Structured Prompting and LLM Ensembling for Multimodal Conversational Aspect-based Sentiment Analysis

TL;DR

This paper introduces a structured prompting and ensembling approach using large language models to improve multimodal conversational aspect-based sentiment analysis, effectively extracting detailed sentiment components and detecting sentiment shifts.

Contribution

It presents a novel structured prompting pipeline and ensemble strategy for large language models to enhance multimodal sentiment analysis in conversations.

Findings

Achieved 47.38% average score on sentiment sextuple extraction.

Achieved 74.12% exact match F1 on sentiment flip detection.

Demonstrated effectiveness of step-wise refinement and ensemble methods.

Abstract

Understanding sentiment in multimodal conversations is a complex yet crucial challenge toward building emotionally intelligent AI systems. The Multimodal Conversational Aspect-based Sentiment Analysis (MCABSA) Challenge invited participants to tackle two demanding subtasks: (1) extracting a comprehensive sentiment sextuple, including holder, target, aspect, opinion, sentiment, and rationale from multi-speaker dialogues, and (2) detecting sentiment flipping, which detects dynamic sentiment shifts and their underlying triggers. For Subtask-I, in the present paper, we designed a structured prompting pipeline that guided large language models (LLMs) to sequentially extract sentiment components with refined contextual understanding. For Subtask-II, we further leveraged the complementary strengths of three LLMs through ensembling to robustly identify sentiment transitions and their triggers. Our system achieved a 47.38% average score on Subtask-I and a 74.12% exact match F1 on Subtask-II, showing the effectiveness of step-wise refinement and ensemble strategies in rich, multimodal sentiment analysis tasks.