English Contrastive Learning Can Learn Universal Cross-lingual Sentence Embeddings

Yaushian Wang, Ashley Wu, Graham Neubig


Abstract
Universal cross-lingual sentence embeddings map semantically similar cross-lingual sentences into a shared embedding space. Aligning cross-lingual sentence embeddings usually requires supervised cross-lingual parallel sentences. In this work, we propose mSimCSE, which extends SimCSE to multilingual settings and reveal that contrastive learning on English data can surprisingly learn high-quality universal cross-lingual sentence embeddings without any parallel data. In unsupervised and weakly supervised settings, mSimCSE significantly improves previous sentence embedding methods on cross-lingual retrieval and multilingual STS tasks. The performance of unsupervised mSimCSE is comparable to fully supervised methods in retrieving low-resource languages and multilingual STS.The performance can be further enhanced when cross-lingual NLI data is available.
Anthology ID:
2022.emnlp-main.621
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
9122–9133
Language:
URL:
https://aclanthology.org/2022.emnlp-main.621
DOI:
10.18653/v1/2022.emnlp-main.621
Bibkey:
Cite (ACL):
Yaushian Wang, Ashley Wu, and Graham Neubig. 2022. English Contrastive Learning Can Learn Universal Cross-lingual Sentence Embeddings. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 9122–9133, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
English Contrastive Learning Can Learn Universal Cross-lingual Sentence Embeddings (Wang et al., EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.621.pdf