Saved in:
Bibliographic Details
Main Authors: Koto, Fajri, Beck, Tilman, Talat, Zeerak, Gurevych, Iryna, Baldwin, Timothy
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.02113
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910317871628288
author Koto, Fajri
Beck, Tilman
Talat, Zeerak
Gurevych, Iryna
Baldwin, Timothy
author_facet Koto, Fajri
Beck, Tilman
Talat, Zeerak
Gurevych, Iryna
Baldwin, Timothy
contents Improving multilingual language models capabilities in low-resource languages is generally difficult due to the scarcity of large-scale data in those languages. In this paper, we relax the reliance on texts in low-resource languages by using multilingual lexicons in pretraining to enhance multilingual capabilities. Specifically, we focus on zero-shot sentiment analysis tasks across 34 languages, including 6 high/medium-resource languages, 25 low-resource languages, and 3 code-switching datasets. We demonstrate that pretraining using multilingual lexicons, without using any sentence-level sentiment data, achieves superior zero-shot performance compared to models fine-tuned on English sentiment datasets, and large language models like GPT--3.5, BLOOMZ, and XGLM. These findings are observable for unseen low-resource languages to code-mixed scenarios involving high-resource languages.
format Preprint
id arxiv_https___arxiv_org_abs_2402_02113
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Zero-shot Sentiment Analysis in Low-Resource Languages Using a Multilingual Sentiment Lexicon
Koto, Fajri
Beck, Tilman
Talat, Zeerak
Gurevych, Iryna
Baldwin, Timothy
Computation and Language
Improving multilingual language models capabilities in low-resource languages is generally difficult due to the scarcity of large-scale data in those languages. In this paper, we relax the reliance on texts in low-resource languages by using multilingual lexicons in pretraining to enhance multilingual capabilities. Specifically, we focus on zero-shot sentiment analysis tasks across 34 languages, including 6 high/medium-resource languages, 25 low-resource languages, and 3 code-switching datasets. We demonstrate that pretraining using multilingual lexicons, without using any sentence-level sentiment data, achieves superior zero-shot performance compared to models fine-tuned on English sentiment datasets, and large language models like GPT--3.5, BLOOMZ, and XGLM. These findings are observable for unseen low-resource languages to code-mixed scenarios involving high-resource languages.
title Zero-shot Sentiment Analysis in Low-Resource Languages Using a Multilingual Sentiment Lexicon
topic Computation and Language
url https://arxiv.org/abs/2402.02113