[2412.15189] Face the Facts! Evaluating RAG-based Fact-checking Pipelines in Realistic Settings

View a PDF of the paper titled Face the Facts! Evaluating RAG-based Fact-checking Pipelines in Realistic Settings, by Daniel Russo and 3 other authors

View PDF
HTML (experimental)

Abstract:Natural Language Processing and Generation systems have recently shown the potential to complement and streamline the costly and time-consuming job of professional fact-checkers. In this work, we lift several constraints of current state-of-the-art pipelines for automated fact-checking based on the Retrieval-Augmented Generation (RAG) paradigm. Our goal is to benchmark, under more realistic scenarios, RAG-based methods for the generation of verdicts – i.e., short texts discussing the veracity of a claim – evaluating them on stylistically complex claims and heterogeneous, yet reliable, knowledge bases. Our findings show a complex landscape, where, for example, LLM-based retrievers outperform other retrieval techniques, though they still struggle with heterogeneous knowledge bases; larger models excel in verdict faithfulness, while smaller models provide better context adherence, with human evaluations favouring zero-shot and one-shot approaches for informativeness, and fine-tuned models for emotional alignment.