wiki_table_questions

• Description:

The dataset contains pairs table-question, and the respective answer. The questions require multi-step reasoning and various data operations such as comparison, aggregation, and arithmetic computation. The tables were randomly selected among Wikipedia tables with at least 8 rows and 5 columns.

(As per the documentation usage notes)

• Dev: Mean accuracy over three (not five) splits of the training data. In other words, train on 'split-{1,2,3}-train' and test on 'split-{1,2,3}-dev', respectively, then average the accuracy.

• Test: Train on 'train' and test on 'test'.

• Additional Documentation: Explore on Papers With Code north_east

• Homepage: https://ppasupat.github.io/WikiTableQuestions/#usage-notes

• Source code: tfds.structured.wiki_table_questions.WikiTableQuestions

• Versions:

  • 1.0.0 (default): Initial release.

• Download size: 65.36 MiB

• Dataset size: 237.24 MiB

• Auto-cached (documentation): Yes

• Splits:

• Feature structure:

FeaturesDict({
    'input_text': FeaturesDict({
        'context': string,
        'table': Sequence({
            'column_header': string,
            'content': string,
            'row_number': int16,
        }),
    }),
    'target_text': string,
})

• Feature documentation:

• Supervised keys (See as_supervised doc): ('input_text', 'target_text')

• Figure (tfds.show_examples): Not supported.

• Examples (tfds.as_dataframe):

• Citation:

@inproceedings{pasupat-liang-2015-compositional,
    title = "Compositional Semantic Parsing on Semi-Structured Tables",
    author = "Pasupat, Panupong  and
      Liang, Percy",
    booktitle = "Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
    month = jul,
    year = "2015",
    address = "Beijing, China",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/P15-1142",
    doi = "10.3115/v1/P15-1142",
    pages = "1470--1480",
}