Combining Static and Contextualised Multilingual Embeddings

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3A10457003" target="_blank" >RIV/00216208:11320/22:10457003 - isvavai.cz</a>

Result on the web

<a href="https://aclanthology.org/2022.findings-acl.182" target="_blank" >https://aclanthology.org/2022.findings-acl.182</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2022.findings-acl.182" target="_blank" >10.18653/v1/2022.findings-acl.182</a>

Result language

angličtina

Original language name

Combining Static and Contextualised Multilingual Embeddings

Original language description

Static and contextual multilingual embeddings have complementary strengths. Static embeddings, while less expressive than contextual language models, can be more straightforwardly aligned across multiple languages. We combine the strengths of static and contextual models to improve multilingual representations. We extract static embeddings for 40 languages from XLM-R, validate those embeddings with cross-lingual word retrieval, and then align them using VecMap. This results in high-quality, highly multilingual static embeddings. Then we apply a novel continued pre-training approach to XLM-R, leveraging the high quality alignment of our static embeddings to better align the representation space of XLM-R. We show positive results for multiple complex semantic tasks. We release the static embeddings and the continued pre-training code. Unlike most previous work, our continued pre-training approach does not require parallel text.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

—

Continuities

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year

2022

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

Findings of the Association for Computational Linguistics: ACL 2022

ISBN

978-1-955917-25-4

ISSN

—

e-ISSN

—

Number of pages

14

Pages from-to

2316-2329

Publisher name

Association for Computational Linguistics

Place of publication

Stroudsburg, PA, USA

Event location

Dublin, Ireland

Event date

May 22, 2022

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

000828767402030