Zero-Shot Learning and Key Points Are All You Need for Automated Fact-Checking
Mohammad Ghiasvand Mohammadkhani, Ali Ghiasvand Mohammadkhani, Hamid Beigy
TL;DR
This work addresses automated fact-checking within the AVeriTeC framework by advocating a zero-shot prompting approach using large language models. It introduces ZSL-KeP, a simple pipeline that constructs diverse key points from claims, performs extensive retrieval with open-web references, and generates evidence in a question-answer format to justify a verdict, all without fine-tuning. The method yields improvements over the baseline in retrieval and overall AVeriTeC scores and ranks 10th among 23 submissions, demonstrating that a compact zero-shot, retrieval-driven strategy can be competitive for open-web fact-checking. The approach highlights the practical potential of leveraging long-context LLMs with structured retrieval to produce transparent, evidence-backed decisions, while acknowledging limitations related to input length and future opportunities with stronger models and longer contexts.
Abstract
Automated fact-checking is an important task because determining the accurate status of a proposed claim within the vast amount of information available online is a critical challenge. This challenge requires robust evaluation to prevent the spread of false information. Modern large language models (LLMs) have demonstrated high capability in performing a diverse range of Natural Language Processing (NLP) tasks. By utilizing proper prompting strategies, their versatility due to their understanding of large context sizes and zero-shot learning ability enables them to simulate human problem-solving intuition and move towards being an alternative to humans for solving problems. In this work, we introduce a straightforward framework based on Zero-Shot Learning and Key Points (ZSL-KeP) for automated fact-checking, which despite its simplicity, performed well on the AVeriTeC shared task dataset by robustly improving the baseline and achieving 10th place.