotional Communication Between Animal and Vegetable Are Put to Rest. American Scientist 67 (3): 337-344. Such paranormal claims are distinct from the ability of plants to sense and respond to the environment via chemical and related stimuli .", "edit_actions": [{"type": "R", "before": "pain and fear. The idea is not accepted by the scientific community, as", "after": "a range of emotions or parapsychological states. Since", "start_char_pos": 172, "end_char_pos": 243, "major_intent": "style", "raw_intents": ["style", "others", "clarity"]}, {"type": "R", "before": ".", "after": ",", "start_char_pos": 272, "end_char_pos": 273, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "style"]}, {"type": "R", "before": "Paranormal claims in regard to", "after": "paranormal claims regarding", "start_char_pos": 805, "end_char_pos": 835, "major_intent": "fluency", "raw_intents": ["clarity", "fluency", "fluency"]}, {"type": "D", "before": "to be", "after": null, "start_char_pos": 868, "end_char_pos": 873, "major_intent": "fluency", "raw_intents": ["fluency", "others", "clarity"]}, {"type": "R", "before": "The idea is distinct from measured plant perception and chemical communication", "after": "Such paranormal claims are distinct from the ability of plants to sense and respond to the environment via chemical and related stimuli", "start_char_pos": 1233, "end_char_pos": 1311, "major_intent": "clarity", "raw_intents": ["clarity", "others", "clarity"]}], "sents_char_pos": [0, 186, 292, 346, 381, 460, 487, 508, 548, 581, 600, 663, 804, 916, 930, 958, 973, 1010, 1039, 1196, 1232], "domain": "wiki"}
{"doc_id": "8489599", "revision_depth": 1, "before_revision": "Naseer Ahmad Malhi () (15 August 1911 12 July 1991) was a Pakistani politician, known for playing a pivotal role in the formation of the Islamic Republic of Pakistan. Malhi is recognised as one of the nation's founding fathers . Birth and family Naseer Ahmad Malhi was born as the second son of Chaudhry Ghulam Haider Malhi, in 1913 , in the town of Baddomalhi, in Sialkot district. Malhi's father was the leading land-owning farmer of the district, one of the elite of Punjab, was noted for his philanthropy, and was decorated by the British Governor for his services to the community. Malhi's great-grandfather, Chaudhry Ali Gohar Malhi, served as Governor of Punjab during the reign of Maharaja Ranjit Singh who ruled Punjab from 1801 to 1839. Malhi had two older sisters, two younger sisters, and an older stepbrother. He fathered three sons and two daughters, of whom only two sons and one daughter survived adolescence. Education Malhi received his primary and secondary education at G.H. (Ghulam Haider) Muslim High School, named after his father, who was encouraged by Maulana Muhammad Ali to found this school in 1918, to educate aspiring students. The school was noted for offering free tuition and books to economically-disadvantaged pupils. The educational expenses of these students were personally met by Malhi's father. Post-independence After the independence of Pakistan, Malhi remained a prominent legislator. He was appointed as the nation's first education minister. It was Jinnah's decision to appoint Malhi as Pakistan's second prime minister, after Liaquat Ali Khan. Jinnah personally notified Malhi of this decision; however, due to Jinnah's sudden death, the writ was never executed. Malhi led Pakistan's delegation at the Geneva Conference in 1955. At this conference he delivered a speech defending the rights of Afro-Asian countries and was awarded a gold medal. Following this success, Malhi led Pakistans delegation to the United Nations, where he addressed the United Nations General Assembly on the political challenges facing South Asia. His speech won acclaim for its political insight, and he was well received by the Assembly. Legacy Malhi died on 12 July 1991.", "after_revision": "Naseer Ahmad Malhi () (15 August 1911 12 July 1991) was a Pakistani politician, known for playing a pivotal role in the formation of the Islamic Republic of Pakistan. Malhi is recognised as one of the nation's activists of the Pakistan Movement . Birth and family Naseer Ahmad Malhi was born as the second son of Chaudhry Ghulam Haider Malhi, in 1911 , in the town of Baddomalhi, in Sialkot district. Malhi's father was the leading land-owning farmer of the district, one of the elite of Punjab, was noted for his philanthropy, and was decorated by the British Governor for his services to the community. Malhi's great-grandfather, Chaudhry Ali Gohar Malhi, served as Governor of Punjab during the reign of Maharaja Ranjit Singh who ruled Punjab from 1801 to 1839. Education Naseer Ahmad Malhi received his primary and secondary education at G.H. (Ghulam Haider) Muslim High School, named after his father, who was encouraged by Sir Syed Ahmad Khan to found this school in 1918, to educate aspiring students. The school was noted for offering free tuition and books to economically-disadvantaged pupils. The educational expenses of these students were personally met by Malhi's father. Post-independence After the independence of Pakistan, Malhi remained a prominent legislator. He was appointed as the nation's first education minister. Malhi led Pakistan's delegation at the Geneva Conference in 1955. At this conference he delivered a speech defending the rights of Afro-Asian countries and was awarded a gold medal. Following this success, Malhi led Pakistans delegation to the United Nations, where he addressed the United Nations General Assembly on the political challenges facing South Asia. Death and legacy Naseer Ahmad Malhi died on 12 July 1991.", "edit_actions": [{"type": "R", "before": "founding fathers", "after": "activists of the Pakistan Movement", "start_char_pos": 210, "end_char_pos": 226, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "1913", "after": "1911", "start_char_pos": 328, "end_char_pos": 332, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "Malhi had two older sisters, two younger sisters, and an older stepbrother. He fathered three sons and two daughters, of whom only two sons and one daughter survived adolescence.", "after": null, "start_char_pos": 747, "end_char_pos": 925, "major_intent": "coherence", "raw_intents": ["coherence", "clarity", "coherence"]}, {"type": "R", "before": "Education", "after": "Education Naseer Ahmad", "start_char_pos": 926, "end_char_pos": 935, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}, {"type": "R", "before": "Maulana Muhammad Ali", "after": "Sir Syed Ahmad Khan", "start_char_pos": 1077, "end_char_pos": 1097, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "D", "before": "It was Jinnah's decision to appoint Malhi as Pakistan's second prime minister, after Liaquat Ali Khan. Jinnah personally notified Malhi of this decision; however, due to Jinnah's sudden death, the writ was never executed.", "after": null, "start_char_pos": 1487, "end_char_pos": 1708, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "D", "before": "His speech won acclaim for its political insight, and he was well received by the Assembly.", "after": null, "start_char_pos": 2071, "end_char_pos": 2162, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}, {"type": "R", "before": "Legacy", "after": "Death and legacy Naseer Ahmad", "start_char_pos": 2163, "end_char_pos": 2169, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 166, 382, 586, 746, 822, 925, 1157, 1252, 1334, 1427, 1486, 1589, 1640, 1708, 1774, 1890, 2070, 2162]}
{"doc_id": "8489599", "revision_depth": 2, "before_revision": "The Sialkot Convention  was a milestone in MuslimHindu politics. It is widely regarded by Pakistani historians as the landmark event which brought the Muslim League into prominence in undivided Punjab. It broke the Unionist Party's hold over the Muslims of Punjab and swayed their sentiments towards the Muslim League and the Pakistan movement, paving the way for the eventual formation of Pakistan.  In his efforts to elevate Pakistan's fledgling profile , Malhi hosted an historic dinner for Eleanor Roosevelt, President of the United Nations General Assembly and former First Lady of the United States. With the advent of martial law in the early sixties, Malhi became a vociferous opponent of the government of General Ayub Khan. In 1965  he allied himself with and supported Fatima Jinnah in her bid to democratically dislodge Ayub Khan from his assumed office. Despite his withdrawal from active political life, Malhi remained an established member of Pakistans elite intellectual and political circles. Death and legacy Naseer Ahmad Malhi died on 12 July 1991. Muhammad Ali Jinnah considered him among his most able stalwarts. Pakistan recognises him as one of its most dedicated founding fathers.", "after_revision": "The Sialkot Convention in May 1944 was a milestone in MuslimHindu politics. It is widely regarded by Pakistani historians as the landmark event which brought the Muslim League into prominence in undivided Punjab. It broke the Unionist Party's hold over the Muslims of Punjab and swayed their sentiments towards the Muslim League and the Pakistan movement, paving the way for the eventual formation of Pakistan. Brief History of Sialkot (scroll down to Colonial Sialkot to read about Sialkot Convention (1944) arranged by Naseer Ahmad Malhi) District Courts Sialkot website, Retrieved 21 January 2020 In his efforts to elevate Pakistan's image , Malhi hosted an historic dinner for Eleanor Roosevelt, President of the United Nations General Assembly and former First Lady of the United States. With the advent of martial law in the early sixties, Malhi became a vociferous opponent of the government of General Ayub Khan. In 1965 , he allied himself with and supported Fatima Jinnah in her bid to democratically dislodge Ayub Khan from his assumed office. Death Naseer Ahmad Malhi died on 12 July 1991. ", "edit_actions": [{"type": "A", "before": null, "after": "in May 1944", "start_char_pos": 23, "end_char_pos": 23, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "A", "before": null, "after": "Brief History of Sialkot (scroll down to Colonial Sialkot to read about Sialkot Convention (1944) arranged by Naseer Ahmad Malhi) District Courts Sialkot website, Retrieved 21 January 2020", "start_char_pos": 400, "end_char_pos": 400, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "fledgling profile", "after": "image", "start_char_pos": 438, "end_char_pos": 455, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "A", "before": null, "after": ",", "start_char_pos": 742, "end_char_pos": 742, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": "Despite his withdrawal from active political life, Malhi remained an established member of Pakistans elite intellectual and political circles.", "after": null, "start_char_pos": 867, "end_char_pos": 1009, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "Death and legacy", "after": "Death", "start_char_pos": 1010, "end_char_pos": 1026, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "Muhammad Ali Jinnah considered him among his most able stalwarts. Pakistan recognises him as one of its most dedicated founding fathers.", "after": null, "start_char_pos": 1068, "end_char_pos": 1204, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "clarity"]}], "sents_char_pos": [0, 64, 201, 399, 605, 733, 866, 1009, 1067, 1133], "domain": "wiki"}
{"doc_id": "8556500", "revision_depth": 1, "before_revision": "Practitioners of oil pulling claim it is capable of improving oral and whole-body health, including a benefit in conditions such as headaches, migraines, diabetes mellitus, asthma, and acne, as well as whitening teeth. Its promoters claim it works by pulling out toxins, which are known as ama in Ayurveda, but there is no credible evidence to support some of these claims and only limited and inconclusive evidence to support other some of these claims, primarily having to do with oral health.  A review of the evidence published in the scientific journal Heliyon (august 7, 2020) URL found \"The limited evidence suggests that oil pulling with coconut oil may have a beneficial effect on improving oral health and dental hygiene. Future clinical trials are of merit considering the universal availability of the intervention. Prospective research should have a robust design with rigorous execution to provide a higher quality of evidence... Oil pulling with coconut oil could be used as a adjunct to normal preventative regimes to improve oral health and dental hygiene although further studies are needed to determine the level of effectiveness.\" This article was deemed significant enought for the toothpaste maker Colgate to cite it on its website URL", "after_revision": "Some practitioners of oil pulling claim it is capable of improving oral and whole-body health, including a benefit in conditions such as headaches, migraines, diabetes mellitus, asthma, and acne, as well as whitening teeth. Some of its promoters claim it works by pulling out toxins, which are known as ama in Ayurveda, . There is no credible evidence to support the claims about toxins. There is only limited and inconclusive evidence to support others of these claims, primarily having to do with oral health. The scientific case for oil pulling for oral health rests on multiple studies that have found that oil pulling, particularly with coconut oil, reduces bacteria and inflammation in the mouth. A review of the evidence published in the scientific journal Heliyon (august 7, 2020) URL found \"The limited evidence suggests that oil pulling with coconut oil may have a beneficial effect on improving oral health and dental hygiene. Future clinical trials are of merit considering the universal availability of the intervention. Prospective research should have a robust design with rigorous execution to provide a higher quality of evidence... Oil pulling with coconut oil could be used as a adjunct to normal preventative regimes to improve oral health and dental hygiene although further studies are needed to determine the level of effectiveness.\" This article was deemed significant enought for the toothpaste maker Colgate to cite it on its website URL", "edit_actions": [{"type": "R", "before": "Practitioners", "after": "Some practitioners", "start_char_pos": 0, "end_char_pos": 13, "major_intent": "clarity", "raw_intents": ["clarity", "style", "fluency"]}, {"type": "R", "before": "Its", "after": "Some of its", "start_char_pos": 219, "end_char_pos": 222, "major_intent": "style", "raw_intents": ["style", "coherence", "style"]}, {"type": "R", "before": "but there", "after": ". There", "start_char_pos": 307, "end_char_pos": 316, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "fluency"]}, {"type": "R", "before": "some of these claims and", "after": "the claims about toxins. There is", "start_char_pos": 352, "end_char_pos": 376, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": "other some", "after": "others", "start_char_pos": 427, "end_char_pos": 437, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "The scientific case for oil pulling for oral health rests on multiple studies that have found that oil pulling, particularly with coconut oil, reduces bacteria and inflammation in the mouth.", "start_char_pos": 496, "end_char_pos": 496, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 218, 495, 731, 827, 943, 1150], "domain": "wiki"}
{"doc_id": "8556500", "revision_depth": 2, "before_revision": "Oil pulling is an alternative medical practice in which an edible oil is swished around the mouth for a period of time and then spat out, similar to mouthwash. There is only limited and inconclusive evidence to support the claims made for the benefits of oil pulling. Some practitioners of oil pulling claim it is capable of improving oral and whole-body health, including a benefit in conditions such as headaches, migraines, diabetes mellitus, asthma, and acne, as well as whitening teeth. Some of its promoters claim it works by pulling out toxins, which are known as ama in Ayurveda, . There is no credible evidence to support the claims about toxins. There is only limited and inconclusive evidence to support others of these claims, primarily having to do with oral health. The scientific case for oil pulling for oral health rests on multiple studies that have found that oil pulling, particularly with coconut oil, reduces bacteria and inflammation in the mouth. A review of the evidence published in the scientific journal Heliyon (august 7, 2020) URL found \"The limited evidence suggests that oil pulling with coconut oil may have a beneficial effect on improving oral health and dental hygiene. Future clinical trials are of merit considering the universal availability of the intervention. Prospective research should have a robust design with rigorous execution to provide a higher quality of evidence... Oil pulling with coconut oil could be used as a adjunct to normal preventative regimes to improve oral health and dental hygiene although further studies are needed to determine the level of effectiveness.\" This article was deemed significant enought for the toothpaste maker Colgate to cite it on its website URL", "after_revision": "Oil pulling is an alternative medical practice in which an edible oil is swished around the mouth for a period of time and then spat out, similar to mouthwash. There is no evidence to support the claims made for the benefits of oil pulling. Practitioners of oil pulling claim it is capable of improving oral and whole-body health, including a benefit in conditions such as headaches, migraines, diabetes mellitus, asthma, and acne, as well as whitening teeth. Its promoters claim it works by pulling out toxins, which are known as ama in Ayurveda, but there is no credible evidence to support this.", "edit_actions": [{"type": "R", "before": "only limited and inconclusive", "after": "no", "start_char_pos": 169, "end_char_pos": 198, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "Some practitioners", "after": "Practitioners", "start_char_pos": 268, "end_char_pos": 286, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "coherence"]}, {"type": "R", "before": "Some of its", "after": "Its", "start_char_pos": 492, "end_char_pos": 503, "major_intent": "style", "raw_intents": ["style", "clarity", "style"]}, {"type": "R", "before": ". There", "after": "but there", "start_char_pos": 588, "end_char_pos": 595, "major_intent": "coherence", "raw_intents": ["coherence", "style", "coherence"]}, {"type": "D", "before": "the claims about toxins. There is only limited and inconclusive evidence to support others of these claims, primarily having to do with oral health.", "after": null, "start_char_pos": 631, "end_char_pos": 779, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "The scientific case for oil pulling for oral health rests on multiple studies that have found that oil pulling, particularly with coconut oil, reduces bacteria and inflammation in the mouth. A review of the evidence published in the scientific journal Heliyon (august 7, 2020) URL found \"The limited evidence suggests that oil pulling with coconut oil may have a beneficial effect on improving oral health and dental hygiene. Future clinical trials are of merit considering the universal availability of the intervention. Prospective research should have a robust design with rigorous execution to provide a higher quality of evidence... Oil pulling with coconut oil could be used as a adjunct to normal preventative regimes to improve oral health and dental hygiene although further studies are needed to determine the level of effectiveness.\" This article was deemed significant enought for the toothpaste maker Colgate to cite it on its website URL", "after": "this.", "start_char_pos": 780, "end_char_pos": 1731, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 159, 267, 491, 589, 655, 779, 970, 1205, 1301, 1417, 1624], "domain": "wiki"}
{"doc_id": "938546", "revision_depth": 1, "before_revision": "Habsburg Spain was the Spain of the 16th and 17th centuries (15161700) when it was ruled by kings from the House of Habsburg (also associated with its role in the history of Central and Eastern Europe). The Habsburg Hispanic Monarchs (chiefly Charles I and Philip II) reached the zenith of their influence and power ruling the Spanish Empire. They controlled territories over the five continents including the Americas, the East Indies, the Low Countries, Belgium, Luxembourg and territories now in Italy, France and Germany in Europe, the Portuguese Empire from 1580 to 1640, and various other territories such as small enclaves like Ceuta and Oran in North Africa. This period of Spanish history has also been referred to as the \"Age of Expansion\". With the Habsburgs, Spain was one of the greatest political and military powers in Europe and the world for much of the sixteenth and seventeenth centuries. The Habsburgyears also ushered in the Spanish Golden Age of arts and literature . Some of the world's most outstanding writers and painters lived in this period , including Teresa of vila, Pedro Caldern de la Barca, Miguel de Cervantes, Francisco de Quevedo, Diego Velzquez, El Greco, Domingo de Soto, Francisco Surez and Francisco de Vitoria. 150px|Arms of Charles I, representing his territories in Spain (top) and his other European possessions (bottom) Spain or \"the Spains\", referring to Spanish territories across different continents in this period, initially covered the entire Iberian peninsula, including the Kingdoms of Aragon, Castile, Len, Navarre and, from 1580, Portugal.", "after_revision": "Habsburg Spain is a contemporary historiographical term referred to the Spain of the 16th and 17th centuries (15161700) when it was ruled by kings from the House of Habsburg (also associated with its role in the history of Central and Eastern Europe). The Habsburg Hispanic Monarchs (chiefly Charles I and Philip II) reached the zenith of their influence and power ruling the Spanish Empire. They controlled territories over the five continents including the Americas, the East Indies, the Low Countries, Belgium, Luxembourg and territories now in Italy, France and Germany in Europe, the Portuguese Empire from 1580 to 1640, and various other territories such as small enclaves like Ceuta and Oran in North Africa. This period of Spanish history has also been referred to as the \"Age of Expansion\". With the Habsburgs, Spain was the foremost political and military power in Europe and the world for much of the sixteenth and seventeenth centuries. During the Habsburg's period, Spain ushered in the Spanish Golden Age of arts and literature producing some of the world's most outstanding writers and painters and influential intellectuals , including Teresa of vila, Pedro Caldern de la Barca, Miguel de Cervantes, Francisco de Quevedo, Diego Velzquez, El Greco, Domingo de Soto, Francisco Surez and Francisco de Vitoria. 150px|Arms of Charles I, representing his territories in Spain (top) and his other European possessions (bottom) Spain or \"the Spains\", referring to Spanish territories across different continents in this period, initially covered the entire Iberian peninsula, including the Kingdoms of Aragon, Castile, Len, Navarre and, from 1580, Portugal.", "edit_actions": [{"type": "R", "before": "was", "after": "is a contemporary historiographical term referred to", "start_char_pos": 15, "end_char_pos": 18, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "clarity", "meaning-changed"]}, {"type": "R", "before": "one of the greatest", "after": "the foremost", "start_char_pos": 781, "end_char_pos": 800, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "powers", "after": "power", "start_char_pos": 824, "end_char_pos": 830, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "R", "before": "The Habsburgyears also", "after": "During the Habsburg's period, Spain", "start_char_pos": 908, "end_char_pos": 930, "major_intent": "clarity", "raw_intents": ["coherence", "clarity", "clarity"]}, {"type": "R", "before": ". Some", "after": "producing some", "start_char_pos": 988, "end_char_pos": 994, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}, {"type": "R", "before": "lived in this period", "after": "and influential intellectuals", "start_char_pos": 1048, "end_char_pos": 1068, "major_intent": "clarity", "raw_intents": ["meaning-changed", "clarity", "clarity"]}], "sents_char_pos": [0, 202, 342, 666, 750, 907, 1251], "domain": "wiki"}
{"doc_id": "965390", "revision_depth": 1, "before_revision": "The matching hypothesis (also known as the matching phenomenon) is derived from the discipline of social psychology and was first proposed by  Elaine Hatfield and her colleagues in 1966, Walster, E., Aronson, V., Abrahams, D., & Rottman, L. (1966). Importance of physical attractiveness in dating behavior. Journal of Personality and Social Psychology, 4(5), 508-516 . which suggests why people become attracted to their partner. It claims that people are more likely to form and succeed in a committed relationship with someone who is equally socially desirable. This is often researched in the form of physical attraction .", "after_revision": "The matching hypothesis (also known as the matching phenomenon) argues that people are more likely to form and succeed in a committed relationship with someone who is equally socially desirable, typically in the form of physical attraction. The hypothesis is derived from the discipline of social psychology and was first proposed by American social psychologist Elaine Hatfield and her colleagues in 1966. Walster, E., Aronson, V., Abrahams, D., & Rottman, L. (1966). Importance of physical attractiveness in dating behavior. Journal of Personality and Social Psychology, 4(5), 508-516 .", "edit_actions": [{"type": "R", "before": "is", "after": "argues that people are more likely to form and succeed in a committed relationship with someone who is equally socially desirable, typically in the form of physical attraction. The hypothesis is", "start_char_pos": 64, "end_char_pos": 66, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "coherence", "meaning-changed"]}, {"type": "A", "before": null, "after": "American social psychologist", "start_char_pos": 142, "end_char_pos": 142, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "1966,", "after": "1966.", "start_char_pos": 181, "end_char_pos": 186, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "fluency"]}, {"type": "D", "before": ". which suggests why people become attracted to their partner. It claims that people are more likely to form and succeed in a committed relationship with someone who is equally socially desirable. This is often researched in the form of physical attraction", "after": null, "start_char_pos": 367, "end_char_pos": 623, "major_intent": "coherence", "raw_intents": ["coherence", "coherence", "coherence"]}], "sents_char_pos": [0, 248, 306, 429, 563], "domain": "wiki"}
{"doc_id": "990732", "revision_depth": 1, "before_revision": "The Royals selected Greinke in the first round of the 2002 MLB draft, after he won the Gatorade National Player of the Year Award as a high school senior. After playing in the minor leagues, he made his MLB debut in 2004. His career was nearly derailed by his battles with depression and anxiety in 2005 and 2006, and he missed most of the 2006 season. He returned in 2007 as a relief pitcher, before rejoining the starting rotation in 2008 and developing into one of the top pitchers in the game. In 2009, he appeared in the MLB All-Star Game, led the major leagues in earned run average, and won the American League Cy Young Award. In 2019, he was traded to the Astros in a blockbuster deal,  becoming a notable third piece in their historic starting rotation on the way to their second World Series berth in three years. Owing to his unusual pitching mechanics, eephus style, and relatively high batting average, Greinke is often considered one of the most unique starting pitchers in baseball. He is a six-time All-Star, and six-time Gold Glove  Award winner.", "after_revision": "The Royals selected Greinke in the first round of the 2002 MLB draft, after he won the Gatorade National Player of the Year Award as a high school senior. After playing in the minor leagues, he made his MLB debut in 2004. His career was nearly derailed by his battles with depression and anxiety in 2005 and 2006, and he missed most of the 2006 season. He returned in 2007 as a relief pitcher, before rejoining the starting rotation in 2008 and developing into one of the top pitchers in the game. In 2009, he appeared in the MLB All-Star Game, led the major leagues in earned run average, and won the American League Cy Young Award. In 2019, he was traded to the Astros in a blockbuster deal, soon thereafter reaching his 200th career win and becoming a notable third piece in the team's historic starting rotation during their quest for a second World Series berth in three years. Owing to his unusual pitching mechanics, eephus style, and relatively high batting average, Greinke is often considered one of the most unique starting pitchers in baseball. He is a six-time All-Star, six-time Gold Glove Award winner, and two-time Silver Slugger Award winner.", "edit_actions": [{"type": "A", "before": null, "after": "soon thereafter reaching his 200th career win and", "start_char_pos": 694, "end_char_pos": 694, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "their", "after": "the team's", "start_char_pos": 729, "end_char_pos": 734, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "clarity"]}, {"type": "R", "before": "on the way to their", "after": "during their quest for a", "start_char_pos": 762, "end_char_pos": 781, "major_intent": "clarity", "raw_intents": ["clarity", "clarity", "style"]}, {"type": "D", "before": "and", "after": null, "start_char_pos": 1025, "end_char_pos": 1028, "major_intent": "clarity", "raw_intents": ["clarity", "coherence", "fluency"]}, {"type": "A", "before": null, "after": "Award winner, and two-time Silver Slugger", "start_char_pos": 1049, "end_char_pos": 1049, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 154, 221, 352, 497, 633, 823, 997], "domain": "wiki"}
{"doc_id": "990732", "revision_depth": 2, "before_revision": "In 2009, he appeared in the MLB All-Star Game, led the major leagues in earned run average, and won the American League Cy Young Award. In 2019 , he was traded to the Astros in a blockbuster deal , soon thereafterreaching his 200th career win and becoming a notable third piece  in the team's historic starting rotation during their quest for a second World Series berth in three years.", "after_revision": "In 2009, Greinke appeared in the MLB All-Star Game, led the major leagues in earned run average, and won the American League Cy Young Award. After nearly a decade of hopping between franchises , he was sent to the Astros in a blockbuster deal at the 2019 trade deadline. Soon thereafter, Greinke reached his 200th career win and became a notable third piece behind Justin Verlander and Gerrit Cole in the team's historic starting rotation during their quest for a second World Series berth in three years.", "edit_actions": [{"type": "R", "before": "he", "after": "Greinke", "start_char_pos": 9, "end_char_pos": 11, "major_intent": "meaning-changed", "raw_intents": ["clarity", "meaning-changed", "meaning-changed"]}, {"type": "R", "before": "In 2019", "after": "After nearly a decade of hopping between franchises", "start_char_pos": 136, "end_char_pos": 143, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "coherence"]}, {"type": "R", "before": "traded", "after": "sent", "start_char_pos": 153, "end_char_pos": 159, "major_intent": "style", "raw_intents": ["clarity", "style", "style"]}, {"type": "R", "before": ", soon thereafterreaching", "after": "at the 2019 trade deadline. Soon thereafter, Greinke reached", "start_char_pos": 196, "end_char_pos": 221, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "clarity"]}, {"type": "R", "before": "becoming", "after": "became", "start_char_pos": 247, "end_char_pos": 255, "major_intent": "fluency", "raw_intents": ["fluency", "fluency", "clarity"]}, {"type": "A", "before": null, "after": "behind Justin Verlander and Gerrit Cole", "start_char_pos": 278, "end_char_pos": 278, "major_intent": "meaning-changed", "raw_intents": ["meaning-changed", "meaning-changed", "meaning-changed"]}], "sents_char_pos": [0, 135], "domain": "wiki"}