Dataset Card for "squadshifts"

Dataset Summary

SquadShifts consists of four new test sets for the Stanford Question Answering Dataset (SQuAD) from four different domains: Wikipedia articles, New York
Times articles, Reddit comments, and Amazon product reviews. Each dataset was generated using the same data generating pipeline, Amazon Mechanical Turk interface, and data cleaning code as the original SQuAD v1.1 dataset. The "new-wikipedia" dataset measures overfitting on the original SQuAD v1.1 dataset. The "new-york-times", "reddit", and "amazon" datasets measure robustness to natural distribution shifts. We encourage SQuAD model developers to also evaluate their methods on these new datasets!

Supported Tasks and Leaderboards

More Information Needed

Languages

More Information Needed

Dataset Structure

Data Instances

amazon

• Size of downloaded dataset files: 16.50 MB
• Size of the generated dataset: 9.44 MB
• Total amount of disk used: 25.94 MB

An example of 'test' looks as follows.

{
    "answers": {
        "answer_start": [25],
        "text": ["amazon"]
    },
    "context": "This is a paragraph from amazon.",
    "id": "090909",
    "question": "Where is this paragraph from?",
    "title": "amazon dummy data"
}

new_wiki

• Size of downloaded dataset files: 16.50 MB
• Size of the generated dataset: 7.86 MB
• Total amount of disk used: 24.37 MB

An example of 'test' looks as follows.

{
    "answers": {
        "answer_start": [25],
        "text": ["wikipedia"]
    },
    "context": "This is a paragraph from wikipedia.",
    "id": "090909",
    "question": "Where is this paragraph from?",
    "title": "new_wiki dummy data"
}

nyt

• Size of downloaded dataset files: 16.50 MB
• Size of the generated dataset: 10.79 MB
• Total amount of disk used: 27.29 MB

An example of 'test' looks as follows.

{
    "answers": {
        "answer_start": [25],
        "text": ["new york times"]
    },
    "context": "This is a paragraph from new york times.",
    "id": "090909",
    "question": "Where is this paragraph from?",
    "title": "nyt dummy data"
}

reddit

• Size of downloaded dataset files: 16.50 MB
• Size of the generated dataset: 9.47 MB
• Total amount of disk used: 25.97 MB

An example of 'test' looks as follows.

{
    "answers": {
        "answer_start": [25],
        "text": ["reddit"]
    },
    "context": "This is a paragraph from reddit.",
    "id": "090909",
    "question": "Where is this paragraph from?",
    "title": "reddit dummy data"
}

Data Fields

The data fields are the same among all splits.

amazon

• id: a string feature.
• title: a string feature.
• context: a string feature.
• question: a string feature.
• answers: a dictionary feature containing:
  • text: a string feature.
  • answer_start: a int32 feature.

new_wiki

• id: a string feature.
• title: a string feature.
• context: a string feature.
• question: a string feature.
• answers: a dictionary feature containing:
  • text: a string feature.
  • answer_start: a int32 feature.

nyt

• id: a string feature.
• title: a string feature.
• context: a string feature.
• question: a string feature.
• answers: a dictionary feature containing:
  • text: a string feature.
  • answer_start: a int32 feature.

reddit

• id: a string feature.
• title: a string feature.
• context: a string feature.
• question: a string feature.
• answers: a dictionary feature containing:
  • text: a string feature.
  • answer_start: a int32 feature.

Data Splits

name	test
amazon	9885
new_wiki	7938
nyt	10065
reddit	9803

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

All the datasets are distributed under the CC BY 4.0 license.

Citation Information

@InProceedings{pmlr-v119-miller20a,
  title = {The Effect of Natural Distribution Shift on Question Answering Models},
  author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},
  booktitle = {Proceedings of the 37th International Conference on Machine Learning},
  pages = {6905--6916},
  year = {2020},
  editor = {III, Hal Daumé and Singh, Aarti},
  volume = {119},
  series = {Proceedings of Machine Learning Research},
  month = {13--18 Jul},
  publisher = {PMLR},
  pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},
  url = {https://proceedings.mlr.press/v119/miller20a.html},
}

Contributions

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