Résumé
This PhD project, started in January 2025, aims to develop a knowledge-driven system to explain the impact of extreme climatic events and environmental changes on the health of Maasai communities in Kenya and Tanzania. It leverages knowledge graphs to structure and interconnect heterogeneous data sources (health, climate, local practices) and integrates a Retrieval-Augmented Generation (RAG) system. The knowledge graph serves as a structured repository, facilitating the extraction and organization of relevant information. The RAG system utilizes these structured insights to generate context-aware, traceable, and comprehensible explanations tailored to different stakeholders, including local communities, researchers, and policymakers. Conducted within the MOSAIC project, this research combines artificial intelligence, semantic web technologies, and environmental health studies to enhance scientific understanding and support informed decision-making.