ParaShoot: A Hebrew Question Answering Dataset

Omri Keren; Omer Levy

ParaShoot: A Hebrew Question Answering Dataset

Omri Keren, Omer Levy

School of Computer Science

Tel Aviv University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

NLP research in Hebrew has largely focused on morphology and syntax, where rich annotated datasets in the spirit of Universal Dependencies are available. Semantic datasets, however, are in short supply, hindering crucial advances in the development of NLP technology in Hebrew. In this work, we present ParaShoot, the first question answering dataset in modern Hebrew. The dataset follows the format and crowdsourcing methodology of SQuAD, and contains approximately 3000 annotated examples, similar to other question-answering datasets in low-resource languages. We provide the first baseline results using recently-released BERT-style models for Hebrew, showing that there is significant room for improvement on this task.

Original language	English
Title of host publication	Proceedings of the 3rd Workshop on Machine Reading for Question Answering
Editors	Adam Fisch, Alon Talmor, Danqi Chen, Eunsol Choi, Minjoon Seo, Patrick Lewis, Robin Jia, Sewon Min
Place of Publication	Punta Cana, Dominican Republic
Publisher	Association for Computational Linguistics
Pages	106-112
Number of pages	7
ISBN (Electronic)	978-1-954085-95-4
State	Published - 1 Nov 2021
Event	3rd Workshop on Machine Reading for Question Answering - Hybrid (Online and Co-located with EMNLP 2021 in the Dominican Republic Duration: 10 Nov 2021 → 10 Nov 2021 Conference number: 3

Workshop

Workshop	3rd Workshop on Machine Reading for Question Answering
Abbreviated title	MRQA 2021
Period	10/11/21 → 10/11/21

Access to Document

https://aclanthology.org/2021.mrqa-1.11

Cite this

@inproceedings{3cb8705a5838484aa9b978520645ecd9,

title = "ParaShoot: A Hebrew Question Answering Dataset",

abstract = "NLP research in Hebrew has largely focused on morphology and syntax, where rich annotated datasets in the spirit of Universal Dependencies are available. Semantic datasets, however, are in short supply, hindering crucial advances in the development of NLP technology in Hebrew. In this work, we present ParaShoot, the first question answering dataset in modern Hebrew. The dataset follows the format and crowdsourcing methodology of SQuAD, and contains approximately 3000 annotated examples, similar to other question-answering datasets in low-resource languages. We provide the first baseline results using recently-released BERT-style models for Hebrew, showing that there is significant room for improvement on this task.",

author = "Omri Keren and Omer Levy",

year = "2021",

month = nov,

day = "1",

language = "אנגלית",

pages = "106--112",

editor = "Adam Fisch and Alon Talmor and Danqi Chen and Eunsol Choi and Minjoon Seo and Patrick Lewis and Robin Jia and Sewon Min",

booktitle = "Proceedings of the 3rd Workshop on Machine Reading for Question Answering",

publisher = "Association for Computational Linguistics",

note = "3rd Workshop on Machine Reading for Question Answering ; Conference date: 10-11-2021 Through 10-11-2021",

}

Keren, O & Levy, O 2021, ParaShoot: A Hebrew Question Answering Dataset. in A Fisch, A Talmor, D Chen, E Choi, M Seo, P Lewis, R Jia & S Min (eds), Proceedings of the 3rd Workshop on Machine Reading for Question Answering. Association for Computational Linguistics, Punta Cana, Dominican Republic, pp. 106-112, 3rd Workshop on Machine Reading for Question Answering, 10/11/21. <https://aclanthology.org/2021.mrqa-1.11>

ParaShoot: A Hebrew Question Answering Dataset. / Keren, Omri; Levy, Omer.
Proceedings of the 3rd Workshop on Machine Reading for Question Answering. ed. / Adam Fisch; Alon Talmor; Danqi Chen; Eunsol Choi; Minjoon Seo; Patrick Lewis; Robin Jia; Sewon Min. Punta Cana, Dominican Republic: Association for Computational Linguistics, 2021. p. 106-112.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - ParaShoot: A Hebrew Question Answering Dataset

AU - Keren, Omri

AU - Levy, Omer

N1 - Conference code: 3

PY - 2021/11/1

Y1 - 2021/11/1

N2 - NLP research in Hebrew has largely focused on morphology and syntax, where rich annotated datasets in the spirit of Universal Dependencies are available. Semantic datasets, however, are in short supply, hindering crucial advances in the development of NLP technology in Hebrew. In this work, we present ParaShoot, the first question answering dataset in modern Hebrew. The dataset follows the format and crowdsourcing methodology of SQuAD, and contains approximately 3000 annotated examples, similar to other question-answering datasets in low-resource languages. We provide the first baseline results using recently-released BERT-style models for Hebrew, showing that there is significant room for improvement on this task.

AB - NLP research in Hebrew has largely focused on morphology and syntax, where rich annotated datasets in the spirit of Universal Dependencies are available. Semantic datasets, however, are in short supply, hindering crucial advances in the development of NLP technology in Hebrew. In this work, we present ParaShoot, the first question answering dataset in modern Hebrew. The dataset follows the format and crowdsourcing methodology of SQuAD, and contains approximately 3000 annotated examples, similar to other question-answering datasets in low-resource languages. We provide the first baseline results using recently-released BERT-style models for Hebrew, showing that there is significant room for improvement on this task.

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

SP - 106

EP - 112

BT - Proceedings of the 3rd Workshop on Machine Reading for Question Answering

A2 - Fisch, Adam

A2 - Talmor, Alon

A2 - Chen, Danqi

A2 - Choi, Eunsol

A2 - Seo, Minjoon

A2 - Lewis, Patrick

A2 - Jia, Robin

A2 - Min, Sewon

PB - Association for Computational Linguistics

CY - Punta Cana, Dominican Republic

T2 - 3rd Workshop on Machine Reading for Question Answering

Y2 - 10 November 2021 through 10 November 2021

ER -

ParaShoot: A Hebrew Question Answering Dataset

Abstract

Workshop

Access to Document

Fingerprint

Cite this