AfrIFact：面向非洲语言的文化信息检索、证据抽取与事实核查数据集
AfrIFact: Cultural Information Retrieval, Evidence Extraction and Fact Checking for African Languages
研究者发布 AfrIFact 数据集，覆盖信息检索、证据抽取与事实核查全流程，涵盖十种非洲语言和英语。评测显示最好的嵌入模型也缺乏跨语言检索能力，LLM 在非洲语言上的多语言事实核查能力不足，few-shot 提示让 AfriqueQwen-14B 提升最多 43%，任务微调再提升事实核查准确率最多 26%。
Authors: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
Abstract: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.
| Comments: | Accepted at 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026) |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2604.00706 [cs.CL] |
| (or arXiv:2604.00706v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2604.00706 arXiv-issued DOI via DataCite |
Submission history
From: Israel Abebe Azime [view email]
[v1]
Wed, 1 Apr 2026 10:08:37 UTC (2,950 KB)
[v2]
Wed, 29 Apr 2026 14:10:39 UTC (2,950 KB)
[v3]
Wed, 21 Oct 2026 10:50:09 UTC (1,295 KB)
来源：arXiv cs.CL · arxiv.org