A web-centric semantic mediation approach for spatial information systems

Semantics-related issues are at the heart of web-centric information systems and emerging spatial applications that require an integrated access to collections of heterogeneous data sources. We present an ontology-based semantic mediation approach and its application to spatial system interoperability. The approach is based on three contexts described by ontologies used to capture the semantics of data sources and to resolve their semantic discrepancies. The global ontology context defines generic application domain mediation concepts while the ontology local contexts are devoted to the description of local concepts. The cooperation contexts provide semantic concepts that encapsulate three key components: (1) semantic roles defined by ontological agreements on the global ontology, (2) virtual views defined on the local ontologies, and (3) context transformation mappings used to define inter-ontology relationships. We illustrate the proposed mediation approach on spatial information systems, relying on the description and reasoning capabilities of OWL to establish relations between concepts of the different ontologies.