Ontological Bridge Building - Using Ontologies to Merge Spatial Datasets

This paper discusses the approaches we have taken to addressing issues of ontology merging and mapping from an ontology to a database, which arise from the semantic data integration problem. We present a proposal for curtailing the extent of an ontology module referenced within another ontology, so that the entire ontology need not be imported if only a small subset is required for reuse. As the national mapping agency for Great Britain, we are particularly concerned with the integration of spatial data, so the second contribution of this paper is to describe how we incorporate a spatial relations ontology into a data ontology that describes the mapping from domain classes to database. Since the spatial relations ontology is based on the Egenhofer 9 Intersection model employed by the Oracle spatial database, spatial SQL queries can be automatically generated from the data ontology, or a SPARQL query using the domain ontology terms.