Explainable Multimodal Aspect-Based Sentiment Analysis with Dependency-guided Large Language Model

TL;DR

This paper introduces a generative, explainable multimodal aspect-based sentiment analysis framework using large language models, which jointly predicts sentiments and generates natural language explanations, improving interpretability and accuracy.

Contribution

It reformulates MABSA as a generative task with a dependency-guided strategy, creating a unified, explainable approach that enhances reasoning and provides natural language explanations.

Findings

Achieves higher sentiment classification accuracy.

Produces faithful, aspect-grounded explanations.

Enhances interpretability with dependency syntax guidance.

Abstract

Multimodal aspect-based sentiment analysis (MABSA) aims to identify aspect-level sentiments by jointly modeling textual and visual information, which is essential for fine-grained opinion understanding in social media. Existing approaches mainly rely on discriminative classification with complex multimodal fusion, yet lacking explicit sentiment explainability. In this paper, we reformulate MABSA as a generative and explainable task, proposing a unified framework that simultaneously predicts aspect-level sentiment and generates natural language explanations. Based on multimodal large language models (MLLMs), our approach employs a prompt-based generative paradigm, jointly producing sentiment and explanation. To further enhance aspect-oriented reasoning capabilities, we propose a dependency-syntax-guided sentiment cue strategy. This strategy prunes and textualizes the aspect-centered dependency syntax tree, guiding the model to distinguish different sentiment aspects and enhancing its explainability. To enable explainability, we use MLLMs to construct new datasets with sentiment explanations to fine-tune. Experiments show that our approach not only achieves consistent gains in sentiment classification accuracy, but also produces faithful, aspect-grounded explanations.