Abstract
We present an easy and efficient method to extend existing sentence embedding models to new languages. This allows to create multilingual versions from previously monolingual models. The training is based on the idea that a translated sentence should be mapped to the same location in the vector space as the original sentence. We use the original (monolingual) model to generate sentence embeddings for the source language and then train a new system on translated sentences to mimic the original model. Compared to other methods for training multilingual sentence embeddings, this approach has several advantages: It is easy to extend existing models with relatively few samples to new languages, it is easier to ensure desired properties for the vector space, and the hardware requirements for training are lower. We demonstrate the effectiveness of our approach for 50+ languages from various language families. Code to extend sentence embeddings models to more than 400 languages is publicly available.- Anthology ID:
- 2020.emnlp-main.365
- Volume:
- Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
- Month:
- November
- Year:
- 2020
- Address:
- Online
- Editors:
- Bonnie Webber, Trevor Cohn, Yulan He, Yang Liu
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 4512–4525
- Language:
- URL:
- https://aclanthology.org/2020.emnlp-main.365
- DOI:
- 10.18653/v1/2020.emnlp-main.365
- Cite (ACL):
- Nils Reimers and Iryna Gurevych. 2020. Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 4512–4525, Online. Association for Computational Linguistics.
- Cite (Informal):
- Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation (Reimers & Gurevych, EMNLP 2020)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-1/2020.emnlp-main.365.pdf
- Code
- UKPLab/sentence-transformers + additional community code
- Data
- BUCC, KorNLI, KorSTS, MultiNLI, OpenSubtitles, SNLI, WikiMatrix