Fact Checking
Fact-checking research aims to automate the verification of claims, combating the spread of misinformation across various media. Current efforts focus on improving evidence retrieval using techniques like contrastive learning and leveraging large language models (LLMs) for claim verification and explanation generation, often incorporating knowledge graphs and multimodal data (text and images). These advancements are crucial for enhancing the accuracy and efficiency of fact-checking, with implications for journalism, public health communication, and broader efforts to mitigate the impact of misinformation.
Papers
SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models
Potsawee Manakul, Adian Liusie, Mark J. F. Gales
Automated Query Generation for Evidence Collection from Web Search Engines
Nestor Prieto-Chavana, Julie Weeds, David Weir
FactReranker: Fact-guided Reranker for Faithful Radiology Report Summarization
Qianqian Xie, Jiayu Zhou, Yifan Peng, Fei Wang