BOUTEF: A Multilingual Corpus for FakeNews in North Africa -- Language as a Weapon
Researchers have developed BOUTEF, a new multilingual corpus aimed at studying fake news in North Africa, specifically Algeria and Tunisia. This dataset includes fake and genuine narratives, user comments, and debunking information across various languages and dialects, including Arabic dialects, French, and English. The analysis reveals that fake news often uses emotionally charged language and sensational framing to increase virality, while debunking content is more factual. The corpus is intended to advance research in fake news detection and low-resource language processing. AI
IMPACT Provides a valuable resource for developing and testing AI models for fake news detection in under-resourced multilingual contexts.