DISERTO: Semantics-Based Tool for Automatic and Virtual Data Integration

In various domains, the continually increasing volume of heterogeneous data generated from different sources is overwhelming. However, the exploitation of this data is still limited while, in most cases, sources are not interoperable, and hence, data are not linked. These issues are due to the semantic, syntactic, and schematic heterogeneity of the data. In this work, we propose a semantic virtual Data Integration TOol based on Semantic Enhancement and RML (RDF Mapping Language) mappings called DISERTO. It automatically generates RML mappings through domain ontology and thesaurus to categorize the semantics behind the data. It follows three main steps: (1) extracting relevant metadata (data schema), (2) mapping metadata to a domain ontology by taking advantages of RDF quads, and finally (3) generating RML mappings. To validate the tool, we provide a case study based on real data provided by the Sahara and Sahel Observatory (OSS) and the National Oceanic and Atmospheric Administration (NOAA). It shows a semantic annotation followed by an RML mapping generation of a raster input image and CSV files.