A Query-By-Example Approach to Compose SPARQL Queries in the GF Framework for Ontology-Based Data Access

The gap between legacy data sources and semantic web technologies can be bridged with the Ontology-Based Data Access (OBDA) methodology by providing protocols and tools for translating old data into ontologies. Querying modern ontologies in OWL/RDF files and understood as networks of objects having describing properties and inter-linked by relations requires writing SPARQL queries, where it is essential to have a technical proficiency not usually in the hands of lay users. For studying this problem, we developed a prototype tool called GF to capable of executing arbitrary SPARQL queries posed against OWL/RDF ontologies obtained by OBDA from H2 relational databases as well as Excel and CSV spreadsheets. The tool allows to visually specify a subset of SPARQL selection and construction queries using a Query-By-Example approach. Our research shows that, given an OWL/RDF ontology obtained from any suitable data source, our implementation can express an important proper subset of queries and translate them into executable SPARQL queries.