Initiative Details
Title:
Ghana Maternal Health Q&A Dataset
Details:
Introducing the #Ghana #Maternal #Health Q&A #Dataset
A collection of 20,000 real-world, code-mixed Twi–English questions and answers covering the full spectrum of maternal health.
Most maternal-health conversations in Ghana don’t happen in polished English. They happen in the lively mix of Twi and English we use when we’re worried, asking for help, or trying to explain a symptom clearly. This dataset captures that reality at scale, enabling smarter and culturally grounded health-AI systems.
What’s inside the dataset?
> 17 major topics including #hypertension, #bleeding, #ANC routines, #nutrition, #fetal #movement, #labor, #postpartum #care and more
> Detailed metadata: urgency levels, pregnancy stages, care settings
> Natural code-mixing patterns that reflect how Ghanaians communicate around health concerns
🚀 What can you build with it?
> Maternal-health chatbots
> RAG systems for frontline health workers
> Mobile health apps
> NLP research on Ghanaian languages and code-mixing
> Clinical training tools and public-health communication assistants
This dataset was created by the Ghana AI Research Network (GAIN) team as part of our commitment to support developers, researchers, and innovators shaping the future of AI in Ghana and across Africa.
If you're building #AI for #maternal #health or #social #impact, this is a resource worth exploring, see Kaggle: Ghana Maternal Health Q&A Dataset https://www.kaggle.com/datasets/ghanaairesnet/ghana-maternal-health-twi-english-q-and-a-dataset