AfriData Commons is a platform designed to bridge the gap in AI training datasets by providing a centralized repository for African community-curated data. By allowing users to upload, manage, and discover datasets representing diverse African norms, cultures, and languages, the platform promotes inclusivity in data-driven systems.
Modern AI systems often lack representation from African contexts, leading to biased outcomes. AfriData Commons addresses this by:
- Enabling users to upload datasets in various formats.
- Providing powerful search, filter, and recommendation tools.
- Promoting dataset discovery and accessibility within African communities.
- User registration with email verification.
- Secure login and authentication.
- Profile management (bio, contact, profile picture).
- Upload datasets with metadata (CSV, JSON, XML, etc.).
- Download datasets with format preservation.
- Dataset access control based on user roles.
- Search datasets by keyword, tags, or filters.
- Advanced search by date, file type, and user-specific filters.
- Personalized recommendations.
- View trending datasets and topics.
- Engagement-based recommendations.
- Frontend: HTML/CSS/JS
- Backend: Django + Django REST Framework
- Database: PostgreSQL
- Search Engine: Elasticsearch (planned)
- Deployment: Docker + CI/CD pipeline
- Diagram: See
Documentation/erd.dot
- Python 3.10+
- MySql
- pip / pipenv
- Graphviz (for ERD generation)
git clone https://github.com/OmbongiFelix/AfriData-Commons.git
cd Afridata-Commons