Research On SPARQL Query Over Relational Databases Based On SQL Convertion

Semantic Web is a extension for the function of current Web, making the information in the semantic Web can be understood by computers, enhancing the interaction and the collaboration between human and computer. Semantic Web uses RDF to describe the various resources on the Web, and gives them semantics. SPARQL is the recommendatory query language for RDF resources. Its principle of query is based on graph pattern matching. The users can express their semantic query as SPARQL, which can inquire on the semantic Web, so as to achieve the purpose of semantic inquiring.But when because of the needs of management of RDF semantic infomation, storing the RDF information in the relational database, the efficiency of SPARQL inquiring RDF will be greatly reduced. And Because the standard language of relational database is SQL, so it has a long significance to study how can convert SPARQL query into SQL query for the efficiency increasing.This paper mainly studies how to make the SPARQL query with the semantic query demands of users, and inquire the RDF resources stored in the relational database directly and efficiently, based on SQL convertion, with the guarantee of recall and precision. This paper firstly deeply studies the inquiring mechanisms of SPARQL, and the converting from the basic SPARQL query to SQL query; Then based on the basic query convertion, puts forward the realization of complex graph models and query models, and then puts forward the realization method of SPARQL query; Finally, according to the realization method, we realize a experiment system SPARQLtoSQL. The experimental system uses a RDF ontology, which is stored in a relational database, as experimental case, analyzes and compares the results with other methods, which effectively proves the feasibility, correctness and effectiveness of SPARQLtoSQL.