Multi-Stage Framework with Refinement Based Point Set Registration for Unsupervised Bi-Lingual Word Alignment

Silviu Vlad Oprea, Sourav Dutta, Haytham Assem


Abstract
Cross-lingual alignment of word embeddings are important in knowledge transfer across languages, for improving machine translation and other multi-lingual applications. Current unsupervised approaches relying on learning structure-preserving transformations, using adversarial networks and refinement strategies, suffer from instability and convergence issues. This paper proposes BioSpere, a novel multi-stage framework for unsupervised mapping of bi-lingual word embeddings onto a shared vector space, by combining adversarial initialization, refinement procedure and point set registration. Experiments for parallel dictionary induction and word similarity demonstrate state-of-the-art unsupervised results for BioSpere on diverse languages – showcasing robustness against variable adversarial performance.
Anthology ID:
2022.coling-1.92
Volume:
Proceedings of the 29th International Conference on Computational Linguistics
Month:
October
Year:
2022
Address:
Gyeongju, Republic of Korea
Venue:
COLING
SIG:
Publisher:
International Committee on Computational Linguistics
Note:
Pages:
1089–1097
Language:
URL:
https://aclanthology.org/2022.coling-1.92
DOI:
Bibkey:
Cite (ACL):
Silviu Vlad Oprea, Sourav Dutta, and Haytham Assem. 2022. Multi-Stage Framework with Refinement Based Point Set Registration for Unsupervised Bi-Lingual Word Alignment. In Proceedings of the 29th International Conference on Computational Linguistics, pages 1089–1097, Gyeongju, Republic of Korea. International Committee on Computational Linguistics.
Cite (Informal):
Multi-Stage Framework with Refinement Based Point Set Registration for Unsupervised Bi-Lingual Word Alignment (Oprea et al., COLING 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.coling-1.92.pdf