Responsible resources
How can consent, clear purpose, and documented provenance shape useful language resources?
GamVoiceResearch
Our research direction starts with responsible language-resource development and thoughtful speech and NLP experimentation.
The research context
Underrepresentation in AI involves more than the languages an interface accepts. It also concerns the availability of language data, evaluation methods, research resources, and linguistic quality, alongside appropriate access and stewardship.
Our research direction focuses on strengthening these foundations for Gambian languages through documented methods, language expertise, and responsible resource development. Research questions guide this work; validated results will be shared when ready.
How can consent, clear purpose, and documented provenance shape useful language resources?
What can carefully developed resources make possible for future speech and natural-language-processing experiments?
How can clear guidelines, human review, and disagreement resolution improve consistency?
How should writing conventions and variation be documented with linguistic and community input?
How can documented methods and appropriate access support work that others can understand and assess?
How can researchers and language experts shape useful questions together while respecting data boundaries?
Research outputs
Research outputs and publications will be listed here when they are available for public release.