DELL: Generating Reactions and Explanations for LLM-Based Misinformation Detection

TL;DR

This paper introduces DELL, a framework that leverages large language models to generate reactions and explanations, enhancing misinformation detection accuracy and interpretability across multiple datasets.

Contribution

DELL innovatively integrates LLM-generated reactions, explanations, and expert merging to improve misinformation detection performance and calibration.

Findings

DELL outperforms state-of-the-art baselines by up to 16.8% in macro f1-score.

Generated reactions and explanations significantly aid misinformation detection.

LLM-guided expert merging produces better-calibrated predictions.

Abstract

Large language models are limited by challenges in factuality and hallucinations to be directly employed off-the-shelf for judging the veracity of news articles, where factual accuracy is paramount. In this work, we propose DELL that identifies three key stages in misinformation detection where LLMs could be incorporated as part of the pipeline: 1) LLMs could \emph{generate news reactions} to represent diverse perspectives and simulate user-news interaction networks; 2) LLMs could \emph{generate explanations} for proxy tasks (e.g., sentiment, stance) to enrich the contexts of news articles and produce experts specializing in various aspects of news understanding; 3) LLMs could \emph{merge task-specific experts} and provide an overall prediction by incorporating the predictions and confidence scores of varying experts. Extensive experiments on seven datasets with three LLMs demonstrate that DELL outperforms state-of-the-art baselines by up to 16.8\% in macro f1-score. Further analysis reveals that the generated reactions and explanations are greatly helpful in misinformation detection, while our proposed LLM-guided expert merging helps produce better-calibrated predictions.