Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

The Future of Combating Rumors? Retrieval, Discrimination, and Generation

Junhao Xu, Longdi Xian, Zening Liu, Mingliang Chen, Qiuyang Yin, Fenghua Song

TL;DR

This work proposes a comprehensive debunking process that not only detects rumors but also provides explanatory generated content to refute the authenticity of the information.

Abstract

Artificial Intelligence Generated Content (AIGC) technology development has facilitated the creation of rumors with misinformation, impacting societal, economic, and political ecosystems, challenging democracy. Current rumor detection efforts fall short by merely labeling potentially misinformation (classification task), inadequately addressing the issue, and it is unrealistic to have authoritative institutions debunk every piece of information on social media. Our proposed comprehensive debunking process not only detects rumors but also provides explanatory generated content to refute the authenticity of the information. The Expert-Citizen Collective Wisdom (ECCW) module we designed aensures high-precision assessment of the credibility of information and the retrieval module is responsible for retrieving relevant knowledge from a Real-time updated debunking database based on information keywords. By using prompt engineering techniques, we feed results and knowledge into a LLM (Large Language Model), achieving satisfactory discrimination and explanatory effects while eliminating the need for fine-tuning, saving computational costs, and contributing to debunking efforts.

The Future of Combating Rumors? Retrieval, Discrimination, and Generation

TL;DR

This work proposes a comprehensive debunking process that not only detects rumors but also provides explanatory generated content to refute the authenticity of the information.

Abstract

Artificial Intelligence Generated Content (AIGC) technology development has facilitated the creation of rumors with misinformation, impacting societal, economic, and political ecosystems, challenging democracy. Current rumor detection efforts fall short by merely labeling potentially misinformation (classification task), inadequately addressing the issue, and it is unrealistic to have authoritative institutions debunk every piece of information on social media. Our proposed comprehensive debunking process not only detects rumors but also provides explanatory generated content to refute the authenticity of the information. The Expert-Citizen Collective Wisdom (ECCW) module we designed aensures high-precision assessment of the credibility of information and the retrieval module is responsible for retrieving relevant knowledge from a Real-time updated debunking database based on information keywords. By using prompt engineering techniques, we feed results and knowledge into a LLM (Large Language Model), achieving satisfactory discrimination and explanatory effects while eliminating the need for fine-tuning, saving computational costs, and contributing to debunking efforts.
Paper Structure (16 sections, 15 equations, 5 figures, 4 tables)

This paper contains 16 sections, 15 equations, 5 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Method
  4. Semantic Encoding
  5. Domain Expert Discrimination
  6. Citizen Perceptual
  7. Collective Wisdom Decision
  8. Retrieval-Augmented Generation
  9. Experiments
  10. Dataset
  11. Baselines
  12. Implementation Details
  13. Performance Comparison
  14. Ablation study
  15. Parameter Sensitivity
  16. ...and 1 more sections

Figures (5)

  • Figure 1: Example of rumor. Both in the academic and industrial sectors, only the first half of the work has been done so far. However, for users influenced by information, this is not sufficient to convince them. Therefore, the second half of the work is what reality demands.
  • Figure 2: The comprehensive framework for debunking processes.In the bottom-left corner, it represents the user's inquiries about the authenticity of the rumors they need to verify. In the top-left corner lies the process of constructing a database of debunking knowledge vectors. The bottom-right corner depicts the simplified workflow of our rumor discrimination network. The top-right corner illustrates how to obtain interpretable debunking content by integrating the debunking results with the retrieved knowledge through RAG (Retrieval-Augmented Generation).
  • Figure 3: Expert Discrimination. The input will be routed through a router to allocate domain experts, and finally, the selected expert information will be summarized.
  • Figure 4: Citizen Perceptual. The input will first pass through semantic noise simulating different people's perspectives, then gain insights from information exchange with others' viewpoints, and finally integrate with one's initial perspective to form a revised viewpoint.
  • Figure 5: Parameter Sensitivity