Datasets:

feradauto

/

NLP4SGPapers

like 2

babych-etal-2007-translating

https://aclanthology.org/2007.mtsummit-papers.5.pdf

Translating from under-resourced languages: comparing direct transfer against pivot translation

In this paper we compare two methods for translating into English from languages for which few MT resources have been developed (e.g. Ukrainian). The first method involves direct transfer using an MT system that is available for this language pair. The second method involves translation via a cognate language, which has more translation resources and one or more advanced translation systems (e.g. Russian for Slavonic languages). The comparison shows that it is possible to achieve better translation quality via the pivot language, leveraging on advanced dictionaries and grammars available for it and on lexical and syntactic similarities between the source and pivot languages. The results suggest that MT development efforts can be efficiently reused for families of closely related languages, and investing in MT for closely related languages can be more productive than developing systems from scratch for new translation directions. We also suggest a method for comparing the performance of a direct and pivot translation routes via automated evaluation of segments with varying translation difficulty.

2007

liu-2010-detecting

https://aclanthology.org/W10-0503.pdf

Detecting Word Misuse in Chinese

Social Network Service (SNS) and personal blogs have become the most popular platform for online communication and sharing information. However because most modern computer keyboards are Latin-based, Asian language speakers (such as Chinese) has to rely on a input system which accepts Romanisation of the characters and convert them into characters or words in that language. In Chinese this form of Romanisation (usually called Pinyin) is highly ambiguous, word misuses often occur because the user choose a wrong candidate or deliverately substitute the word with another character string that has the identical Romanisation to convey certain semantics, or to achieve a sarcasm effect. In this paper we aim to develop a system that can automatically identify such word misuse, and suggest the correct word to be used.

2010

jiang-etal-2020-know

https://aclanthology.org/2020.tacl-1.28.pdf

How Can We Know What Language Models Know?

Recent work has presented intriguing results examining the knowledge contained in language models (LMs) by having the LM fill in the blanks of prompts such as ''Obama is a by profession''. These prompts are usually manually created, and quite possibly sub-optimal; another prompt such as ''Obama worked as a '' may result in more accurately predicting the correct profession. Because of this, given an inappropriate prompt, we might fail to retrieve facts that the LM does know, and thus any given prompt only provides a lower bound estimate of the knowledge contained in an LM. In this paper, we attempt to more accurately estimate the knowledge contained in LMs by automatically discovering better prompts to use in this querying process. Specifically, we propose mining-based and paraphrasing-based methods to automatically generate high-quality and diverse prompts, as well as ensemble methods to combine answers from different prompts. Extensive experiments on the LAMA benchmark for extracting relational knowledge from LMs demonstrate that our methods can improve accuracy from 31.1% to 39.6%, providing a tighter lower bound on what LMs know.

This work was supported by a gift from Bosch Research and NSF award no. 1815287. We would like to thank Paul Michel, Hiroaki Hayashi, Pengcheng Yin, and Shuyan Zhou for their insightful comments and suggestions.

2020

joshi-etal-2013-making

https://aclanthology.org/I13-2006.pdf

Making Headlines in Hindi: Automatic English to Hindi News Headline Translation

News headlines exhibit stylistic peculiarities. The goal of our translation engine 'Making Headlines in Hindi' is to achieve automatic translation of English news headlines to Hindi while retaining the Hindi news headline styles. There are two central modules of our engine: the modified translation unit based on Moses and a co-occurrencebased post-processing unit. The modified translation unit provides two machine translation (MT) models: phrase-based and factor-based (both using in-domain data). In addition, a co-occurrence-based post-processing option may be turned on by a user. Our evaluation shows that this engine handles some linguistic phenomena observed in Hindi news headlines.

2013

tolmachev-etal-2019-shrinking

https://aclanthology.org/N19-1281.pdf

Shrinking Japanese Morphological Analyzers With Neural Networks and Semi-supervised Learning

For languages without natural word boundaries, like Japanese and Chinese, word segmentation is a prerequisite for downstream analysis. For Japanese, segmentation is often done jointly with part of speech tagging, and this process is usually referred to as morphological analysis. Morphological analyzers are trained on data hand-annotated with segmentation boundaries and part of speech tags. A segmentation dictionary or character n-gram information is also provided as additional inputs to the model. Incorporating this extra information makes models large. Modern neural morphological analyzers can consume gigabytes of memory. We propose a compact alternative to these cumbersome approaches which do not rely on any externally provided n-gram or word representations. The model uses only unigram character embeddings, encodes them using either stacked bi-LSTM or a self-attention network, and independently infers both segmentation and part of speech information. The model is trained in an end-to-end and semisupervised fashion, on labels produced by a state-of-the-art analyzer. We demonstrate that the proposed technique rivals performance of a previous dictionary-based state-of-the-art approach and can even surpass it when training with the combination of human-annotated and automatically-annotated data. Our model itself is significantly smaller than the dictionarybased one: it uses less than 15 megabytes of space.

2019

khayrallah-etal-2018-jhu

https://aclanthology.org/W18-6479.pdf

The JHU Parallel Corpus Filtering Systems for WMT 2018

This work describes our submission to the WMT18 Parallel Corpus Filtering shared task. We use a slightly modified version of the Zipporah Corpus Filtering toolkit (Xu and Koehn, 2017), which computes an adequacy score and a fluency score on a sentence pair, and use a weighted sum of the scores as the selection criteria. This work differs from Zipporah in that we experiment with using the noisy corpus to be filtered to compute the combination weights, and thus avoids generating synthetic data as in standard Zipporah.

This work was in part supported by the IARPA MATERIAL project and a Google Faculty Research Award.

2018

ashihara-etal-2019-contextualized

https://aclanthology.org/D19-5552.pdf

Contextualized context2vec

Lexical substitution ranks substitution candidates from the viewpoint of paraphrasability for a target word in a given sentence. There are two major approaches for lexical substitution: (1) generating contextualized word embeddings by assigning multiple embeddings to one word and (2) generating context embeddings using the sentence. Herein we propose a method that combines these two approaches to contextualize word embeddings for lexical substitution. Experiments demonstrate that our method outperforms the current state-ofthe-art method. We also create CEFR-LP, a new evaluation dataset for the lexical substitution task. It has a wider coverage of substitution candidates than previous datasets and assigns English proficiency levels to all target words and substitution candidates.

We thank Professor Christopher G. Haswell for his valuable comments and discussions. We also thank the anonymous reviewers for their valuable comments. This research was supported by the KDDI Foundation.

2019

gonzalez-rubio-etal-2010-saturnalia

http://www.lrec-conf.org/proceedings/lrec2010/pdf/541_Paper.pdf

Saturnalia: A Latin-Catalan Parallel Corpus for Statistical MT

Currently, a great effort is being carried out in the digitalisation of large historical document collections for preservation purposes. The documents in these collections are usually written in ancient languages, such as Latin or Greek, which limits the access of the general public to their content due to the language barrier. Therefore, digital libraries aim not only at storing raw images of digitalised documents, but also to annotate them with their corresponding text transcriptions and translations into modern languages. Unfortunately, ancient languages have at their disposal scarce electronic resources to be exploited by natural language processing techniques. This paper describes the compilation process of a novel Latin-Catalan parallel corpus as a new task for statistical machine translation (SMT). Preliminary experimental results are also reported using a state-of-the-art phrase-based SMT system. The results presented in this work reveal the complexity of the task and its challenging, but interesting nature for future development.

2010

li-etal-2013-multi

https://aclanthology.org/W13-3101.pdf

Multi-document multilingual summarization corpus preparation, Part 1: Arabic, English, Greek, Chinese, Romanian

This document overviews the strategy, effort and aftermath of the MultiLing 2013 multilingual summarization data collection. We describe how the Data Contributors of MultiLing collected and generated a multilingual multi-document summarization corpus on 10 different languages: Arabic, Chinese, Czech, English, French, Greek, Hebrew, Hindi, Romanian and Spanish. We discuss the rationale behind the main decisions of the collection, the methodology used to generate the multilingual corpus, as well as challenges and problems faced per language. This paper overviews the work on Arabic, Chinese, English, Greek, and Romanian languages. A second part, covering the remaining languages, is available as a distinct paper in the MultiLing 2013 proceedings.

2013

liu-etal-2020-hiring

https://aclanthology.org/2020.acl-main.281.pdf

Hiring Now: A Skill-Aware Multi-Attention Model for Job Posting Generation

Writing a good job posting is a critical step in the recruiting process, but the task is often more difficult than many people think. It is challenging to specify the level of education, experience, relevant skills per the company information and job description. To this end, we propose a novel task of Job Posting Generation (JPG) that is cast as a conditional text generation problem to generate job requirements according to the job descriptions. To deal with this task, we devise a data-driven global Skill-Aware Multi-Attention generation model, named SAMA. Specifically, to model the complex mapping relationships between input and output, we design a hierarchical decoder that we first label the job description with multiple skills, then we generate a complete text guided by the skill labels. At the same time, to exploit the prior knowledge about the skills, we further construct a skill knowledge graph to capture the global prior knowledge of skills and refine the generated results. The proposed approach is evaluated on real-world job posting data. Experimental results clearly demonstrate the effectiveness of the proposed method 1 .

Decent Work and Economic Growth

2020

zeng-etal-2019-neural

https://aclanthology.org/D19-1470.pdf

Neural Conversation Recommendation with Online Interaction Modeling

The prevalent use of social media leads to a vast amount of online conversations being produced on a daily basis. It presents a concrete challenge for individuals to better discover and engage in social media discussions. In this paper, we present a novel framework to automatically recommend conversations to users based on their prior conversation behaviors. Built on neural collaborative filtering, our model explores deep semantic features that measure how a user's preferences match an ongoing conversation's context. Furthermore, to identify salient characteristics from interleaving user interactions, our model incorporates graph-structured networks, where both replying relations and temporal features are encoded as conversation context. Experimental results on two large-scale datasets collected from Twitter and Reddit show that our model yields better performance than previous stateof-the-art models, which only utilize lexical features and ignore past user interactions in the conversations.

This work is partially supported by the following HK grants: RGC-GRF (14232816, 14209416, 14204118, 3133237), NSFC (61877020) & ITF (ITS/335/18). Lu Wang is supported in part by National Science Foundation through Grants IIS-1566382 and IIS-1813341. We thank the three anonymous reviewers for the insightful suggestions on various aspects of this work.

2019

hope-etal-2021-extracting

https://aclanthology.org/2021.naacl-main.355.pdf

Extracting a Knowledge Base of Mechanisms from COVID-19 Papers

The COVID-19 pandemic has spawned a diverse body of scientific literature that is challenging to navigate, stimulating interest in automated tools to help find useful knowledge. We pursue the construction of a knowledge base (KB) of mechanisms-a fundamental concept across the sciences, which encompasses activities, functions and causal relations, ranging from cellular processes to economic impacts. We extract this information from the natural language of scientific papers by developing a broad, unified schema that strikes a balance between relevance and breadth. We annotate a dataset of mechanisms with our schema and train a model to extract mechanism relations from papers. Our experiments demonstrate the utility of our KB in supporting interdisciplinary scientific search over COVID-19 literature, outperforming the prominent PubMed search in a study with clinical experts. Our search engine, dataset and code are publicly available. 1 * * Equal contribution. 1 https://covidmechanisms.apps.allenai.org/ … a deep learning framework for design of antiviral candidate drugs Temperature increase can facilitate the destruction of SARS-COV-2 gpl16 antiserum blocks binding of virions to cellular receptors ...food price inflation is an unintended consequence of COVID-19 containment measures Retrieved from CORD-19 papers Ent1: deep learning Ent2: drugs Query mechanism relations

Good Health and Well-Being

Industry, Innovation and Infrastructure

We like to acknowledge a grant from ONR N00014-18-1-2826. Authors would also like to thank anonymous reviewers, members of AI2, UW-NLP and the H2Lab at The University of Washington for their valuable feedback and comments.

2021

paris-vander-linden-1996-building

https://aclanthology.org/C96-2124.pdf

Building Knowledge Bases for the Generation of Software Documentation

Automated text generation requires a underlying knowledge base fl'om which to generate, which is often difficult to produce. Software documentation is one domain in which parts of this knowledge base may be derived automatically. In this paper, we describe DRAFTER, an authoring support tool for generating usercentred software documentation, and in particular, we describe how parts of its required knowledge base can be obtained automatically.

Industry, Innovation and Infrastructure

1996

han-etal-2019-opennre

https://aclanthology.org/D19-3029.pdf

OpenNRE: An Open and Extensible Toolkit for Neural Relation Extraction

OpenNRE is an open-source and extensible toolkit that provides a unified framework to implement neural models for relation extraction (RE). Specifically, by implementing typical RE methods, OpenNRE not only allows developers to train custom models to extract structured relational facts from the plain text but also supports quick model validation for researchers. Besides, OpenNRE provides various functional RE modules based on both TensorFlow and PyTorch to maintain sufficient modularity and extensibility, making it becomes easy to incorporate new models into the framework. Besides the toolkit, we also release an online system to meet real-time extraction without any training and deploying. Meanwhile, the online system can extract facts in various scenarios as well as aligning the extracted facts to Wikidata, which may benefit various downstream knowledge-driven applications (e.g., information retrieval and question answering). More details of the toolkit and online system can be obtained from http://github.com/ thunlp/OpenNRE.

This work is supported by the National Key Research and Development Program of China (No.

2019

albogamy-ramsay-2016-fast

https://aclanthology.org/L16-1238.pdf

Fast and Robust POS tagger for Arabic Tweets Using Agreement-based Bootstrapping

Part-of-Speech (POS) tagging is a key step in many NLP algorithms. However, tweets are difficult to POS tag because they are short, are not always written maintaining formal grammar and proper spelling, and abbreviations are often used to overcome their restricted lengths. Arabic tweets also show a further range of linguistic phenomena such as usage of different dialects, romanised Arabic and borrowing foreign words. In this paper, we present an evaluation and a detailed error analysis of state-of-the-art POS taggers for Arabic when applied to Arabic tweets. On the basis of this analysis, we combine normalisation and external knowledge to handle the domain noisiness and exploit bootstrapping to construct extra training data in order to improve POS tagging for Arabic tweets. Our results show significant improvements over the performance of a number of well-known taggers for Arabic.

The authors would like to thank the anonymous reviewers for their encouraging feedback and insights. Fahad would also like to thank King Saud University for their financial support. Allan Ramsay's contribution to this work was partially supported by Qatar National Research Foundation (grant NPRP-7-1334-6 -039).

2016

danescu-niculescu-mizil-etal-2009-without

https://aclanthology.org/N09-1016.pdf

Without a 'doubt'? Unsupervised Discovery of Downward-Entailing Operators

An important part of textual inference is making deductions involving monotonicity, that is, determining whether a given assertion entails restrictions or relaxations of that assertion. For instance, the statement 'We know the epidemic spread quickly' does not entail 'We know the epidemic spread quickly via fleas', but 'We doubt the epidemic spread quickly' entails 'We doubt the epidemic spread quickly via fleas'. Here, we present the first algorithm for the challenging lexical-semantics problem of learning linguistic constructions that, like 'doubt', are downward entailing (DE). Our algorithm is unsupervised, resource-lean, and effective, accurately recovering many DE operators that are missing from the handconstructed lists that textual-inference systems currently use.

Acknowledgments We thank Roy Bar-Haim, Cleo Condoravdi, and Bill MacCartney for sharing their systems' lists and information about their work with us; Mats Rooth for helpful conversations; Alex Niculescu-Mizil for technical assistance; and Eugene Charniak for reassuring remarks. We also thank Marisa Ferrara Boston, Claire Cardie, Zhong Chen, Yejin Choi, Effi Georgala, Myle Ott, Stephen Purpura, and Ainur Yessenalina at Cornell University, the UT-Austin NLP group, Roy Bar-Haim, Bill MacCartney, and the anonymous reviewers for for their comments on this paper. This paper is based upon work supported in part by DHS grant N0014-07-1-0152, National Science Foundation grant No. BCS-0537606, a Yahoo! Research Alliance gift, a CU Provost's Award for Distinguished Scholarship, and a CU Institute for the Social Sciences Faculty Fellowship. Any opinions, findings, and conclusions or recommendations expressed are those of the authors and do not necessarily reflect the views or official policies, either expressed or implied, of any sponsoring institutions, the U.S. government, or any other entity.

2009

acl-1993-association

https://aclanthology.org/P93-1000.pdf

31st Annual Meeting of the Association for Computational Linguistics

This volume contains the papers prepared for the 31 st Annual Meeting of the Association for Computational Linguistics, held 22-26 June 1993 at The Ohio State University in Columbus, Ohio. The cluster of papers in the final section stems from the student session, featured at the meeting for the 3rd successive year and testifying to the vigor of this emerging tradition. The number and quality of submitted papers was again gratifying, and all authors deserve our collective plaudits for the efforts they invested despite the well-known risks of submitting to a highly selective conference. It was their efforts that once again ensured a Meeting (and Proceedings) reflecting the highest standards in computational linguistics, offering a tour of some of the most significant recent advances and most lively research frontiers. Special thanks go to our invited speakers, Wolfgang Wahlster, Geoff Nunberg and Barbara Partee, for contributing their insights and panache to the conference; to Philip Cohen for concocting and coordinating a varied and relevant tutorial program, and to

We thank the reviewers for providing providing helpful, detailed reviews of the submissions, and for completing the reviews promptly. The careful thought that went into their review comments was obvious and impressive, and we are sure the student authors found the reviews beneficial. The Program Committee included the members of the Planning Committee and the following non-student members: Mary Dal-

1993

hakala-etal-2013-evex

https://aclanthology.org/W13-2004.pdf

EVEX in ST'13: Application of a large-scale text mining resource to event extraction and network construction

During the past few years, several novel text mining algorithms have been developed in the context of the BioNLP Shared Tasks on Event Extraction. These algorithms typically aim at extracting biomolecular interactions from text by inspecting only the context of one sentence. However, when humans interpret biomolecular research articles, they usually build upon extensive background knowledge of their favorite genes and pathways. To make such world knowledge available to a text mining algorithm, it could first be applied to all available literature to subsequently make a more informed decision on which predictions are consistent with the current known data. In this paper, we introduce our participation in the latest Shared Task using the largescale text mining resource EVEX which we previously implemented using state-ofthe-art algorithms, and which was applied to the whole of PubMed and PubMed Central. We participated in the Genia Event Extraction (GE) and Gene Regulation Network (GRN) tasks, ranking first in the former and fifth in the latter.

Computational resources were provided by CSC IT Center for Science Ltd., Espoo, Finland. The work of KH and FG was supported by the Academy of Finland, and of SVL by the Research Foundation Flanders (FWO). YVdP and SVL acknowledge the support from Ghent University (Multidisciplinary Research Partnership Bioinformatics: from nucleotides to networks).

2013

banerjee-etal-2021-scrambled

https://aclanthology.org/2021.mtsummit-research.11.pdf

Scrambled Translation Problem: A Problem of Denoising UNMT

In this paper, we identify an interesting kind of error in the output of Unsupervised Neural Machine Translation (UNMT) systems like Undreamt 1. We refer to this error type as Scrambled Translation problem. We observe that UNMT models which use word shuffle noise (as in case of Undreamt) can generate correct words, but fail to stitch them together to form phrases. As a result, words of the translated sentence look scrambled, resulting in decreased BLEU. We hypothesise that the reason behind scrambled translation problem is 'shuffling noise' which is introduced in every input sentence as a denoising strategy. To test our hypothesis, we experiment by retraining UNMT models with a simple retraining strategy. We stop the training of the Denoising UNMT model after a pre-decided number of iterations and resume the training for the remaining iterations-which number is also pre-decided-using original sentence as input without adding any noise. Our proposed solution achieves significant performance improvement UNMT models that train conventionally. We demonstrate these performance gains on four language pairs, viz., English-French, English-German, English-Spanish, Hindi-Punjabi. Our qualitative and quantitative analysis shows that the retraining strategy helps achieve better alignment as observed by attention heatmap and better phrasal translation, leading to statistically significant improvement in BLEU scores.

2021

jurgens-etal-2014-twitter

https://aclanthology.org/W14-3906.pdf

Twitter Users \#CodeSwitch Hashtags! \#MoltoImportante \#wow

When code switching, individuals incorporate elements of multiple languages into the same utterance. While code switching has been studied extensively in formal and spoken contexts, its behavior and prevalence remains unexamined in many newer forms of electronic communication. The present study examines code switching in Twitter, focusing on instances where an author writes a post in one language and then includes a hashtag in a second language. In the first experiment, we perform a large scale analysis on the languages used in millions of posts to show that authors readily incorporate hashtags from other languages, and in a manual analysis of a subset the hashtags, reveal prolific code switching, with code switching occurring for some hashtags in over twenty languages. In the second experiment, French and English posts from three bilingual cities are analyzed for their code switching frequency and its content.

2014

ilinykh-dobnik-2022-attention

https://aclanthology.org/2022.findings-acl.320.pdf

Attention as Grounding: Exploring Textual and Cross-Modal Attention on Entities and Relations in Language-and-Vision Transformer

We explore how a multi-modal transformer trained for generation of longer image descriptions learns syntactic and semantic representations about entities and relations grounded in objects at the level of masked self-attention (text generation) and crossmodal attention (information fusion). We observe that cross-attention learns the visual grounding of noun phrases into objects and high-level semantic information about spatial relations, while text-to-text attention captures low-level syntactic knowledge between words. This concludes that language models in a multi-modal task learn different semantic information about objects and relations cross-modally and uni-modally (text-only).

The research reported in this paper was supported by a grant from the Swedish Research Council (VR project 2014-39) for the establishment of the Centre for Linguistic Theory and Studies in Probability (CLASP) at the University of Gothenburg.

2022

hovy-2010-distributional

https://aclanthology.org/W10-3401.pdf

Distributional Semantics and the Lexicon

The lexicons used in computational linguistics systems contain morphological, syntactic, and occasionally also some semantic information (such as definitions, pointers to an ontology, verb frame filler preferences, etc.). But the human cognitive lexicon contains a great deal more, crucially, expectations about how a word tends to combine with others: not just general information-extraction-like patterns, but specific instantial expectations. Such information is very useful when it comes to listening in bad aural conditions and reading texts in which background information is taken for granted; without such specific expectation, one would be hard-pressed (and computers are completely unable) to form coherent and richly connected multi-sentence interpretations. Over the past few years, NLP work has increasingly treated topic signature word distributions (also called 'context vectors', 'topic models', etc.) as a de facto replacement for semantics. Whether the task is wordsense disambiguation, certain forms of textual entailment, information extraction, paraphrase learning, and so on, it turns out to be very useful to consider a word(sense) as being defined by the distribution of word(senses) that regu-larly accompany it (in the classic words of Firth, "you shall know a word by the company it keeps"). And this is true not only for individual wordsenses, but also for larger units such as topics: the product of LDA and similar topic characterization engines is similar.

2010

osenova-etal-2010-exploring

http://www.lrec-conf.org/proceedings/lrec2010/pdf/721_Paper.pdf

Exploring Co-Reference Chains for Concept Annotation of Domain Texts

The paper explores the co-reference chains as a way for improving the density of concept annotation over domain texts. The idea extends authors' previous work on relating the ontology to the text terms in two domains-IT and textile. Here IT domain is used. The challenge is to enhance relations among concepts instead of text entities, the latter pursued in most works. Our ultimate goal is to exploit these additional chains for concept disambiguation as well as sparseness resolution at concept level. First, a gold standard was prepared with manually connected links among concepts, anaphoric pronouns and contextual equivalents. This step was necessary not only for test purposes, but also for better orientation in the co-referent types and distribution. Then, two automatic systems were tested on the gold standard. Note that these systems were not designed specially for concept chaining. The conclusion is that the state-of-the-art co-reference resolution systems might address the concept sparseness problem, but not so much the concept disambiguation task. For the latter, word-sense disambiguation systems have to be integrated.

The work reported here is done within the context of the EU project -Language Technology for Lifelong Learning (LTfLL). We would also like to thank the three anonymous reviewers for their valuable remarks as specialists and readers.

2010

nallapati-etal-2016-abstractive

https://aclanthology.org/K16-1028.pdf

Abstractive Text Summarization using Sequence-to-sequence RNNs and Beyond

In this work, we model abstractive text summarization using Attentional Encoder-Decoder Recurrent Neural Networks, and show that they achieve state-of-the-art performance on two different corpora. We propose several novel models that address critical problems in summarization that are not adequately modeled by the basic architecture, such as modeling keywords , capturing the hierarchy of sentence-toword structure, and emitting words that are rare or unseen at training time. Our work shows that many of our proposed models contribute to further improvement in performance. We also propose a new dataset consisting of multi-sentence summaries, and establish performance benchmarks for further research.

2016

reinhard-gibbon-1991-prosodic

https://aclanthology.org/E91-1023.pdf

Prosodic Inheritance and Morphological Generalisations

Prosodic Inheritance (PI) morphology provides uniform treatment of both concatenative and non-concatenative morphological and phonological generalisations using default inheritance. Models of an extensive range of German Umlaut and Arabic intercalation facts, implemented in DATR, show that the PI approach also covers 'hard cases' more homogeneously and more extensively than previous computational treatments.

1991

walker-etal-2012-annotated

http://www.lrec-conf.org/proceedings/lrec2012/pdf/1114_Paper.pdf

An Annotated Corpus of Film Dialogue for Learning and Characterizing Character Style

Interactive story systems often involve dialogue with virtual dramatic characters. However, to date most character dialogue is written by hand. One way to ease the authoring process is to (semi-)automatically generate dialogue based on film characters. We extract features from dialogue of film characters in leading roles. Then we use these character-based features to drive our language generator to produce interesting utterances. This paper describes a corpus of film dialogue that we have collected from the IMSDb archive and annotated for linguistic structures and character archetypes. We extract different sets of features using external sources such as LIWC and SentiWordNet as well as using our own written scripts. The automation of feature extraction also eases the process of acquiring additional film scripts. We briefly show how film characters can be represented by models learned from the corpus, how the models can be distinguished based on different categories such as gender and film genre, and how they can be applied to a language generator to generate utterances that can be perceived as being similar to the intended character model.

2012

zhai-huang-2015-pilot

https://aclanthology.org/2015.mtsummit-papers.5.pdf

A pilot study towards end-to-end MT training

Typical MT training involves several stages, including word alignment, rule extraction, translation model estimation, and parameter tuning. In this paper, different from the traditional pipeline, we investigate the possibility of end-to-end MT training, and propose a framework which combines rule induction and parameter tuning in one single module. Preliminary experiments show that our learned model achieves comparable translation quality to the traditional MT training pipeline. * Work done while Prof. Liang Huang was in City University of New York.

We thank the three anonymous reviewers for the valuable comments, and Kai Zhao for discussions. This project was supported in part by DARPA FA8750-13-2-0041 (DEFT), NSF IIS-1449278, and a Google Faculty Research Award.

2015

kucuk-etal-2014-named

http://www.lrec-conf.org/proceedings/lrec2014/pdf/380_Paper.pdf

Named Entity Recognition on Turkish Tweets

Various recent studies show that the performance of named entity recognition (NER) systems developed for well-formed text types drops significantly when applied to tweets. The only existing study for the highly inflected agglutinative language Turkish reports a drop in F-Measure from 91% to 19% when ported from news articles to tweets. In this study, we present a new named entity-annotated tweet corpus and a detailed analysis of the various tweet-specific linguistic phenomena. We perform comparative NER experiments with a rule-based multilingual NER system adapted to Turkish on three corpora: a news corpus, our new tweet corpus, and another tweet corpus. Based on the analysis and the experimentation results, we suggest system features required to improve NER results for social media like Twitter.

This study is supported in part by a postdoctoral research grant from TÜBİTAK.

2014

gilbert-riloff-2013-domain

https://aclanthology.org/P13-2015.pdf

Domain-Specific Coreference Resolution with Lexicalized Features

Most coreference resolvers rely heavily on string matching, syntactic properties, and semantic attributes of words, but they lack the ability to make decisions based on individual words. In this paper, we explore the benefits of lexicalized features in the setting of domain-specific coreference resolution. We show that adding lexicalized features to off-the-shelf coreference resolvers yields significant performance gains on four domain-specific data sets and with two types of coreference resolution architectures.

This material is based upon work supported by the National Science Foundation under Grant No. IIS-1018314 and the Defense Advanced Research Projects Agency (DARPA) Machine Reading Program under Air Force Research Laboratory (AFRL) prime contract no. FA8750-09-C-0172. Any opinions, findings, and conclusion or recommendations expressed in this material are those of the authors and do not necessarily reflect the view of the DARPA, AFRL, or the U.S. government.

2013

alvez-etal-2018-cross

https://aclanthology.org/L18-1723.pdf

Cross-checking WordNet and SUMO Using Meronymy

We report on the practical application of a black-box testing methodology for the validation of the knowledge encoded in WordNet, SUMO and their mapping by using automated theorem provers. Our proposal is based on the part-whole information provided by WordNet, out of which we automatically create a large set of tests. Our experimental results confirm that the proposed system enables the validation of some pieces of information and also the detection of missing information or inconsistencies among these resources.

This work has been partially funded by the Spanish Projects TUNER (TIN2015-65308-C5-1-R) and GRAMM (TIN2017-86727-C2-2-R), the Basque Project LoRea (GIU15/30) and the UPV/EHU project OEBU (EHUA16/33).

2018

paul-etal-2009-mining

https://aclanthology.org/W09-1111.pdf

Mining the Web for Reciprocal Relationships

In this paper we address the problem of identifying reciprocal relationships in English. In particular we introduce an algorithm that semi-automatically discovers patterns encoding reciprocity based on a set of simple but effective pronoun templates. Using a set of most frequently occurring patterns, we extract pairs of reciprocal pattern instances by searching the web. Then we apply two unsupervised clustering procedures to form meaningful clusters of such reciprocal instances. The pattern discovery procedure yields an accuracy of 97%, while the clustering procedures indicate accuracies of 91% and 82%. Moreover, the resulting set of 10,882 reciprocal instances represent a broad-coverage resource.

Partnership for the goals

2009

hatzivassiloglou-mckeown-1995-quantitative

https://aclanthology.org/P95-1027.pdf

A Quantitative Evaluation of Linguistic Tests for the Automatic Prediction of Semantic Markedness

We present a corpus-based study of methods that have been proposed in the linguistics literature for selecting the semantically unmarked term out of a pair of antonymous adjectives. Solutions to this problem are applicable to the more general task of selecting the positive term from the pair. Using automatically collected data, the accuracy and applicability of each method is quantified, and a statistical analysis of the significance of the results is performed. We show that some simple methods are indeed good indicators for the answer to the problem while other proposed methods fail to perform better than would be attributable to chance. In addition, one of the simplest methods, text frequency, dominates all others. We also apply two generic statistical learning methods for combining the indications of the individual methods, and compare their performance to the simple methods. The most sophisticated complex learning method offers a small, but statistically significant, improvement over the original tests.

This work was supported jointly by the Advanced Research Projects Agency and the Office of Naval Research under contract N00014-89-J-1782, and by the National Science Foundation under contract GER-90-24069. It was conducted under the auspices of the Columbia University CAT in High Performance Computing and Communications in Healthcare, a New York State Center for Advanced Technology supported by the New York State Science and Technology Foundation. We wish to thank Judith Klavans, Rebecca Passonneau, and the anonymous reviewers for providing us with useful comments on earlier versions of the paper.

1995

bangalore-etal-2006-finite

https://aclanthology.org/2006.iwslt-evaluation.2.pdf

Finite-state transducer-based statistical machine translation using joint probabilities

In this paper, we present our system for statistical machine translation that is based on weighted finite-state transducers. We describe the construction of the transducer, the estimation of the weights, acquisition of phrases (locally ordered tokens) and the mechanism we use for global reordering. We also present a novel approach to machine translation that uses a maximum entropy model for parameter estimation and contrast its performance to the finite-state translation model on the IWSLT Chinese-English data sets.

2006

rohith-ramakrishnan-etal-2021-analysis

https://aclanthology.org/2021.paclic-1.75.pdf

Analysis of Text-Semantics via Efficient Word Embedding using Variational Mode Decomposition

In this paper, we propose a novel method which establishes a newborn relation between Signal Processing and Natural Language Processing (NLP) method via Variational Mode Decomposition (VMD). Unlike the modern Neural Network approaches for NLP which are complex and often masked from the end user, our approach involving Term Frequency-Inverse Document Frequency (TF-IDF) aided with VMD dials down the complexity retaining the performance with transparency. The performance in terms of Machine Learning based approaches and semantic relationships of words along with the methodology of the above mentioned approach is analyzed and discussed in this paper.

2021

di-eugenio-glass-2004-squibs

https://aclanthology.org/J04-1005.pdf

Squibs and Discussions: The Kappa Statistic: A Second Look

In recent years, the kappa coefficient of agreement has become the de facto standard for evaluating intercoder agreement for tagging tasks. In this squib, we highlight issues that affect κ and that the community has largely neglected. First, we discuss the assumptions underlying different computations of the expected agreement component of κ. Second, we discuss how prevalence and bias affect the κ measure.

This work is supported by grant N00014-00-1-0640 from the Office of Naval Research. Thanks to Janet Cahn and to the anonymous reviewers for comments on earlier drafts.

2004

le-roux-etal-2013-combining

https://aclanthology.org/D13-1116.pdf

Combining PCFG-LA Models with Dual Decomposition: A Case Study with Function Labels and Binarization

It has recently been shown that different NLP models can be effectively combined using dual decomposition. In this paper we demonstrate that PCFG-LA parsing models are suitable for combination in this way. We experiment with the different models which result from alternative methods of extracting a grammar from a treebank (retaining or discarding function labels, left binarization versus right binarization) and achieve a labeled Parseval F-score of 92.4 on Wall Street Journal Section 23-this represents an absolute improvement of 0.7 and an error reduction rate of 7% over a strong PCFG-LA product-model baseline. Although we experiment only with binarization and function labels in this study, there is much scope for applying this approach to other grammar extraction strategies.

We are grateful to the reviewers for their helpful comments. We also thank Joachim Wagner for providing feedback on an early version of the paper. This work has been partially funded by the Labex EFL (ANR/CGI). 9 Their other system relying on the self-trained version of the BLLIP parser achieves 92.6 F1. ACL-08: HLT, pages 586-594.

2013

marie-fujita-2019-unsupervised-joint

https://aclanthology.org/P19-1312.pdf

Unsupervised Joint Training of Bilingual Word Embeddings

State-of-the-art methods for unsupervised bilingual word embeddings (BWE) train a mapping function that maps pre-trained monolingual word embeddings into a bilingual space. Despite its remarkable results, unsupervised mapping is also well-known to be limited by the dissimilarity between the original word embedding spaces to be mapped. In this work, we propose a new approach that trains unsupervised BWE jointly on synthetic parallel data generated through unsupervised machine translation. We demonstrate that existing algorithms that jointly train BWE are very robust to noisy training data and show that unsupervised BWE jointly trained significantly outperform unsupervised mapped BWE in several cross-lingual NLP tasks.

We would like to thank the reviewers for their useful comments and suggestions. A part of this work was conducted under the program "Promotion of Global Communications Plan: Research, Development, and Social Demonstration of Multilingual Speech Translation Technology" of the Ministry of Internal Affairs and Communications (MIC), Japan. 13 We used the News Commentary corpora provided by WMT for en→de and en→fr to train SMT systems performing at 15.4 and 20.1 BLEU points on Newstest2016 en-de and Newstest2014 en-fr, respectively.

2019

kuo-etal-2012-exploiting

https://aclanthology.org/P12-2067.pdf

Exploiting Latent Information to Predict Diffusions of Novel Topics on Social Networks

This paper brings a marriage of two seemly unrelated topics, natural language processing (NLP) and social network analysis (SNA). We propose a new task in SNA which is to predict the diffusion of a new topic, and design a learning-based framework to solve this problem. We exploit the latent semantic information among users, topics, and social connections as features for prediction. Our framework is evaluated on real data collected from public domain. The experiments show 16% AUC

Peace, Justice and Strong Institutions

This work was also supported by National Science Council, National Taiwan University and Intel Corporation under Grants NSC 100-2911-I-002-001, and 101R7501.

2012

druck-etal-2009-active

https://aclanthology.org/D09-1009.pdf

Active Learning by Labeling Features

Methods that learn from prior information about input features such as generalized expectation (GE) have been used to train accurate models with very little effort. In this paper, we propose an active learning approach in which the machine solicits "labels" on features rather than instances. In both simulated and real user experiments on two sequence labeling tasks we show that our active learning method outperforms passive learning with features as well as traditional active learning with instances. Preliminary experiments suggest that novel interfaces which intelligently solicit labels on multiple features facilitate more efficient annotation.

We thank Kedar Bellare for helpful discussions and Gau-

2009

wang-etal-2021-enpar

https://aclanthology.org/2021.eacl-main.251.pdf

ENPAR:Enhancing Entity and Entity Pair Representations for Joint Entity Relation Extraction

Current state-of-the-art systems for joint entity relation extraction (Luan et al., 2019; Wadden et al., 2019) usually adopt the multi-task learning framework. However, annotations for these additional tasks such as coreference resolution and event extraction are always equally hard (or even harder) to obtain. In this work, we propose a pre-training method ENPAR to improve the joint extraction performance. EN-PAR requires only the additional entity annotations that are much easier to collect. Unlike most existing works that only consider incorporating entity information into the sentence encoder, we further utilize the entity pair information. Specifically, we devise four novel objectives, i.e., masked entity typing, masked entity prediction, adversarial context discrimination, and permutation prediction, to pretrain an entity encoder and an entity pair encoder. Comprehensive experiments show that the proposed pre-training method achieves significant improvement over BERT on ACE05, SciERC, and NYT, and outperforms current state-of-the-art on ACE05.

The authors wish to thank the reviewers for their helpful comments and suggestions. This research is (partially) supported by NSFC (62076097

2021

ling-etal-2015-contexts

https://aclanthology.org/D15-1161.pdf

Not All Contexts Are Created Equal: Better Word Representations with Variable Attention

We introduce an extension to the bag-ofwords model for learning words representations that take into account both syntactic and semantic properties within language. This is done by employing an attention model that finds within the contextual words, the words that are relevant for each prediction. The general intuition of our model is that some words are only relevant for predicting local context (e.g. function words), while other words are more suited for determining global context, such as the topic of the document. Experiments performed on both semantically and syntactically oriented tasks show gains using our model over the existing bag of words model. Furthermore, compared to other more sophisticated models, our model scales better as we increase the size of the context of the model.

The PhD thesis of Wang Ling is supported by FCT grant SFRH/BD/51157/2010. This research was supported in part by the U.S. Army Research Laboratory, the U.S. Army Research Office under contract/grant number W911NF-10-1-0533 and NSF IIS-1054319 and FCT through the plurianual contract UID/CEC/50021/2013 and grant number SFRH/BPD/68428/2010.

2015

xu-etal-2002-study

https://aclanthology.org/P02-1025.pdf

A Study on Richer Syntactic Dependencies for Structured Language Modeling

We study the impact of richer syntactic dependencies on the performance of the structured language model (SLM) along three dimensions: parsing accuracy (LP/LR), perplexity (PPL) and worderror-rate (WER, N-best re-scoring). We show that our models achieve an improvement in LP/LR, PPL and/or WER over the reported baseline results using the SLM on the UPenn Treebank and Wall Street Journal (WSJ) corpora, respectively. Analysis of parsing performance shows correlation between the quality of the parser (as measured by precision/recall) and the language model performance (PPL and WER). A remarkable fact is that the enriched SLM outperforms the baseline 3-gram model in terms of WER by 10% when used in isolation as a second pass (N-best re-scoring) language model.

2002

barron-cedeno-etal-2016-convkn

https://aclanthology.org/S16-1138.pdf

ConvKN at SemEval-2016 Task 3: Answer and Question Selection for Question Answering on Arabic and English Fora

We describe our system, ConvKN, participating to the SemEval-2016 Task 3 "Community Question Answering". The task targeted the reranking of questions and comments in real-life web fora both in English and Arabic. ConvKN combines convolutional tree kernels with convolutional neural networks and additional manually designed features including text similarity and thread specific features. For the first time, we applied tree kernels to syntactic trees of Arabic sentences for a reranking task. Our approaches obtained the second best results in three out of four tasks. The only task we performed averagely is the one where we did not use tree kernels in our classifier.

This research is developed by the Arabic Language Technologies (ALT) group at the Qatar Computing Research Institute (QCRI), HBKU, Qatar Foundation in collaboration with MIT. It is part of the Interactive sYstems for Answer Search (IYAS) project. This work has been partially supported by the EC project CogNet, 671625 (H2020-ICT-2014-2, Research and Innovation action) and by an IBM Faculty Award.

2016

schwenk-2012-continuous

https://aclanthology.org/C12-2104.pdf

Continuous Space Translation Models for Phrase-Based Statistical Machine Translation

This paper presents a new approach to perform the estimation of the translation model probabilities of a phrase-based statistical machine translation system. We use neural networks to directly learn the translation probability of phrase pairs using continuous representations. The system can be easily trained on the same data used to build standard phrase-based systems. We provide experimental evidence that the approach seems to be able to infer meaningful translation probabilities for phrase pairs not seen in the training data, or even predict a list of the most likely translations given a source phrase. The approach can be used to rescore n-best lists, but we also discuss an integration into the Moses decoder. A preliminary evaluation on the English/French IWSLT task achieved improvements in the BLEU score and a human analysis showed that the new model often chooses semantically better translations. Several extensions of this work are discussed.

This work was partially financed by the French government (COSMAT, ANR-09-CORD-004), the European Commission (MATECAT, ICT-2011.4.2 -287688) and the DARPA BOLT project.

2012

plank-etal-2016-multilingual

https://aclanthology.org/P16-2067.pdf

Multilingual Part-of-Speech Tagging with Bidirectional Long Short-Term Memory Models and Auxiliary Loss

Bidirectional long short-term memory (bi-LSTM) networks have recently proven successful for various NLP sequence modeling tasks, but little is known about their reliance to input representations, target languages, data set size, and label noise. We address these issues and evaluate bi-LSTMs with word, character, and unicode byte embeddings for POS tagging. We compare bi-LSTMs to traditional POS taggers across languages and data sizes. We also present a novel bi-LSTM model, which combines the POS tagging loss function with an auxiliary loss function that accounts for rare words. The model obtains state-of-the-art performance across 22 languages, and works especially well for morphologically complex languages. Our analysis suggests that bi-LSTMs are less sensitive to training data size and label corruptions (at small noise levels) than previously assumed.

We thank the anonymous reviewers for their feedback. AS is funded by the ERC Starting Grant LOWLANDS No. 313695. YG is supported by The Israeli Science Foundation (grant number 1555/15) and a Google Research Award.

2016

jiang-etal-2021-lnn

https://aclanthology.org/2021.acl-long.64.pdf

LNN-EL: A Neuro-Symbolic Approach to Short-text Entity Linking

Entity linking (EL), the task of disambiguating mentions in text by linking them to entities in a knowledge graph, is crucial for text understanding, question answering or conversational systems. Entity linking on short text (e.g., single sentence or question) poses particular challenges due to limited context. While prior approaches use either heuristics or blackbox neural methods, here we propose LNN-EL, a neuro-symbolic approach that combines the advantages of using interpretable rules based on first-order logic with the performance of neural learning. Even though constrained to using rules, LNN-EL performs competitively against SotA black-box neural approaches, with the added benefits of extensibility and transferability. In particular, we show that we can easily blend existing rule templates given by a human expert, with multiple types of features (priors, BERT encodings, box embeddings, etc), and even scores resulting from previous EL methods, thus improving on such methods. For instance, on the LC-QuAD-1.0 dataset, we show more than 4% increase in F1 score over previous SotA. Finally, we show that the inductive bias offered by using logic results in learned rules that transfer well across datasets, even without fine tuning, while maintaining high accuracy. * Equal contribution; Author Hang Jiang did this work while interning at IBM.

We thank Ibrahim Abdelaziz, Pavan Kapanipathi, Srinivas Ravishankar, Berthold Reinwald, Salim Roukos and anonymous reviewers for their valuable inputs and feedback.

2021

jimenez-lopez-becerra-bonache-2016-machine

https://aclanthology.org/W16-4101.pdf

Could Machine Learning Shed Light on Natural Language Complexity?

In this paper, we propose to use a subfield of machine learning-grammatical inference-to measure linguistic complexity from a developmental point of view. We focus on relative complexity by considering a child learner in the process of first language acquisition. The relevance of grammatical inference models for measuring linguistic complexity from a developmental point of view is based on the fact that algorithms proposed in this area can be considered computational models for studying first language acquisition. Even though it will be possible to use different techniques from the field of machine learning as computational models for dealing with linguistic complexity-since in any model we have algorithms that can learn from data-, we claim that grammatical inference models offer some advantages over other tools.

This research has been supported by the Ministerio de Economía y Competitividad under the project number FFI2015-69978-P (MINECO/FEDER) of the Programa Estatal de Fomento de la Investigación Científica y Técnica de Excelencia, Subprograma Estatal de Generación de Conocimiento.

2016

song-etal-2010-active

https://aclanthology.org/W10-4121.pdf

Active Learning Based Corpus Annotation

Opinion Mining aims to automatically acquire useful opinioned information and knowledge in subjective texts. Research of Chinese Opinioned Mining requires the support of annotated corpus for Chinese opinioned-subjective texts. To facilitate the work of corpus annotators, this paper implements an active learning based annotation tool for Chinese opinioned elements which can identify topic, sentiment, and opinion holder in a sentence automatically.

The author of this paper would like to thank Information Retrieval Lab, Harbin Institute of Technology for providing the tool (LTP) used in experiments. This research was supported by National Natural Science Foundation of China Grant No.60773087.

2010

rello-basterrechea-2010-automatic

https://aclanthology.org/W10-0301.pdf

Automatic conjugation and identification of regular and irregular verb neologisms in Spanish

In this paper, a novel system for the automatic identification and conjugation of Spanish verb neologisms is presented. The paper describes a rule-based algorithm consisting of six steps which are taken to determine whether a new verb is regular or not, and to establish the rules that the verb should follow in its conjugation. The method was evaluated on 4,307 new verbs and its performance found to be satisfactory both for irregular and regular neologisms. The algorithm also contains extra rules to cater for verb neologisms in Spanish that do not exist as yet, but are inferred to be possible in light of existing cases of new verb creation in Spanish.

We would like to express or gratitude to the Molino de Ideas s.a. engineering team who have successfully implemented the method, specially to Daniel Ayuso de Santos and Alejandro de Pablos López.

2010

saint-dizier-2016-argument

https://aclanthology.org/L16-1156.pdf

Argument Mining: the Bottleneck of Knowledge and Language Resources

Given a controversial issue, argument mining from natural language texts (news papers, and any form of text on the Internet) is extremely challenging: domain knowledge is often required together with appropriate forms of inferences to identify arguments. This contribution explores the types of knowledge that are required and how they can be paired with reasoning schemes, language processing and language resources to accurately mine arguments. We show via corpus analysis that the Generative Lexicon, enhanced in different manners and viewed as both a lexicon and a domain knowledge representation, is a relevant approach. In this paper, corpus annotation for argument mining is first developed, then we show how the generative lexicon approach must be adapted and how it can be paired with language processing patterns to extract and specify the nature of arguments. Our approach to argument mining is thus knowledge driven

2016

pasca-2015-interpreting

https://aclanthology.org/N15-1037.pdf

Interpreting Compound Noun Phrases Using Web Search Queries

A weakly-supervised method is applied to anonymized queries to extract lexical interpretations of compound noun phrases (e.g., "fortune 500 companies"). The interpretations explain the subsuming role ("listed in") that modifiers (fortune 500) play relative to heads (companies) within the noun phrases. Experimental results over evaluation sets of noun phrases from multiple sources demonstrate that interpretations extracted from queries have encouraging coverage and precision. The top interpretation extracted is deemed relevant for more than 70% of the noun phrases.

The paper benefits from comments from Jutta Degener, Mihai Surdeanu and Susanne Riehemann. Data extracted by Haixun Wang and Jian Li is the source of the IsA vocabulary of noun phrases used in the evaluation.

2015

haffari-etal-2011-ensemble

https://aclanthology.org/P11-2125.pdf

An Ensemble Model that Combines Syntactic and Semantic Clustering for Discriminative Dependency Parsing

We combine multiple word representations based on semantic clusters extracted from the (Brown et al., 1992) algorithm and syntactic clusters obtained from the Berkeley parser (Petrov et al., 2006) in order to improve discriminative dependency parsing in the MST-Parser framework (McDonald et al., 2005). We also provide an ensemble method for combining diverse cluster-based models. The two contributions together significantly improves unlabeled dependency accuracy from 90.82% to 92.13%.

This research was partially supported by NSERC, Canada (RGPIN: 264905). We would like to thank Terry Koo for his help with the cluster-based features for dependency parsing and Ryan McDonald for the MSTParser source code which we modified and used for the experiments in this paper.

2011

chung-2005-market

https://aclanthology.org/Y05-1007.pdf

MARKET Metaphors: Chinese, English and Malay

In this paper, MARKET metaphors used by different communities (Chinese, Malay and English) are laid out based on the frequency counts of these metaphors and their occurrences in different syntactic positions. The results show that certain types of metaphors have preferences for different syntactic positions for 'market.' For instance, MARKET IS A PERSON in all three languages prefers to place 'market' in the subject position. In addition to this finding, the choice of metaphor types by different speech communities may also reflect their perspectives regarding their country's economy. This is evidenced by the fewer instances of MARKET IS COMPETITION in the English data. The instances that describe how the market falls (plunges and crashes) may reflect the speakers' concerns with the maintenance of their power in the market rather than the competitiveness of their market. Therefore, through using quantitative data, this paper is able to infer the economic status of these speech communities. This can be done not only through analyzing the semantic meanings of the metaphors but also their interface with syntax.

2005

gliwa-etal-2019-samsum

https://aclanthology.org/D19-5409.pdf

SAMSum Corpus: A Human-annotated Dialogue Dataset for Abstractive Summarization

This paper introduces the SAMSum Corpus, a new dataset with abstractive dialogue summaries. We investigate the challenges it poses for automated summarization by testing several models and comparing their results with those obtained on a corpus of news articles. We show that model-generated summaries of dialogues achieve higher ROUGE scores than the model-generated summaries of news-in contrast with human evaluators' judgement. This suggests that a challenging task of abstractive dialogue summarization requires dedicated models and non-standard quality measures. To our knowledge, our study is the first attempt to introduce a high-quality chatdialogues corpus, manually annotated with abstractive summarizations, which can be used by the research community for further studies.

We would like to express our sincere thanks to Tunia Błachno, Oliwia Ebebenge, Monika Jędras and Małgorzata Krawentek for their huge contribution to the corpus collection -without their ideas, management of the linguistic task and verification of examples we would not be able to create this paper. We are also grateful for the reviewers' helpful comments and suggestions.

2019

klafka-ettinger-2020-spying

https://aclanthology.org/2020.acl-main.434.pdf

Spying on Your Neighbors: Fine-grained Probing of Contextual Embeddings for Information about Surrounding Words

Although models using contextual word embeddings have achieved state-of-the-art results on a host of NLP tasks, little is known about exactly what information these embeddings encode about the context words that they are understood to reflect. To address this question, we introduce a suite of probing tasks that enable fine-grained testing of contextual embeddings for encoding of information about surrounding words. We apply these tasks to examine the popular BERT, ELMo and GPT contextual encoders, and find that each of our tested information types is indeed encoded as contextual information across tokens, often with nearperfect recoverability-but the encoders vary in which features they distribute to which tokens, how nuanced their distributions are, and how robust the encoding of each feature is to distance. We discuss implications of these results for how different types of models break down and prioritize word-level context information when constructing token embeddings.

We would like to thank Itamar Francez and Sam Wiseman for helpful discussion, and anonymous reviewers for their valuable feedback. This material is based upon work supported by the National Science Foundation under Award No. 1941160.

2020

huang-etal-2022-distilling

https://aclanthology.org/2022.fever-1.3.pdf

Distilling Salient Reviews with Zero Labels

Many people read online reviews to learn about real-world entities of their interest. However, majority of reviews only describes general experiences and opinions of the customers, and may not reveal facts that are specific to the entity being reviewed. In this work, we focus on a novel task of mining from a review corpus sentences that are unique for each entity. We refer to this task as Salient Fact Extraction. Salient facts are extremely scarce due to their very nature. Consequently, collecting labeled examples for training supervised models is tedious and cost-prohibitive. To alleviate this scarcity problem, we develop an unsupervised method ZL-Distiller, which leverages contextual language representations of the reviews and their distributional patterns to identify salient sentences about entities. Our experiments on multiple domains (hotels, products, and restaurants) show that ZL-Distiller achieves state-of-theart performance and further boosts the performance of other supervised/unsupervised algorithms for the task. Furthermore, we show that salient sentences mined by ZL-Distiller provide unique and detailed information about entities, which benefit downstream NLP applications including question answering and summarization. * Work done during internship at Megagon Labs. † Work done while at Megagon Labs. The Fifth Workshop on Fact Extraction and VERification (FEVER). Co-located with Association for Computational Linguistics 2022.

2022

lowe-etal-1994-language

https://aclanthology.org/H94-1087.pdf

Language Identification via Large Vocabulary Speaker Independent Continuous Speech Recognition

The goal of this study is to evaluate the potential for using large vocabulary continuous speech recognition as an engine for automatically classifying utterances according to the language being spoken. The problem of language identification is often thought of as being separate from the problem of speech recognition. But in this paper, as in Dragon's earlier work on topic and speaker identification, we explore a unifying approach to all three message classification problems based on the underlying stochastic process which gives rise to speech. We discuss the theoretical framework upon which our message classification systems are built and report on a series of experiments in which this theory is tested, using large vocabulary continuous speech recognition to distinguish English from Spanish.

1994

carroll-etal-2000-engineering

https://aclanthology.org/W00-2007.pdf

Engineering a Wide-Coverage Lexicalized Grammar

We discuss a number of practical issues that have arisen in the development of a wide-coverage lexicalized grammar for English. In particular, we consider the way in which the design of the •~rammar and of its encoding was infiuenced by issues relating to the size of the grammar.

2000

penkale-2013-tailor

https://aclanthology.org/2013.tc-1.13.pdf

Tailor-made quality-controlled translation

Traditional 'one-size-fits-all' models are failing to meet businesses' requirements. To support the growing demand for cost-effective translation, fine-grained control of quality is required, enabling fit-for-purpose content to be delivered at predictable quality and cost levels. This paper argues for customisable levels of quality, detailing the variables which can be altered to achieve a certain level of quality, and showing how this model can be implemented within Lingo24's Coach translation platform.

2013

patrick-li-2009-cascade

https://aclanthology.org/U09-1014.pdf

A Cascade Approach to Extracting Medication Events

Information Extraction, from the electronic clinical record is a comparatively new topic for computational linguists. In order to utilize the records to improve the efficiency and quality of health care, the knowledge content should be automatically encoded; however this poses a number of challenges for Natural Language Processing (NLP). In this paper, we present a cascade approach to discover the medicationrelated information (MEDICATION, DOSAGE, MODE, FREQUENCY, DURATION, REASON, and CONTEXT) from narrative patient records. The prototype of this system was used to participate the i2b2 2009 medication extraction challenge. The results show better than 90% accuracy on 5 out of 7 entities used in the study.

Good Health and Well-Being

We would like to acknowledge the contribution of Stephen Crawshaw, Yefeng Wang and other members in the Health Information Technologies Research Laboratory.Deidentified clinical records used in this research were provided by the i2b2 National Center for Biomedical Computing funded by U54LM008748 and were originally prepared for the Shared Tasks for Challenges in NLP for Clinical Data organized by Dr. Ozlem Uzuner, i2b2 and SUNY.

2009

brook-weiss-etal-2021-qa

https://aclanthology.org/2021.emnlp-main.778.pdf

QA-Align: Representing Cross-Text Content Overlap by Aligning Question-Answer Propositions

Multi-text applications, such as multidocument summarization, are typically required to model redundancies across related texts. Current methods confronting consolidation struggle to fuse overlapping information. In order to explicitly represent content overlap, we propose to align predicate-argument relations across texts, providing a potential scaffold for information consolidation. We go beyond clustering coreferring mentions, and instead model overlap with respect to redundancy at a propositional level, rather than merely detecting shared referents. Our setting exploits QA-SRL, utilizing question-answer pairs to capture predicate-argument relations, facilitating laymen annotation of cross-text alignments. We employ crowd-workers for constructing a dataset of QA-based alignments, and present a baseline QA alignment model trained over our dataset. Analyses show that our new task is semantically challenging, capturing content overlap beyond lexical similarity and complements cross-document coreference with proposition-level links, offering potential use for downstream tasks.

We would like to thank the anonymous reviewers for their thorough and insightful comments. The work described herein was supported in part by grants from Intel Labs, Facebook, and the Israel Science Foundation grant 1951/17.

2021

jappinen-etal-1988-locally

https://aclanthology.org/C88-1056.pdf

Locally Governed Trees and Dependecncy Parsing

~ paper desc[J.~s the notion of ]pcall.y gove~:ned t~:ees as a n<x]el of sttuCtu17ally [estrfcted dependency st~:uctures of sentenc.eSo 2~n abstract umchine and its supporting softwa~:e to*. the building of local ly goqerned t~:ees is intr_(~iuced. The rest of the paper dis~:usse,q [.ew uuaM]iguous~ ~<-'].]-for,~d local]y governed i:~:ees can be parsed ill l.i[~ea~: tia~ ~~en cxertain ~'tructural ~x~nstr~int's a~e in fuzce o

1988

zhang-etal-2016-jate

https://aclanthology.org/L16-1359.pdf

JATE 2.0: Java Automatic Term Extraction with Apache Solr

Automatic Term Extraction (ATE) or Recognition (ATR) is a fundamental processing step preceding many complex knowledge engineering tasks. However, few methods have been implemented as public tools and in particular, available as open-source freeware. Further, little effort is made to develop an adaptable and scalable framework that enables customization, development, and comparison of algorithms under a uniform environment. This paper introduces JATE 2.0, a complete remake of the free Java Automatic Term Extraction Toolkit (Zhang et al., 2008) delivering new features including: (1) highly modular, adaptable and scalable ATE thanks to integration with Apache Solr, the open source free-text indexing and search platform; (2) an extended collection of state-of-the-art algorithms. We carry out experiments on two well-known benchmarking datasets and compare the algorithms along the dimensions of effectiveness (precision) and efficiency (speed and memory consumption). To the best of our knowledge, this is by far the only free ATE library offering a flexible architecture and the most comprehensive collection of algorithms.

2016

salaberri-etal-2015-brol

https://aclanthology.org/R15-1072.pdf

bRol: The Parser of Syntactic and Semantic Dependencies for Basque

This paper presents bRol, the first fully automatic system to be developed for the parsing of syntactic and semantic dependencies in Basque. The parser has been built according to the settings established for the CoNLL-2009 Shared Task (Hajič et al., 2009), therefore, bRol can be thought of as a standard parser with scores comparable to the ones reported in the shared task. A second-order graphbased MATE parser has been used as the syntactic dependency parser. The semantic model, on the other hand, uses the traditional four-stage SRL pipeline. The system has a labeled attachment score of 80.51%, a labeled semantic F 1 of 75.10, and a labeled macro F 1 of 77.80.

Haritz Salaberri holds a PhD grant from the University of the Basque Country. In addition, this work has been supported by the EXTRECM project (Grant No. TIN2013-46616-C2-1-R) and IXA Group, research group of type A (2010-2015)(IT34410).

2015

darwish-etal-2017-arabic

https://aclanthology.org/W17-1302.pdf

Arabic Diacritization: Stats, Rules, and Hacks

In this paper, we present a new and fast state-of-the-art Arabic diacritizer that guesses the diacritics of words and then their case endings. We employ a Viterbi decoder at word-level with back-off to stem, morphological patterns, and transliteration and sequence labeling based diacritization of named entities. For case endings, we use Support Vector Machine (SVM) based ranking coupled with morphological patterns and linguistic rules to properly guess case endings. We achieve a low word level diacritization error of 3.29% and 12.77% without and with case endings respectively on a new multi-genre free of copyright test set. We are making the diacritizer available for free for research purposes.

2017

hovy-2015-demographic

https://aclanthology.org/P15-1073.pdf

Demographic Factors Improve Classification Performance

Extra-linguistic factors influence language use, and are accounted for by speakers and listeners. Most natural language processing (NLP) tasks to date, however, treat language as uniform. This assumption can harm performance. We investigate the effect of including demographic information on performance in a variety of text-classification tasks. We find that by including age or gender information, we consistently and significantly improve performance over demographic-agnostic models. These results hold across three text-classification tasks in five languages.

Thanks toŽeljko Agić, David Bamman, Jacob Eisenstein, Stephan Gouws, Anders Johannsen, Barbara Plank, Anders Søgaard, and Svitlana Volkova for their invaluable feedback, as well as to the anonymous reviewers, whose comments helped improve the paper. The author was supported under ERC Starting Grant LOWLANDS No. 313695.

2015

acl-1987-association

https://aclanthology.org/P87-1000.pdf

25th Annual Meeting of the Association for Computational Linguistics

The Twenty-Fifth Annual Meeting of the Association for Computational Linguistics offers the membership a chance to acknowledge and benefit from the wide range of developments in computational linguistics in the past several years. The papers in the program reflect the growing interaction between computational linguists trained in information processing approaches and those trained in linguistic disciplines. At the same time the program's papers report on new developments in many areas of computational linguistics that are now the forefront of research in our field. Because of the exponentially increasing number of papers submitted this year, the committee chose to reduce the number of panels and invited talks in favor of accepting additional papers.

1987

klein-nabi-2021-attention-based

https://aclanthology.org/2021.findings-emnlp.208.pdf

Attention-based Contrastive Learning for Winograd Schemas

Self-supervised learning has recently attracted considerable attention in the NLP community for its ability to learn discriminative features using a contrastive objective (Qu et al., 2020; Klein and Nabi, 2020). This paper investigates whether contrastive learning can be extended to Transfomer attention to tackling the Winograd Schema Challenge. To this end, we propose a novel self-supervised framework, leveraging a contrastive loss directly at the level of self-attention. Experimental analysis of our attention-based models on multiple datasets demonstrates superior commonsense reasoning capabilities. The proposed approach outperforms all comparable unsupervised approaches while occasionally surpassing supervised ones. 1

2021

pardo-etal-2010-computational

https://aclanthology.org/W10-1601.pdf

Computational Linguistics in Brazil: An Overview

In this paper we give an overview of Computational Linguistics / Natural Language Processing in Brazil, describing the general research scenario, the main research groups, existing events and journals, and the perceived challenges, among other relevant information. We also identify opportunities for collaboration.

The authors are grateful to SBC, CEPLN, FAPESP, and CAPES for supporting this work and the realization of STIL 2009, where part of the data shown in this paper was presented.

2010

zhou-huang-2019-towards

https://aclanthology.org/W19-8661.pdf

Towards Generating Math Word Problems from Equations and Topics

A math word problem is a narrative with a specific topic that provides clues to the correct equation with numerical quantities and variables therein. In this paper, we focus on the task of generating math word problems. Previous works are mainly templatebased with pre-defined rules. We propose a novel neural network model to generate math word problems from the given equations and topics. First, we design a fusion mechanism to incorporate the information of both equations and topics. Second, an entity-enforced loss is introduced to ensure the relevance between the generated math problem and the equation. Automatic evaluation results show that the proposed model significantly outperforms the baseline models. In human evaluations, the math word problems generated by our model are rated as being more relevant (in terms of solvability of the given equations and relevance to topics) and natural (i.e., grammaticality, fluency) than the baseline models.

Quality Education

We would like to thank the annotators for their efforts in the evaluation process. Thanks to the anonymous reviewers for their helpful comments and suggestions.

2019

green-2018-proposed

https://aclanthology.org/W18-5213.pdf

Proposed Method for Annotation of Scientific Arguments in Terms of Semantic Relations and Argument Schemes

This paper presents a proposed method for annotation of scientific arguments in biological/biomedical journal articles. Semantic entities and relations are used to represent the propositional content of arguments in instances of argument schemes. We describe an experiment in which we encoded the arguments in a journal article to identify issues in this approach. Our catalogue of argument schemes and a copy of the annotated article are now publically available.

Industry, Innovation and Infrastructure

The analysis of the CRAFT article was done with the help of Michael Branon and Bishwa Giri, who were supported by a UNCG 2016 Summer Faculty Excellence Research Grant.

2018

sui-etal-2000-information

https://aclanthology.org/P00-1060.pdf

An Information-Theory-Based Feature Type Analysis for the Modeling of Statistical Parsing

The paper proposes an information-theorybased method for feature types analysis in probabilistic evaluation modelling for statistical parsing. The basic idea is that we use entropy and conditional entropy to measure whether a feature type grasps some of the information for syntactic structure prediction. Our experiment quantitatively analyzes several feature types' power for syntactic structure prediction and draws a series of interesting conclusions.

2000

takeshita-etal-2020-existing

https://aclanthology.org/2020.gebnlp-1.5.pdf

Can Existing Methods Debias Languages Other than English? First Attempt to Analyze and Mitigate Japanese Word Embeddings

It is known that word embeddings exhibit biases inherited from the corpus, and those biases reflect social stereotypes. Recently, many studies have been conducted to analyze and mitigate biases in word embeddings. Unsupervised Bias Enumeration (UBE) (Swinger et al., 2019) is one of approach to analyze biases for English, and Hard Debias (Bolukbasi et al., 2016) is the common technique to mitigate gender bias. These methods focused on English, or, in smaller extent, on Indo-European languages. However, it is not clear whether these methods can be generalized to other languages. In this paper, we apply these analyzing and mitigating methods, UBE and Hard Debias, to Japanese word embeddings. Additionally, we examine whether these methods can be used for Japanese. We experimentally show that UBE and Hard Debias cannot be sufficiently adapted to Japanese embeddings.

2020

golding-schabes-1996-combining

https://aclanthology.org/P96-1010.pdf

Combining Trigram-Based and Feature-Based Methods for Context-Sensitive Spelling Correction

This paper addresses the problem of correcting spelling errors that result in valid, though unintended words (such as peace and piece, or quiet and quite) and also the problem of correcting particular word usage errors (such as amount and number, or among and between). Such corrections require contextual information and are not handled by conventional spelling programs such as Unix spell. First, we introduce a method called Trigrams that uses part-of-speech trigrams to encode the context. This method uses a small number of parameters compared to previous methods based on word trigrams. However, it is effectively unable to distinguish among words that have the same part of speech. For this case, an alternative feature-based method called Bayes performs better; but Bayes is less effective than Trigrams when the distinction among words depends on syntactic constraints. A hybrid method called Tribayes is then introduced that combines the best of the previous two methods. The improvement in performance of Tribayes over its components is verified experimentally. Tribayes is also compared with the grammar checker in Microsoft Word, and is found to have substantially higher performance.

1996

shnarch-etal-2013-plis

https://aclanthology.org/P13-4017.pdf

PLIS: a Probabilistic Lexical Inference System

This paper presents PLIS, an open source Probabilistic Lexical Inference System which combines two functionalities: (i) a tool for integrating lexical inference knowledge from diverse resources, and (ii) a framework for scoring textual inferences based on the integrated knowledge. We provide PLIS with two probabilistic implementation of this framework. PLIS is available for download and developers of text processing applications can use it as an off-the-shelf component for injecting lexical knowledge into their applications. PLIS is easily configurable, components can be extended or replaced with user generated ones to enable system customization and further research. PLIS includes an online interactive viewer, which is a powerful tool for investigating lexical inference processes.

The authors thank Eden Erez for his help with the interactive viewer and Miquel Esplà Gomis for the bilingual dictionaries. This work was partially supported by the European Community's 7 th Framework Programme (FP7/2007-2013) under grant agreement no. 287923 (EXCITEMENT) and the Israel Science Foundation grant 880/12.

2013

sogaard-johannsen-2012-robust

https://aclanthology.org/C12-2114.pdf

Robust Learning in Random Subspaces: Equipping NLP for OOV Effects

Inspired by work on robust optimization we introduce a subspace method for learning linear classifiers for natural language processing that are robust to out-of-vocabulary effects. The method is applicable in live-stream settings where new instances may be sampled from different and possibly also previously unseen domains. In text classification and part-of-speech (POS) tagging, robust perceptrons and robust stochastic gradient descent (SGD) with hinge loss achieve average error reductions of up to 18% when evaluated on out-of-domain data.

2012