rocstories-combined / README.md
shawon's picture
Update README.md
9598341 verified
metadata
dataset_info:
  features:
    - name: title
      dtype: string
    - name: sentences
      sequence: string
    - name: shuffled_sentences
      sequence: string
    - name: gold_order
      sequence: int64
  splits:
    - name: train
      num_bytes: 54656181
      num_examples: 98161
  download_size: 32722430
  dataset_size: 54656181
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Dataset Card for Dataset Name

This dataset is a merged version of the Spring 2016 and Winter 2017 versions of the ROCStories Dataset. You can request the dataset from using the form on the website as well.

Dataset Details

Dataset Description

  • Curated by: Nasrin Mostafazadeh, Nathanael Chambers, Xiaodong He, Devi Parikh, Dhruv Batra, Lucy Vanderwende, Pushmeet Kohli, James Allen
  • Language(s) (NLP): English

Dataset Sources

Uses

Sentences Ordering, Sentence Comprehension, Evaluation of Language Models on Sentence Ordering.

Dataset Structure

The dataset contains 98161 stories, each containing 5 sentences. Each instance or story includes a title, the original and the shuffled order of the sentences and a list of integers as gold order for evaluating prediction from a model.

Citation

BibTeX:

@misc{mostafazadeh2016corpus,
      title={A Corpus and Evaluation Framework for Deeper Understanding of Commonsense Stories}, 
      author={Nasrin Mostafazadeh and Nathanael Chambers and Xiaodong He and Devi Parikh and Dhruv Batra and Lucy Vanderwende and Pushmeet Kohli and James Allen},
      year={2016},
      eprint={1604.01696},
      archivePrefix={arXiv},
      primaryClass={id='cs.CL' full_name='Computation and Language' is_active=True alt_name='cmp-lg' in_archive='cs' is_general=False description='Covers natural language processing. Roughly includes material in ACM Subject Class I.2.7. Note that work on artificial languages (programming languages, logics, formal systems) that does not explicitly address natural-language issues broadly construed (natural-language processing, computational linguistics, speech, text retrieval, etc.) is not appropriate for this area.'}
}

Dataset Card Authors

Shawon Ashraf