Datasets:
lmqg
/

Modalities:
Text
Languages:
Korean
ArXiv:
Libraries:
Datasets
License:
qag_koquad / README.md
asahi417's picture
Update README.md
cc90041
metadata
license: cc-by-sa-4.0
pretty_name: SQuAD for question generation
language: ko
multilinguality: monolingual
size_categories: 1k<n<10K
source_datasets: lmqg/qg_koquad
task_categories:
  - text-generation
task_ids:
  - language-modeling
tags:
  - question-generation

Dataset Card for "lmqg/qag_koquad"

Dataset Description

Dataset Summary

This is the question & answer generation dataset based on the KOQuAD.

Supported Tasks and Leaderboards

  • question-answer-generation: The dataset is assumed to be used to train a model for question & answer generation. Success on this task is typically measured by achieving a high BLEU4/METEOR/ROUGE-L/BERTScore/MoverScore (see our paper for more in detail).

Languages

Korean (ko)

Dataset Structure

An example of 'train' looks as follows.

{
  "paragraph": ""3.13 ๋งŒ์„ธ์šด๋™" ์€ 1919๋…„ 3.13์ผ ์ „์ฃผ์—์„œ ์ผ์–ด๋‚œ ๋งŒ์„ธ์šด๋™์ด๋‹ค. ์ง€์—ญ ์ธ์‚ฌ๋“ค๊ณผ ํ•จ๊ป˜ ์‹ ํฅํ•™๊ต ํ•™์ƒ๋“ค์ด ์ฃผ๋„์ ์ธ ์—ญํ• ์„ ํ•˜๋ฉฐ, ๋งŒ์„ธ์šด๋™์„ ์ด๋Œ์—ˆ๋‹ค. ๋ฐ•ํƒœ๋ จ, ๊น€์‹ ๊ทน ๋“ฑ ์ „์ฃผ ์ง€๋„์ž๋“ค์€ ๊ตฐ์‚ฐ์—์„œ 4์ผ๊ณผ 5์ผ ๋…๋ฆฝ๋งŒ์„ธ ์‹œ์œ„๊ฐ€ ๊ฐํ–‰๋๋‹ค๋Š” ์†Œ์‹์— ๋“ฃ๊ณ  ์ค€๋น„ํ•˜๊ณ  ์žˆ์—ˆ๋‹ค. ์ฒœ๋„๊ต์™€ ๋ฐ•ํƒœ๋ จ ์‹ ๊ฐ„ํšŒ ์ด๋ฌด์ง‘์—์„œ ํ•„์š”ํ•œ ํƒœ๊ทน๊ธฐ๋ฅผ ์ธ์‡„ํ•˜๊ธฐ๋กœ ํ–ˆ์—ˆ๋‹ค. ์„œ์šธ์„ ๋น„๋กฏํ•œ ๋‹ค๋ฅธ ์ง€๋ฐฉ์—์„œ ์‹œ์œ„๊ฐ€ ๊ณ„์†๋˜์ž ์ผ๋ณธ๊ฒฝ์ฐฐ์€ ์‹ ํฅํ•™๊ต์™€ ๊ธฐ์ „ํ•™๊ต๋ฅผ ๋น„๋กฏํ•œ ์ „์ฃผ์‹œ๋‚ด ํ•™๊ต์— ๊ฐ•์ œ ๋ฐฉํ•™์กฐ์น˜๋ฅผ ์ทจํ–ˆ๋‹ค. ์ด์— ์ตœ์ข…์‚ผ ๋“ฑ ์‹ ํฅํ•™๊ต ํ•™์ƒ 5๋ช…์€ ๋ฐค์„ ์ด์šฉํ•ด ์‹ ํฅํ•™๊ต ์ง€ํ•˜์‹ค์—์„œ ํƒœ๊ทน๊ธฐ ๋“ฑ ์ธ์‡„๋ฌผ์„ ๋งŒ๋“ค์—ˆ๋‹ค. ์ค€๋น„๋ฅผ ๋งˆ์นœ ์ด๋“ค์€ 13์ผ ์žฅํ„ฐ๋กœ ๋ชจ์ด๊ธฐ ์‹œ์ž‘ํ–ˆ๊ณ , ์ฑ„์†Œ๊ฐ€๋งˆ๋‹ˆ๋กœ ์œ„์žฅํ•œ ํƒœ๊ทน๊ธฐ๋ฅผ ์žฅํ„ฐ๋กœ ์‹ค์–ด ๋‚˜๋ฅด๊ณ  ๊ฑฐ์‚ฌ ์ง์ „ ์‹œ์žฅ ์ž…๊ตฌ์ธ ์™„์‚ฐ๋™๊ณผ ์ „์ฃผ๊ต ๊ฑด๋„ˆํŽธ์—์„œ ๊ตฐ์ค‘๋“ค์—๊ฒŒ ์€๋ฐ€ํžˆ ๋ฐฐ๋ถ€ํ–ˆ๋‹ค. ๋‚ฎ 12์‹œ20๋ถ„๊ป˜ ์‹ ํฅํ•™๊ต์™€ ๊ธฐ์ „ํ•™๊ต ํ•™์ƒ ๋ฐ ์ฒœ๋„๊ต๋„ ๋“ฑ์€ ํƒœ๊ทน๊ธฐ๋ฅผ ๋“ค๊ณ  ๋งŒ์„ธ๋ฅผ ๋ถˆ๋ €๋‹ค. ๋‚จ๋ฌธ ๋ฐ– ์‹œ์žฅ, ์ œ2๋ณดํ†ตํ•™๊ต(ํ˜„ ์™„์‚ฐ์ดˆ๋“ฑํ•™๊ต)์—์„œ ๋ชจ์—ฌ ์ธ์‡„๋ฌผ์„ ๋ฟŒ๋ฆฌ๋ฉฐ ์‹œ๊ฐ€์ง€๋กœ ๊ตฌ๋ณด๋กœ ํ–‰์ง„ํ–ˆ๋‹ค. ์‹œ์œ„๋Š” ์˜คํ›„ 11์‹œ๊นŒ์ง€ ์„œ๋„ˆ์ฐจ๋ก€ ๊ณ„์†๋๋‹ค. ๋˜ ๋‹ค์Œ๋‚  ์˜คํ›„ 3์‹œ์—๋„ ๊ตฐ์ค‘์ด ๋ชจ์—ฌ ๋งŒ์„ธ๋ฅผ ๋ถˆ๋ €๋‹ค. ์ดํ›„ ๊ณ ํ˜•์ง„, ๋‚จ๊ถํ˜„, ๊น€๋ณ‘ํ•™, ๊น€์ ์‡ , ์ด๊ธฐ๊ณค, ๊น€๊ฒฝ์‹  ๋“ฑ ์‹ ํฅํ•™๊ต ํ•™์ƒ๋“ค์€ ์‹œ์œ„๋ฅผ ์ฃผ๋„ํ–ˆ๋‹ค๋Š” ํ˜์˜๋กœ ๋ชจ๋‘ ์‹คํ˜• 1๋…„์„ ์–ธ๋„ ๋ฐ›์•˜๋‹ค. ์ด์™ธ ์‹ ํฅํ•™๊ต ํ•™์ƒ 3๋ช…์€ ์ผ์ œ์˜ ๊ณ ๋ฌธ์— ์˜ฅ์‚ฌํ•œ ๊ฒƒ์œผ๋กœ ์•Œ๋ ค์กŒ๋‹ค. ๋˜ ์‹œ์œ„๋ฅผ ์ง€๋„ํ•œ ๊น€์ธ์ „ ๋ชฉ์‚ฌ๋Š” ์ดํ›„ ์ค‘๊ตญ ์ƒํ•ด๋กœ ๊ฑฐ์ฒ˜๋ฅผ ์˜ฎ๊ฒจ ์ž„์‹œ์ •๋ถ€์—์„œ ํ™œ๋™ํ–ˆ๋‹ค. ํ˜„์žฌ ์‹ ํฅํ•™๊ต ๊ต๋ฌธ ์˜†์— ๋งŒ์„ธ์šด๋™ ๊ธฐ๋…๋น„๊ฐ€ ์„ธ์›Œ์ ธ ์žˆ๋‹ค.",
  "questions": [ "๋งŒ์„ธ์šด๋™ ๊ธฐ๋…๋น„๊ฐ€ ์„ธ์›Œ์ ธ ์žˆ๋Š” ๊ณณ์€?", "์ผ๋ณธ๊ฒฝ์ฐฐ์˜ ๊ฐ•์ œ ๋ฐฉํ•™์กฐ์น˜์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ํ•™์ƒ๋“ค์€ ์‹ ํฅํ•™๊ต ์ง€ํ•˜์‹ค์— ๋ชจ์—ฌ์„œ ์–ด๋–ค ์ธ์‡„๋ฌผ์„ ๋งŒ๋“ค์—ˆ๋Š”๊ฐ€?", "์—ฌ๋Ÿฌ ์ง€๋ฐฉ์—์„œ ์‹œ์œ„๊ฐ€ ์ผ์–ด๋‚˜์ž ์ผ๋ณธ๊ฒฝ์ฐฐ์ด ์ „์ฃผ์‹œ๋‚ด ํ•™๊ต์— ๊ฐํ–‰ํ•œ ์กฐ์น˜๋Š” ๋ฌด์—‡์ธ๊ฐ€?", "์ง€์—ญ์ธ์‚ฌ๋“ค๊ณผ ์‹ ํฅ๊ณ ๋“ฑํ•™๊ต ํ•™์ƒ๋“ค์ด ์ฃผ๋„์ ์ธ ์—ญํ• ์„ ํ•œ 3.13 ๋งŒ์„ธ์šด๋™์ด ์ผ์–ด๋‚œ ํ•ด๋Š”?", "์‹ ํฅํ•™๊ต ํ•™์ƒ๋“ค์€ ์‹œ์œ„๋ฅผ ์ฃผ๋„ํ–ˆ๋‹ค๋Š” ํ˜์˜๋กœ ๋ชจ๋‘ ์‹คํ˜• ๋ช‡๋…„์„ ์–ธ๋„ ๋ฐ›์•˜๋Š”๊ฐ€?", "๋งŒ์„ธ์šด๋™์—์„œ ์ฃผ๋„์ ์ธ ์—ญํ• ์„ ํ•œ ์ด๋“ค์€?", "1919๋…„ 3.1 ์šด๋™์ด ์ผ์–ด๋‚œ ์ง€์—ญ์€ ์–ด๋””์ธ๊ฐ€?", "3.13 ๋งŒ์„ธ์šด๋™์ด ์ผ์–ด๋‚œ ๊ณณ์€?" ],
  "answers": [ "์‹ ํฅํ•™๊ต ๊ต๋ฌธ ์˜†", "ํƒœ๊ทน๊ธฐ", "๊ฐ•์ œ ๋ฐฉํ•™์กฐ์น˜", "1919๋…„", "1๋…„", "์‹ ํฅํ•™๊ต ํ•™์ƒ๋“ค", "์ „์ฃผ", "์ „์ฃผ" ],
  "questions_answers": "question: ๋งŒ์„ธ์šด๋™ ๊ธฐ๋…๋น„๊ฐ€ ์„ธ์›Œ์ ธ ์žˆ๋Š” ๊ณณ์€?, answer: ์‹ ํฅํ•™๊ต ๊ต๋ฌธ ์˜† | question: ์ผ๋ณธ๊ฒฝ์ฐฐ์˜ ๊ฐ•์ œ ๋ฐฉํ•™์กฐ์น˜์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ํ•™์ƒ๋“ค์€ ์‹ ํฅํ•™๊ต ์ง€ํ•˜์‹ค์— ๋ชจ์—ฌ์„œ ์–ด๋–ค ์ธ์‡„๋ฌผ์„ ๋งŒ๋“ค์—ˆ๋Š”๊ฐ€?, answer: ํƒœ๊ทน๊ธฐ | question: ์—ฌ๋Ÿฌ ์ง€๋ฐฉ์—์„œ ์‹œ์œ„๊ฐ€ ์ผ์–ด๋‚˜์ž ์ผ๋ณธ๊ฒฝ์ฐฐ์ด ์ „์ฃผ์‹œ๋‚ด ํ•™๊ต์— ๊ฐํ–‰ํ•œ ์กฐ์น˜๋Š” ๋ฌด์—‡์ธ๊ฐ€?, answer: ๊ฐ•์ œ ๋ฐฉํ•™์กฐ์น˜ | question: ์ง€์—ญ์ธ์‚ฌ๋“ค๊ณผ ์‹ ํฅ๊ณ ๋“ฑํ•™๊ต ํ•™์ƒ๋“ค์ด ์ฃผ๋„์ ์ธ ์—ญํ• ์„ ํ•œ 3.13 ๋งŒ์„ธ์šด๋™์ด ์ผ์–ด๋‚œ ํ•ด๋Š”?, answer: 1919๋…„ | question: ์‹ ํฅํ•™๊ต ํ•™์ƒ๋“ค์€ ์‹œ์œ„๋ฅผ ์ฃผ๋„ํ–ˆ๋‹ค๋Š” ํ˜์˜๋กœ ๋ชจ๋‘ ์‹คํ˜• ๋ช‡๋…„์„ ์–ธ๋„ ๋ฐ›์•˜๋Š”๊ฐ€?, answer: 1๋…„ | question: ๋งŒ์„ธ์šด๋™์—์„œ ์ฃผ๋„์ ์ธ ์—ญํ• ์„ ํ•œ ์ด๋“ค์€?, answer: ์‹ ํฅํ•™๊ต ํ•™์ƒ๋“ค | question: 1919๋…„ 3.1 ์šด๋™์ด ์ผ์–ด๋‚œ ์ง€์—ญ์€ ์–ด๋””์ธ๊ฐ€?, answer: ์ „์ฃผ | question: 3.13 ๋งŒ์„ธ์šด๋™์ด ์ผ์–ด๋‚œ ๊ณณ์€?, answer: ์ „์ฃผ"
  }

The data fields are the same among all splits.

  • questions: a list of string features.
  • answers: a list of string features.
  • paragraph: a string feature.
  • questions_answers: a string feature.

Data Splits

train validation test
9600 960 4442

Citation Information

@inproceedings{ushio-etal-2022-generative,
    title = "{G}enerative {L}anguage {M}odels for {P}aragraph-{L}evel {Q}uestion {G}eneration",
    author = "Ushio, Asahi  and
        Alva-Manchego, Fernando  and
        Camacho-Collados, Jose",
    booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
    month = dec,
    year = "2022",
    address = "Abu Dhabi, U.A.E.",
    publisher = "Association for Computational Linguistics",
}