Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Utilizing Large Language Models for Event Deconstruction to Enhance Multimodal Aspect-Based Sentiment Analysis

Xiaoyong Huang, Heli Sun, Qunshu Gao, Wenjie Huang, Ruichen Cao

TL;DR

This paper innovatively introduces Large Language Models (LLMs) for event decomposition and proposes a reinforcement learning framework for Multimodal Aspect-based Sentiment Analysis (MABSA-RL) framework, reducing the complexity of analysis, introducing reinforcement learning to optimize model parameters.

Abstract

With the rapid development of the internet, the richness of User-Generated Contentcontinues to increase, making Multimodal Aspect-Based Sentiment Analysis (MABSA) a research hotspot. Existing studies have achieved certain results in MABSA, but they have not effectively addressed the analytical challenges in scenarios where multiple entities and sentiments coexist. This paper innovatively introduces Large Language Models (LLMs) for event decomposition and proposes a reinforcement learning framework for Multimodal Aspect-based Sentiment Analysis (MABSA-RL) framework. This framework decomposes the original text into a set of events using LLMs, reducing the complexity of analysis, introducing reinforcement learning to optimize model parameters. Experimental results show that MABSA-RL outperforms existing advanced methods on two benchmark datasets. This paper provides a new research perspective and method for multimodal aspect-level sentiment analysis.

Utilizing Large Language Models for Event Deconstruction to Enhance Multimodal Aspect-Based Sentiment Analysis

TL;DR

This paper innovatively introduces Large Language Models (LLMs) for event decomposition and proposes a reinforcement learning framework for Multimodal Aspect-based Sentiment Analysis (MABSA-RL) framework, reducing the complexity of analysis, introducing reinforcement learning to optimize model parameters.

Abstract

With the rapid development of the internet, the richness of User-Generated Contentcontinues to increase, making Multimodal Aspect-Based Sentiment Analysis (MABSA) a research hotspot. Existing studies have achieved certain results in MABSA, but they have not effectively addressed the analytical challenges in scenarios where multiple entities and sentiments coexist. This paper innovatively introduces Large Language Models (LLMs) for event decomposition and proposes a reinforcement learning framework for Multimodal Aspect-based Sentiment Analysis (MABSA-RL) framework. This framework decomposes the original text into a set of events using LLMs, reducing the complexity of analysis, introducing reinforcement learning to optimize model parameters. Experimental results show that MABSA-RL outperforms existing advanced methods on two benchmark datasets. This paper provides a new research perspective and method for multimodal aspect-level sentiment analysis.

Paper Structure

This paper contains 16 sections, 11 equations, 3 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related works
  3. MABSA
  4. Reinforcement Learning
  5. Methodology
  6. Task Formulation
  7. Text Decomposition Module
  8. Multimodal Aspect Prediction and Sentiment Analysis Agent
  9. Sequential Decision Enhancement Module
  10. Experiments
  11. Experimental Setup
  12. Comparative Models
  13. Experimental Results
  14. Ablation Studies
  15. Case Study
  16. ...and 1 more sections

Figures (3)

  • Figure 1: An example of Multimodal Aspect-Based Sentiment Analysis. It includes three aspects: Jarrod Smith, KokomoPost6,Terre Haute, along with their corresponding sentiments. In addition, we present a Sequence Event Set obtained through event decomposition by Qwen-Max-0428qwen1.5.
  • Figure 2: The overview of MABSA-RL.
  • Figure 3: Two examples of predictions made by UMT+TomBERT, VLP-MABSA, and our MABSA-RL.