Inducing Language-Agnostic Multilingual Representations

Wei Zhao, Steffen Eger, Johannes Bjerva, Isabelle Augenstein

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

25 Citations (Scopus)

Abstract

Cross-lingual representations have the potential to make NLP techniques available to the vast majority of languages in the world. However, they currently require large pretraining corpora or access to typologically similar languages. In this work, we address these obstacles by removing language identity signals from multilingual embeddings. We examine three approaches for this: (i) re-aligning the vector spaces of target languages (all together) to a pivot source language; (ii) removing language-specific means and variances, which yields better discriminativeness of embeddings as a by-product; and (iii) increasing input similarity across languages by removing morphological contractions and sentence reordering. We evaluate on XNLI and reference-free MT across 19 typologically diverse languages. Our findings expose the limitations of these approaches-unlike vector normalization, vector space re-alignment and text normalization do not achieve consistent gains across encoders and languages. Due to the approaches' additive effects, their combination decreases the cross-lingual transfer gap by 8.9 points (m-BERT) and 18.2 points (XLM-R) on average across all tasks and languages, however. Our code and models are publicly available.

Original languageEnglish
Title of host publication*SEM 2021 - 10th Conference on Lexical and Computational Semantics, Proceedings of the Conference
EditorsLun-Wei Ku, Vivi Nastase, Ivan Vulic
Number of pages12
PublisherAssociation for Computational Linguistics, ACL Anthology
Publication date2021
Pages229-240
ISBN (Electronic)9781954085770
Publication statusPublished - 2021
Event10th Conference on Lexical and Computational Semantics, *SEM 2021 - Virtual, Bangkok, Thailand
Duration: 5 Aug 20216 Aug 2021

Conference

Conference10th Conference on Lexical and Computational Semantics, *SEM 2021
Country/TerritoryThailand
CityVirtual, Bangkok
Period05/08/202106/08/2021
SponsorACL Special Interest Group on the Lexicon, SIGLEX
Series*SEM 2021 - 10th Conference on Lexical and Computational Semantics, Proceedings of the Conference

Bibliographical note

Funding Information:
We thank the anonymous reviewers for their insightful comments and suggestions, which greatly improved the final version of the paper. This work has been supported by the German Research Foundation as part of the Research Training Group Adaptive Preparation of Information from Heterogeneous Sources (AIPHES) at the Technische Uni-versität Darmstadt under grant No. GRK 1994/1, as well as by the Swedish Research Council under grant agreement No 2019-04129.

Publisher Copyright:
© 2021 Lexical and Computational Semantics

Fingerprint

Dive into the research topics of 'Inducing Language-Agnostic Multilingual Representations'. Together they form a unique fingerprint.

Cite this