Résumé
Ensuring sustainable access to drinking water remains a significant challenge in Madagascar, where data availability and quality are often fragmented and inconsistent. In the Haute-Matsiatra region, monitoring the use of drinking water is hindered by the lack of interoperable information systems that can integrate spatial, statistical, and field data. This article proposes a data model that facilitates the integration of multi-source, multi-scale, multi-access heterogeneous data from various water management entities operating in the Haute Matsiatra region. The proposed data model extends standard vocabularies commonly used to describe water network supply, facilitating semantic interoperability and consistent data integration. The proposed data model facilitates integration using a PostgreSQL/PostGIS environment and open-source GIS tools, enabling more reliable monitoring and data analysis of endpoint drinking water distribution. Preliminary results demonstrate the potential of the data model to conduct water access analysis across the Haute Matsiatra region. The output of this work will contribute to the regional water resources observatory. This approach contributes to the broader challenge of designing interoperable and sustainable semantic infrastructures for water management in developing countries.