Open community of academics, researchers, linguists, students, and NLP enthusiasts dedicated to advancing Hausa NLP. Based at the Department of Computing, Bayero University Kano (BUK), Nigeria. Hausa is the second most spoken language in Africa (~120M L1 + 80M L2 speakers) but remains severely under-resourced for NLP. HausaNLP has built foundational resources including AfriSenti (110K+ tweets, 14 African languages, organized as SemEval-2023 Task 12), NaijaSenti (sentiment corpus for 4 Nigerian languages), HaVG/HaVQA (multimodal Hausa datasets), AfriHate (hate speech for 15 African languages), and BRIGHTER (emotion detection for 28 languages, Best Resource Paper at ACL 2025). Also maintains a resource catalogue at catalog.hausanlp.org and the Kamusun Hausa dictionary.
The primary community driving Hausa NLP. Hausa is spoken across Nigeria, Niger, Ghana, Cameroon, and Chad -- core WCA territory. Their sentiment analysis tools (AfriSenti, NaijaSenti) are directly useful for analyzing community feedback from U-Report and similar platforms. Hate speech detection (AfriHate, HERDPhobia) relevant for content moderation. Strong Masakhane connections. Lead researcher (Shamsuddeen Muhammad) is a Google DeepMind fellow with high academic impact.
Extremely prolific research group -- multiple best paper awards at top venues (ACL, NAACL, EMNLP). 9 models and 6 datasets on HuggingFace. Also organized SemEval shared tasks in 2023, 2024, and 2025. Idris Abdulmumin (core contributor) is at University of Pretoria DSFSI -- another important connection in the African NLP ecosystem.