The viewer is disabled because this dataset repo requires arbitrary Python code execution. Please consider removing the loading script and relying on automated data support (you can use convert_to_parquet from the datasets library). If this is not possible, please open a discussion for direct help.

Dataset Card for "wiki_split"

Dataset Summary

One million English sentences, each split into two sentences that together preserve the original meaning, extracted from Wikipedia Google's WikiSplit dataset was constructed automatically from the publicly available Wikipedia revision history. Although the dataset contains some inherent noise, it can serve as valuable training data for models that split or merge sentences.

Supported Tasks and Leaderboards

  • Split and Rephrase

Languages

More Information Needed

Dataset Structure

Data Instances

default

  • Size of downloaded dataset files: 100.28 MB
  • Size of the generated dataset: 388.40 MB
  • Total amount of disk used: 488.68 MB

An example of 'train' looks as follows.

{
    "complex_sentence": " '' As she translates from one language to another , she tries to find the appropriate wording and context in English that would correspond to the work in Spanish her poems and stories started to have differing meanings in their respective languages .",
    "simple_sentence_1": "' '' As she translates from one language to another , she tries to find the appropriate wording and context in English that would correspond to the work in Spanish . ",
    "simple_sentence_2": " Ergo , her poems and stories started to have differing meanings in their respective languages ."
}

Data Fields

The data fields are the same among all splits.

default

  • complex_sentence: a string feature.
  • simple_sentence_1: a string feature.
  • simple_sentence_2: a string feature.

Data Splits

name train validation test
default 989944 5000 5000

Dataset Creation

Curation Rationale

More Information Needed

Source Data

Initial Data Collection and Normalization

More Information Needed

Who are the source language producers?

More Information Needed

Annotations

Annotation process

More Information Needed

Who are the annotators?

More Information Needed

Personal and Sensitive Information

More Information Needed

Considerations for Using the Data

Social Impact of Dataset

More Information Needed

Discussion of Biases

More Information Needed

Other Known Limitations

More Information Needed

Additional Information

Dataset Curators

More Information Needed

Licensing Information

The WikiSplit dataset is a verbatim copy of certain content from the publicly available Wikipedia revision history. The dataset is therefore licensed under CC BY-SA 4.0. Any third party content or data is provided "As Is" without any warranty, express or implied.

Citation Information

@inproceedings{botha-etal-2018-learning,
    title = "Learning To Split and Rephrase From {W}ikipedia Edit History",
    author = "Botha, Jan A.  and
      Faruqui, Manaal  and
      Alex, John  and
      Baldridge, Jason  and
      Das, Dipanjan",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D18-1080",
    doi = "10.18653/v1/D18-1080",
    pages = "732--737",
}

Contributions

Thanks to @thomwolf, @patrickvonplaten, @albertvillanova, @lewtun for adding this dataset.

Downloads last month
38

Models trained or fine-tuned on google-research-datasets/wiki_split