Utilizing Weighted Ontology Mappings on Federated SPARQL Querying

摘要

Technologies to allow the heterogeneity of ontologies on the web of data have a key role on the emergence of Linked Open Data. Much research exists on generating better ontology mappings, and also they produce 'weights' (i.e. confidence) of each generated mapping. Although the semantics of such weighted ontology mappings has been discussed, how they can effectively be used in querying remains as a crucial issue. In this paper we show how such weighted ontology mappings can effectively be used in SPARQL-based querying on heterogeneous data sources by slightly extending the syntax of SPARQL query language. We show how such an extended query can be translated to a standard SPARQL query and how the extended query can customize the behavior of processing the query by reflecting the user's demands.