A Survey of Text and Speech Resources for Hausa and Fongbe: Availability, Quality, and Gaps for NLP Development

In three linesInventory of text and speech resources for Hausa (80-100M speakers) and Fongbe (2M speakers). Hausa has diverse parallel corpora and text collections (news, encyclopedic, educational). Fongbe lacks text data but benefits from recent speech collection initiatives. Both languages represented in Masakhane benchmarks (NER, POS tagging).

Summary generated by Claude — human-verified