Fine-tuning multilingual language models in Twitter/X sentiment analysis: a study on Eastern-European V4 languages

TL;DR

This study evaluates fine-tuning multilingual language models for Twitter/X sentiment analysis in Eastern-European V4 languages, demonstrating that small models can outperform larger ones in narrow tasks with limited data.

Contribution

It provides a comparative analysis of various LLMs fine-tuned for sentiment classification in underrepresented languages on Twitter/X data, highlighting effective configurations and the potential of small models.

Findings

Some models outperform larger ones in narrow sentiment tasks

Small training sets can achieve state-of-the-art performance

Optimal settings vary across models and tasks

Abstract

The aspect-based sentiment analysis (ABSA) is a standard NLP task with numerous approaches and benchmarks, where large language models (LLM) represent the current state-of-the-art. We focus on ABSA subtasks based on Twitter/X data in underrepresented languages. On such narrow tasks, small tuned language models can often outperform universal large ones, providing available and cheap solutions. We fine-tune several LLMs (BERT, BERTweet, Llama2, Llama3, Mistral) for classification of sentiment towards Russia and Ukraine in the context of the ongoing military conflict. The training/testing dataset was obtained from the academic API from Twitter/X during 2023, narrowed to the languages of the V4 countries (Czech Republic, Slovakia, Poland, Hungary). Then we measure their performance under a variety of settings including translations, sentiment targets, in-context learning and more, using GPT4 as a reference model. We document several interesting phenomena demonstrating, among others, that some models are much better fine-tunable on multilingual Twitter tasks than others, and that they can reach the SOTA level with a very small training set. Finally we identify combinations of settings providing the best results.