Abstract
Successful methods for unsupervised neural machine translation (UNMT) employ cross-lingual pretraining via self-supervision, often in the form of a masked language modeling or a sequence generation task, which requires the model to align the lexical- and high-level representations of the two languages. While cross-lingual pretraining works for similar languages with abundant corpora, it performs poorly in low-resource and distant languages. Previous research has shown that this is because the representations are not sufficiently aligned. In this paper, we enhance the bilingual masked language model pretraining with lexical-level information by using type-level cross-lingual subword embeddings. Empirical results demonstrate improved performance both on UNMT (up to 4.5 BLEU) and bilingual lexicon induction using our method compared to a UNMT baseline.- Anthology ID:
- 2021.naacl-main.16
- Volume:
- Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
- Month:
- June
- Year:
- 2021
- Address:
- Online
- Venue:
- NAACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 173–180
- Language:
- URL:
- https://aclanthology.org/2021.naacl-main.16
- DOI:
- 10.18653/v1/2021.naacl-main.16
- Cite (ACL):
- Alexandra Chronopoulou, Dario Stojanovski, and Alexander Fraser. 2021. Improving the Lexical Ability of Pretrained Language Models for Unsupervised Neural Machine Translation. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 173–180, Online. Association for Computational Linguistics.
- Cite (Informal):
- Improving the Lexical Ability of Pretrained Language Models for Unsupervised Neural Machine Translation (Chronopoulou et al., NAACL 2021)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/2021.naacl-main.16.pdf
- Code
- alexandra-chron/lexical_xlm_relm