arXiv 2604.007062026-04-29cs.CL

AfrIFact：非洲语言的文化信息检索、证据提取与事实核查

AfrIFact: Cultural Information Retrieval, Evidence Extraction and Fact Checking for African Languages

Masakhane NLP；Saarland University（萨尔兰大学）；Saarland Informatic Campus（萨尔兰信息学院）；National Institute of Informatics（国家信息研究所）；University of Waterloo（滑铁卢大学）；University College London（伦敦大学学院）；Instituto Politécnico Nacional（国家理工学院）；University of Ilorin（伊洛林大学）；Universidad de Deusto（德尤斯大学）；Bayero University（贝亚罗大学）；University of Lagos（拉各斯大学）；Black Swan（黑天鹅）；Addis Ababa University（亚的斯亚贝巴大学）；Universite Alioune Diop（阿利乌内·迪普大学）；McGill University（麦吉尔大学）

Israel Abebe Azime, Jesujoba Oluwadara Alabi, Crystina Zhang, Iffat Maab, Atnafu Lambebo Tonja, Tadesse Destaw Belay, Folasade Peace Alabi, Salomey Osei, Saminu… 展开作者

Israel Abebe Azime, Jesujoba Oluwadara Alabi, Crystina Zhang, Iffat Maab, Atnafu Lambebo Tonja, Tadesse Destaw Belay, Folasade Peace Alabi, Salomey Osei, Saminu Mohammad Aliyu, Nkechinyere Faith Aguobi, Bontu Fufa Balcha, Blessing Kudzaishe Sibanda, Davis David, Mouhamadane Mboup, Daud Abolade, Neo Putini, Philipp Slusallek, David Ifeoluwa Adelani, Dietrich Klakow

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出AfrIFact数据集，针对非洲十种语言和英语进行自动事实核查，发现嵌入模型在跨语言检索能力不足，而少量提示和任务特定微调可显著提升事实核查准确性。

摘要：

评估在线声明的真实性是一项复杂且重要的任务，具有现实影响。当这些声明针对信息获取有限的社区，且内容涉及健康和文化问题时，后果更加严重，尤其是在低资源语言中。本文介绍AfrIFact数据集，涵盖信息检索、证据提取和事实核查的必要步骤，覆盖十种非洲语言和英语。评估结果显示，即使最佳嵌入模型也缺乏跨语言检索能力，文化与新闻文档比医疗领域文档更容易检索，无论是在大规模语料库还是单个文档中。我们还发现，LLMs在非洲语言中缺乏稳健的多语言事实验证能力，而少量提示在AfriqueQwen-14B中可提升性能达43%，任务特定微调进一步提升事实核查准确性达26%。这些发现及AfrIFact数据集的发布，鼓励低资源信息检索、证据检索和事实核查的研究。

英文摘要：

Assessing the veracity of a claim made online is a complex and important task with real-world implications. When these claims are directed at communities with limited access to information and the content concerns issues such as healthcare and culture, the consequences intensify, especially in low-resource languages. In this work, we introduce AfrIFact, a dataset that covers the necessary steps for automatic fact-checking (i.e., information retrieval, evidence extraction, and fact checking), in ten African languages and English. Our evaluation results show that even the best embedding models lack cross-lingual retrieval capabilities, and that cultural and news documents are easier to retrieve than healthcare-domain documents, both in large corpora and in single documents. We show that LLMs lack robust multilingual fact-verification capabilities in African languages, while few-shot prompting improves performance by up to 43% in AfriqueQwen-14B, and task-specific fine-tuning further improves fact-checking accuracy by up to 26%. These findings, along with our release of the AfrIFact dataset, encourage work on low-resource information retrieval, evidence retrieval, and fact checking.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。