Dataset Viewer

The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

allenai/math_qa(複製)

データの内容

数学の文章題(選択式)のデータ。1 行は一問で、Problem(問題文)、options(選択肢)、correct(正解)、Rationale(解説)を持ち、annotated_formula と linear_formula に操作プログラムが入る。AQuA-RAT の問題・選択肢・解説・正解に、新しい表現言語で完全に規定された操作プログラムをアノテーションして作られた(論文は MathQA)。

データの概要

項目 値
行数 37,297(1 構成・3 分割の合計)
列数 7
容量 11.86 KB(ファイル 3 本)
ファイルの形式 py 1
言語 en
公開 public

構成と分割

構成(config) 分割(split)ごとの行数 列数
default test 2,985、train 29,837、validation 4,475 7

列(構成 default)

列 型
Problem string
Rationale string
options string
correct string
annotated_formula string
linear_formula string
category string

行数と列は元のカードの YAML(dataset_info)の記載。データセットビューアでは読めなかった。全データセットの一覧は HYUKI5/dataset-catalog にある。

複製について

allenai/math_qa をリビジョン c4f1cc784c04c4957b50c97858f23893b633eea6 の時点でそのまま複製したもの。中身は変えていない。判断モデル Laya(convaiinnovations/laya)の微調整に使う候補として、調査メモ(Laya 学習用データセット調査)に載せたデータを集めている。

出典

論文

アノテーションの方法

収集の根拠

More Information Needed

初期のデータ収集と正規化

More Information Needed

元の言語データの作成者は誰か?

More Information Needed

アノテーションの過程

More Information Needed

微調整に使うときの注意

  • 調査メモの台帳での区分: A(許諾型のライセンスだけで、疑義の注記が無い)。区分は記載の文言で機械的に付けたもので、法的な判定ではない。学習に入れる前に元のカードとライセンスの原文で確かめる。
  • Laya の入力は 2,048 token まで(ModernBERT-large / mmBERT)。長い本文は先頭から黙って切られるので、切れる行は事前に除くか、要約・切り出しをしてから入れる。
  • 分割は記事・文書単位で行い、同じ本文から作った別の問いを学習側と較正側に分けない。温度と閾値は学習に使っていない分割で別々に当てる。

調査メモの記述

メモ 節 ライセンス(記載) 確認 注記
jev_sources_tasksource_bekko.csv CSV(tasksource-jev-typed-decisions) apache-2.0 / license_use=commercial 確認済 license_use=commercial / [Jev 元データ source] source_annotation_converted(deterministic recast; no model r

元のカード(日本語訳)

以下は元のデータセットのカードの本文を日本語に訳したもの(機械翻訳)。訳に疑問があるときは、末尾の原文を確かめる。


MathQA のデータセットカード

目次

  • データセットの説明
    • データセットの概要
    • 対応タスクとリーダーボード
    • 言語
  • データセットの構造
    • データの実例
    • データのフィールド
    • データの分割
  • データセットの作成
    • 収集の根拠
    • 元データ
    • アノテーション
    • 個人情報と機微な情報
  • データ利用上の注意
    • データセットの社会的影響
    • バイアスについての検討
    • その他の既知の制限
  • 追加情報
    • データセットの管理者
    • ライセンス情報
    • 引用情報
    • 貢献

データセットの説明

データセットの概要

本研究では、大規模な数学文章題のデータセットを提示する。

このデータセットは、新しい表現言語を用いて、AQuA-RAT データセットに完全に規定された操作プログラムをアノテーションすることで収集された。

AQuA-RAT は、問題、選択肢、根拠、正解の選択肢を提供している。

対応タスクとリーダーボード

More Information Needed

言語

More Information Needed

データセットの構造

データの実例

default

  • ダウンロードしたデータセットファイルのサイズ: 7.30 MB
  • 生成されたデータセットのサイズ: 22.96 MB
  • 使用ディスク容量の合計: 30.27 MB

'train' の例は次のとおりである。

{
    "Problem": "a multiple choice test consists of 4 questions , and each question has 5 answer choices . in how many r ways can the test be completed if every question is unanswered ?",
    "Rationale": "\"5 choices for each of the 4 questions , thus total r of 5 * 5 * 5 * 5 = 5 ^ 4 = 625 ways to answer all of them . answer : c .\"",
    "annotated_formula": "power(5, 4)",
    "category": "general",
    "correct": "c",
    "linear_formula": "power(n1,n0)|",
    "options": "a ) 24 , b ) 120 , c ) 625 , d ) 720 , e ) 1024"
}

データのフィールド

データのフィールドは、すべての分割で同じである。

default

  • Problem: string 型の特徴量。
  • Rationale: string 型の特徴量。
  • options: string 型の特徴量。
  • correct: string 型の特徴量。
  • annotated_formula: string 型の特徴量。
  • linear_formula: string 型の特徴量。
  • category: string 型の特徴量。

データの分割

name train validation test
default 29837 4475 2985

データセットの作成

収集の根拠

More Information Needed

元データ

初期のデータ収集と正規化

More Information Needed

元の言語データの作成者は誰か?

More Information Needed

アノテーション

アノテーションの過程

More Information Needed

アノテーターは誰か?

More Information Needed

個人情報と機微な情報

More Information Needed

データ利用上の注意

データセットの社会的影響

More Information Needed

バイアスについての検討

More Information Needed

その他の既知の制限

More Information Needed

追加情報

データセットの管理者

More Information Needed

ライセンス情報

このデータセットは Apache License, Version 2.0 のもとで提供されている。

引用情報

@inproceedings{amini-etal-2019-mathqa,
    title = "{M}ath{QA}: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms",
    author = "Amini, Aida  and
      Gabriel, Saadia  and
      Lin, Shanchuan  and
      Koncel-Kedziorski, Rik  and
      Choi, Yejin  and
      Hajishirzi, Hannaneh",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1245",
    doi = "10.18653/v1/N19-1245",
    pages = "2357--2367",
}

貢献

このデータセットを追加してくれた @thomwolf、@lewtun、@patrickvonplaten に感謝する。

原文(英語)

Dataset Card for MathQA

Dataset Summary

We introduce a large-scale dataset of math word problems.

Our dataset is gathered by using a new representation language to annotate over the AQuA-RAT dataset with fully-specified operational programs.

AQuA-RAT has provided the questions, options, rationale, and the correct options.

Supported Tasks and Leaderboards

More Information Needed

Languages

More Information Needed

Dataset Structure

Data Instances

default

  • Size of downloaded dataset files: 7.30 MB
  • Size of the generated dataset: 22.96 MB
  • Total amount of disk used: 30.27 MB

An example of 'train' looks as follows.

{
    "Problem": "a multiple choice test consists of 4 questions , and each question has 5 answer choices . in how many r ways can the test be completed if every question is unanswered ?",
    "Rationale": "\"5 choices for each of the 4 questions , thus total r of 5 * 5 * 5 * 5 = 5 ^ 4 = 625 ways to answer all of them . answer : c .\"",
    "annotated_formula": "power(5, 4)",
    "category": "general",
    "correct": "c",
    "linear_formula": "power(n1,n0)|",
    "options": "a ) 24 , b ) 120 , c ) 625 , d ) 720 , e ) 1024"
}

Data Fields

The data fields are the same among all splits.

default

  • Problem: a string feature.
  • Rationale: a string feature.
  • options: a string feature.
  • correct: a string feature.
  • annotated_formula: a string feature.
  • linear_formula: a string feature.
  • category: a string feature.

Data Splits

name train validation test
default 29837 4475 2985

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 dataset is licensed under the Apache License, Version 2.0.

Citation Information

@inproceedings{amini-etal-2019-mathqa,
    title = "{M}ath{QA}: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms",
    author = "Amini, Aida  and
      Gabriel, Saadia  and
      Lin, Shanchuan  and
      Koncel-Kedziorski, Rik  and
      Choi, Yejin  and
      Hajishirzi, Hannaneh",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1245",
    doi = "10.18653/v1/N19-1245",
    pages = "2357--2367",
}

Contributions

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

Downloads last month
78