Face the Facts! Evaluating RAG-based Pipelines for Professional Fact-Checking

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Russo, Daniel, Menini, Stefano, Staiano, Jacopo, Guerini, Marco
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908617606692864
author Russo, Daniel
Menini, Stefano
Staiano, Jacopo
Guerini, Marco
author_facet Russo, Daniel
Menini, Stefano
Staiano, Jacopo
Guerini, Marco
contents 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, following professional fact-checking practices, 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.
format Preprint
id arxiv_https___arxiv_org_abs_2412_15189
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Face the Facts! Evaluating RAG-based Pipelines for Professional Fact-Checking
Russo, Daniel
Menini, Stefano
Staiano, Jacopo
Guerini, Marco
Computation and Language
Computers and Society
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, following professional fact-checking practices, 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.
title Face the Facts! Evaluating RAG-based Pipelines for Professional Fact-Checking
topic Computation and Language
Computers and Society
url https://arxiv.org/abs/2412.15189