Structured Prompting and LLM Ensembling for Multimodal Conversational Aspect-based Sentiment Analysis
This work addresses the challenge of understanding complex sentiment dynamics in multimodal conversations, which is incremental as it applies existing LLM techniques to a specific domain task.
The paper tackled multimodal conversational aspect-based sentiment analysis by designing a structured prompting pipeline for extracting sentiment sextuples and an LLM ensembling approach for detecting sentiment flipping, achieving a 47.38% average score on the first subtask and a 74.12% exact match F1 on the second.
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.