Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PAMAS: Self-Adaptive Multi-Agent System with Perspective Aggregation for Misinformation Detection

Zongwei Wang, Min Gao, Junliang Yu, Tong Chen, Chenghua Lin

TL;DR

PAMAS tackles misinformation detection by mitigating information drowning through perspective aggregation in a hierarchical, self-adaptive multi-agent system. It introduces Auditors, Coordinators, and a Decision-Maker to hierarchically amplify anomaly cues from diverse feature subsets and synthesize robust judgments. The framework adds topology adaptation, targeted correction, and confidence-guided routing to improve efficiency and scalability, with empirical validation on three benchmark datasets showing superior accuracy and token efficiency. This work advances trustworthy, scalable misinformation detection and opens pathways to multimodal extensions and human-in-the-loop collaboration.

Abstract

Misinformation on social media poses a critical threat to information credibility, as its diverse and context-dependent nature complicates detection. Large language model-empowered multi-agent systems (MAS) present a promising paradigm that enables cooperative reasoning and collective intelligence to combat this threat. However, conventional MAS suffer from an information-drowning problem, where abundant truthful content overwhelms sparse and weak deceptive cues. With full input access, agents tend to focus on dominant patterns, and inter-agent communication further amplifies this bias. To tackle this issue, we propose PAMAS, a multi-agent framework with perspective aggregation, which employs hierarchical, perspective-aware aggregation to highlight anomaly cues and alleviate information drowning. PAMAS organizes agents into three roles: Auditors, Coordinators, and a Decision-Maker. Auditors capture anomaly cues from specialized feature subsets; Coordinators aggregate their perspectives to enhance coverage while maintaining diversity; and the Decision-Maker, equipped with evolving memory and full contextual access, synthesizes all subordinate insights to produce the final judgment. Furthermore, to improve efficiency in multi-agent collaboration, PAMAS incorporates self-adaptive mechanisms for dynamic topology optimization and routing-based inference, enhancing both efficiency and scalability. Extensive experiments on multiple benchmark datasets demonstrate that PAMAS achieves superior accuracy and efficiency, offering a scalable and trustworthy way for misinformation detection.

PAMAS: Self-Adaptive Multi-Agent System with Perspective Aggregation for Misinformation Detection

TL;DR

PAMAS tackles misinformation detection by mitigating information drowning through perspective aggregation in a hierarchical, self-adaptive multi-agent system. It introduces Auditors, Coordinators, and a Decision-Maker to hierarchically amplify anomaly cues from diverse feature subsets and synthesize robust judgments. The framework adds topology adaptation, targeted correction, and confidence-guided routing to improve efficiency and scalability, with empirical validation on three benchmark datasets showing superior accuracy and token efficiency. This work advances trustworthy, scalable misinformation detection and opens pathways to multimodal extensions and human-in-the-loop collaboration.

Abstract

Misinformation on social media poses a critical threat to information credibility, as its diverse and context-dependent nature complicates detection. Large language model-empowered multi-agent systems (MAS) present a promising paradigm that enables cooperative reasoning and collective intelligence to combat this threat. However, conventional MAS suffer from an information-drowning problem, where abundant truthful content overwhelms sparse and weak deceptive cues. With full input access, agents tend to focus on dominant patterns, and inter-agent communication further amplifies this bias. To tackle this issue, we propose PAMAS, a multi-agent framework with perspective aggregation, which employs hierarchical, perspective-aware aggregation to highlight anomaly cues and alleviate information drowning. PAMAS organizes agents into three roles: Auditors, Coordinators, and a Decision-Maker. Auditors capture anomaly cues from specialized feature subsets; Coordinators aggregate their perspectives to enhance coverage while maintaining diversity; and the Decision-Maker, equipped with evolving memory and full contextual access, synthesizes all subordinate insights to produce the final judgment. Furthermore, to improve efficiency in multi-agent collaboration, PAMAS incorporates self-adaptive mechanisms for dynamic topology optimization and routing-based inference, enhancing both efficiency and scalability. Extensive experiments on multiple benchmark datasets demonstrate that PAMAS achieves superior accuracy and efficiency, offering a scalable and trustworthy way for misinformation detection.
Paper Structure (33 sections, 21 equations, 4 figures, 3 tables, 1 algorithm)

This paper contains 33 sections, 21 equations, 4 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. PAMAS: Self-Adaptive Multi-Agent System with Perspective Aggregation
  4. Agent Roles
  5. Auditor
  6. Coordinator
  7. Decision-Maker
  8. Initialization of PAMAS
  9. Decision-Maker Self-Learning.
  10. Auditor Assignment.
  11. Relation Construction.
  12. The Optimization of PAMAS
  13. Forward Pass.
  14. Topology Adaptation.
  15. Backward Targeted Correction.
  16. ...and 18 more sections

Figures (4)

  • Figure 1: Comparison between the existing MAS design and our proposed PAMAS.
  • Figure 2: The framework of PAMAS, including the process of initialization, optimization, and inference.
  • Figure 3: Impact of topology adaptation, target correction, and confidence-guided routing on accuracy–efficiency trade-offs across three datasets.
  • Figure 4: Comparative results of various MAS in terms of accuracy and token consumption across three datasets.