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

---

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 Afri-qwen, 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.

Health parallel data

from datasets import load_dataset

ds = load_dataset("israel/AfrIFact", "health")

Native culture amharic data

from datasets import load_dataset

ds = load_dataset("israel/AfrIFact", "amharic_culture_news")

AfriFact Dataset Distribution

AfriFact Health

Label	Amharic	English	Hausa	Igbo	Twi	Oromo	Shona	Swahili	Wolof	Yoruba	Zulu	Total
SUPPORT	284	284	284	284	284	284	284	284	284	284	284	3124
REFUTES	283	283	283	283	283	283	283	283	283	283	283	3113
NEI	281	281	281	281	281	281	281	281	281	281	281	3091
Total	848	848	848	848	848	848	848	848	848	848	848	9328

---

AfriFact Culture

Label	Amharic	English	Hausa	Igbo	Twi	Oromo	Shona	Swahili	Wolof	Yoruba	Zulu	Total
SUPPORT	285	263	282	278	278	271	291	307	246	275	286	3062
REFUTES	253	267	213	272	272	271	212	211	256	269	225	2721
NEI	275	283	336	263	263	271	310	295	311	269	302	3178
Total	810	810	828	810	810	810	810	810	810	810	810	8928

---

Disclaimers

The views expressed herein do not necessarily represent those of Lacuna Fund, its Steering Committee, or CENIA.

Duty to Acknowledge

This work was carried out with support from lacunafund.org and google.org