Multi-CrossRE A Multi-Lingual Multi-Domain Dataset for Relation Extraction

Elisa Bassignana; Filip Ginter; Sampo Pyysalo; Rob van der Goot; Barbara Plank

Multi-CrossRE A Multi-Lingual Multi-Domain Dataset for Relation Extraction

Elisa Bassignana, Filip Ginter, Sampo Pyysalo, Rob van der Goot, Barbara Plank

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

Abstract

Most research in Relation Extraction (RE) involves the English language, mainly due to the lack of multi-lingual resources. We propose Multi-CrossRE, the broadest multi-lingual dataset for RE, including 26 languages in addition to English, and covering six text domains. Multi-CrossRE is a machine translated version of CrossRE (Bassignana and Plank, 2022), with a sub-portion including more than 200 sentences in seven diverse languages checked by native speakers. We run a baseline model over the 26 new datasets and–as sanity check–over the 26 back-translations to English. Results on the back-translated data are consistent with the ones on the original English CrossRE, indicating high quality of the translation and the resulting dataset.

Originalsprog	Engelsk
Titel	Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa
Forlag	University of Tartu Library
Publikationsdato	2023
Sider	80 - 85
Status	Udgivet - 2023

Emneord

Relation Extraction
Multi-lingual Dataset
Machine Translation
CrossRE
Text Domains

Adgang til dokumentet

https://aclanthology.org/2023.nodalida-1.9.pdf

Citationsformater

@inproceedings{c2cc1e0b3fa343a29daace2c44a58072,

title = "Multi-CrossRE A Multi-Lingual Multi-Domain Dataset for Relation Extraction",

abstract = "Most research in Relation Extraction (RE) involves the English language, mainly due to the lack of multi-lingual resources. We propose Multi-CrossRE, the broadest multi-lingual dataset for RE, including 26 languages in addition to English, and covering six text domains. Multi-CrossRE is a machine translated version of CrossRE (Bassignana and Plank, 2022), with a sub-portion including more than 200 sentences in seven diverse languages checked by native speakers. We run a baseline model over the 26 new datasets and–as sanity check–over the 26 back-translations to English. Results on the back-translated data are consistent with the ones on the original English CrossRE, indicating high quality of the translation and the resulting dataset.",

keywords = "Relation Extraction, Multi-lingual Dataset, Machine Translation, CrossRE, Text Domains, Relation Extraction, Multi-lingual Dataset, Machine Translation, CrossRE, Text Domains",

author = "Elisa Bassignana and Filip Ginter and Sampo Pyysalo and {van der Goot}, Rob and Barbara Plank",

year = "2023",

language = "English",

pages = "80 -- 85",

booktitle = "Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa",

publisher = "University of Tartu Library",

}

Multi-CrossRE A Multi-Lingual Multi-Domain Dataset for Relation Extraction. / Bassignana, Elisa; Ginter, Filip; Pyysalo, Sampo et al.
Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa. University of Tartu Library, 2023. s. 80 - 85.

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

TY - GEN

T1 - Multi-CrossRE A Multi-Lingual Multi-Domain Dataset for Relation Extraction

AU - Bassignana, Elisa

AU - Ginter, Filip

AU - Pyysalo, Sampo

AU - van der Goot, Rob

AU - Plank, Barbara

PY - 2023

Y1 - 2023

N2 - Most research in Relation Extraction (RE) involves the English language, mainly due to the lack of multi-lingual resources. We propose Multi-CrossRE, the broadest multi-lingual dataset for RE, including 26 languages in addition to English, and covering six text domains. Multi-CrossRE is a machine translated version of CrossRE (Bassignana and Plank, 2022), with a sub-portion including more than 200 sentences in seven diverse languages checked by native speakers. We run a baseline model over the 26 new datasets and–as sanity check–over the 26 back-translations to English. Results on the back-translated data are consistent with the ones on the original English CrossRE, indicating high quality of the translation and the resulting dataset.

AB - Most research in Relation Extraction (RE) involves the English language, mainly due to the lack of multi-lingual resources. We propose Multi-CrossRE, the broadest multi-lingual dataset for RE, including 26 languages in addition to English, and covering six text domains. Multi-CrossRE is a machine translated version of CrossRE (Bassignana and Plank, 2022), with a sub-portion including more than 200 sentences in seven diverse languages checked by native speakers. We run a baseline model over the 26 new datasets and–as sanity check–over the 26 back-translations to English. Results on the back-translated data are consistent with the ones on the original English CrossRE, indicating high quality of the translation and the resulting dataset.

KW - Relation Extraction

KW - Multi-lingual Dataset

KW - Machine Translation

KW - CrossRE

KW - Text Domains

KW - Relation Extraction

KW - Multi-lingual Dataset

KW - Machine Translation

KW - CrossRE

KW - Text Domains

M3 - Article in proceedings

SP - 80

EP - 85

BT - Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa

PB - University of Tartu Library

ER -

Multi-CrossRE A Multi-Lingual Multi-Domain Dataset for Relation Extraction

Abstract

Emneord

Adgang til dokumentet

Fingeraftryk

Citationsformater