@inproceedings{downey-etal-2024-targeted,
title = "Targeted Multilingual Adaptation for Low-resource Language Families",
author = "Downey, C. M. and
Blevins, Terra and
Serai, Dhwani and
Parikh, Dwija and
Steinert-Threlkeld, Shane",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.918",
doi = "10.18653/v1/2024.findings-emnlp.918",
pages = "15647--15663",
abstract = "Massively multilingual models are known to have limited utility in any one language, and to perform particularly poorly on low-resource languages. By contrast, targeted multinguality has been shown to benefit low-resource languages. To test this approach more rigorously, we systematically study best practices for adapting a pre-trained model to a language family. Focusing on the Uralic family as a test case, we adapt XLM-R under various configurations to model 15 languages; we then evaluate the performance of each experimental setting on two downstream tasks and 11 evaluation languages. Our adapted models significantly outperform mono- and multilingual baselines. A regression analysis reveals that adapted vocabulary size is relatively unimportant for low-resource languages, and that low-resource languages can be aggressively up-sampled during training at little detriment to performance in high-resource languages. These results introduce new best practices for performing language adaptation in a targeted setting.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="downey-etal-2024-targeted">
<titleInfo>
<title>Targeted Multilingual Adaptation for Low-resource Language Families</title>
</titleInfo>
<name type="personal">
<namePart type="given">C</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Downey</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Terra</namePart>
<namePart type="family">Blevins</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dhwani</namePart>
<namePart type="family">Serai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dwija</namePart>
<namePart type="family">Parikh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shane</namePart>
<namePart type="family">Steinert-Threlkeld</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yaser</namePart>
<namePart type="family">Al-Onaizan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohit</namePart>
<namePart type="family">Bansal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Nung</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Massively multilingual models are known to have limited utility in any one language, and to perform particularly poorly on low-resource languages. By contrast, targeted multinguality has been shown to benefit low-resource languages. To test this approach more rigorously, we systematically study best practices for adapting a pre-trained model to a language family. Focusing on the Uralic family as a test case, we adapt XLM-R under various configurations to model 15 languages; we then evaluate the performance of each experimental setting on two downstream tasks and 11 evaluation languages. Our adapted models significantly outperform mono- and multilingual baselines. A regression analysis reveals that adapted vocabulary size is relatively unimportant for low-resource languages, and that low-resource languages can be aggressively up-sampled during training at little detriment to performance in high-resource languages. These results introduce new best practices for performing language adaptation in a targeted setting.</abstract>
<identifier type="citekey">downey-etal-2024-targeted</identifier>
<identifier type="doi">10.18653/v1/2024.findings-emnlp.918</identifier>
<location>
<url>https://aclanthology.org/2024.findings-emnlp.918</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>15647</start>
<end>15663</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Targeted Multilingual Adaptation for Low-resource Language Families
%A Downey, C. M.
%A Blevins, Terra
%A Serai, Dhwani
%A Parikh, Dwija
%A Steinert-Threlkeld, Shane
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Findings of the Association for Computational Linguistics: EMNLP 2024
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F downey-etal-2024-targeted
%X Massively multilingual models are known to have limited utility in any one language, and to perform particularly poorly on low-resource languages. By contrast, targeted multinguality has been shown to benefit low-resource languages. To test this approach more rigorously, we systematically study best practices for adapting a pre-trained model to a language family. Focusing on the Uralic family as a test case, we adapt XLM-R under various configurations to model 15 languages; we then evaluate the performance of each experimental setting on two downstream tasks and 11 evaluation languages. Our adapted models significantly outperform mono- and multilingual baselines. A regression analysis reveals that adapted vocabulary size is relatively unimportant for low-resource languages, and that low-resource languages can be aggressively up-sampled during training at little detriment to performance in high-resource languages. These results introduce new best practices for performing language adaptation in a targeted setting.
%R 10.18653/v1/2024.findings-emnlp.918
%U https://aclanthology.org/2024.findings-emnlp.918
%U https://doi.org/10.18653/v1/2024.findings-emnlp.918
%P 15647-15663
Markdown (Informal)
[Targeted Multilingual Adaptation for Low-resource Language Families](https://aclanthology.org/2024.findings-emnlp.918) (Downey et al., Findings 2024)
ACL