Masakhane

Language research built as a community

Masakhane organizes African-language natural language processing research with participation from researchers and language speakers. Its community model broadens who can contribute data, linguistic knowledge and experiments, rather than treating African languages only as evaluation targets for systems built elsewhere. [1][2]

Its own history places the formal community launch in 2019 at the Deep Learning Indaba in Nairobi. That retrospective supports a year-level date. The community is not automatically identical to a later legal research foundation, so this record does not infer a corporate parent or legal incorporation date. [2]

MasakhaNER is a concrete output: the 2021 paper presents named-entity recognition resources and experiments for ten African languages, with released data and models. Its October journal publication is evidence of the research, rather than proof of the first day every artifact became public. [3]