Authors: Guanghua Li, Wensheng Lu, Wei Zhang, Defu Lian, Kezhong Lu, Rui Mao, Kai Shu, Hao Liao

Published on: March 14, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2403.09747

Summary

• What is new: Introduces a novel, retrieval-augmented LLMs framework for fake news detection, using a multi-round retrieval strategy from web sources.
• Why this is important: Existing fake news detection methods suffer from outdated or incomplete data and limitations in handling stale and long-tail knowledge.
• What the research proposes: A retrieval-augmented LLMs framework with a multi-round retrieval strategy for dynamic and relevant evidence gathering.
• Results: The framework outperforms existing methods in accuracy and provides human-readable explanations to improve interpretability.

Technical Details

Technological frameworks used: Retrieval-augmented Large Language Models (LLMs)

Models used: Multi-round retrieval strategy model

Data used: Three real-world datasets

Potential Impact

News organizations, social media platforms, and fact-checking agencies could benefit or be disrupted.

Want to implement this idea in a business?

We have generated a startup concept here: VeritasAI.