Evaluating Zero-Shot Multilingual Aspect-Based Sentiment Analysis with Large Language Models

TL;DR

This paper evaluates the effectiveness of large language models in zero-shot multilingual aspect-based sentiment analysis, revealing their potential and limitations compared to fine-tuned models across various prompting strategies.

Contribution

It provides a comprehensive empirical assessment of LLMs for multilingual ABSA under zero-shot conditions, exploring multiple prompting techniques and highlighting areas for improvement.

Findings

LLMs show promise but lag behind fine-tuned models in ABSA.

Simpler zero-shot prompts often outperform complex strategies.

Performance varies across languages, with high-resource languages like English performing better.

Abstract

Aspect-based sentiment analysis (ABSA), a sequence labeling task, has attracted increasing attention in multilingual contexts. While previous research has focused largely on fine-tuning or training models specifically for ABSA, we evaluate large language models (LLMs) under zero-shot conditions to explore their potential to tackle this challenge with minimal task-specific adaptation. We conduct a comprehensive empirical evaluation of a series of LLMs on multilingual ABSA tasks, investigating various prompting strategies, including vanilla zero-shot, chain-of-thought (CoT), self-improvement, self-debate, and self-consistency, across nine different models. Results indicate that while LLMs show promise in handling multilingual ABSA, they generally fall short of fine-tuned, task-specific models. Notably, simpler zero-shot prompts often outperform more complex strategies, especially in high-resource languages like English. These findings underscore the need for further refinement of LLM-based approaches to effectively address ABSA task across diverse languages.