LexC-Gen: Generating Data for Extremely Low-Resource Languages with Large Language Models and Bilingual Lexicons

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yong, Zheng-Xin, Menghini, Cristina, Bach, Stephen H.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909367216898048
author Yong, Zheng-Xin
Menghini, Cristina
Bach, Stephen H.
author_facet Yong, Zheng-Xin
Menghini, Cristina
Bach, Stephen H.
contents Data scarcity in low-resource languages can be addressed with word-to-word translations from labeled task data in high-resource languages using bilingual lexicons. However, bilingual lexicons often have limited lexical overlap with task data, which results in poor translation coverage and lexicon utilization. We propose lexicon-conditioned data generation LexC-Gen, a method that generates low-resource-language classification task data at scale. Specifically, LexC-Gen first uses high-resource-language words from bilingual lexicons to generate lexicon-compatible task data, and then it translates them into low-resource languages with bilingual lexicons via word translation. Across 17 extremely low-resource languages, LexC-Gen generated data is competitive with expert-translated gold data, and yields on average 5.6 and 8.9 points improvement over existing lexicon-based word translation methods on sentiment analysis and topic classification tasks respectively. Through ablation study, we show that conditioning on bilingual lexicons is the key component of LexC-Gen. LexC-Gen serves as a potential solution to close the performance gap between open-source multilingual models, such as BLOOMZ and Aya-101, and state-of-the-art commercial models like GPT-4o on low-resource-language tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2402_14086
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LexC-Gen: Generating Data for Extremely Low-Resource Languages with Large Language Models and Bilingual Lexicons
Yong, Zheng-Xin
Menghini, Cristina
Bach, Stephen H.
Computation and Language
Artificial Intelligence
Machine Learning
Data scarcity in low-resource languages can be addressed with word-to-word translations from labeled task data in high-resource languages using bilingual lexicons. However, bilingual lexicons often have limited lexical overlap with task data, which results in poor translation coverage and lexicon utilization. We propose lexicon-conditioned data generation LexC-Gen, a method that generates low-resource-language classification task data at scale. Specifically, LexC-Gen first uses high-resource-language words from bilingual lexicons to generate lexicon-compatible task data, and then it translates them into low-resource languages with bilingual lexicons via word translation. Across 17 extremely low-resource languages, LexC-Gen generated data is competitive with expert-translated gold data, and yields on average 5.6 and 8.9 points improvement over existing lexicon-based word translation methods on sentiment analysis and topic classification tasks respectively. Through ablation study, we show that conditioning on bilingual lexicons is the key component of LexC-Gen. LexC-Gen serves as a potential solution to close the performance gap between open-source multilingual models, such as BLOOMZ and Aya-101, and state-of-the-art commercial models like GPT-4o on low-resource-language tasks.
title LexC-Gen: Generating Data for Extremely Low-Resource Languages with Large Language Models and Bilingual Lexicons
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2402.14086