Research On Mapping Relational Database To RDF (S)

With the semantic Web's phenomenal growth, to build ontologies by using exiting data stored in relational databases has gotten more and more attention. Resource Description Framework (RDF) and Resource Description Framework Schema (RDFS) play an important role in semantic web. Presently, RDF and RDFS have been widely used in various fields of Semantic Web. Virtually, most of data are stored in relational databases. Therefor, how to create ontologies based on data stored in relational databases is a highly meaningful research orientation. In order to make full use of existing information in the relational database to enrich the data on the Semantic Web, this thesis proposes to build RDF(S) ontologies with the existing relational databases automatically.This thesis studies an approach to map relational database to RDF(S) ontologies. Firstly, this thesis gives formal definition of the relational database schema and the RDF(S) data model respectively by means of analyzing and comparing them. Then, extracts semantic information in relational database by analyzing primarykeys, attributes, foreign keys and partial data, and according to which, groups of mapping rules from relational database to RDF(S) are proposed including rules for the tables, fields and records. These rules are demonstrated in detail with examples. Finally, a prototype system for mapping relational database to RDF(S) is developed based on these mapping rules, and the correctness and superiority of which are verified through experiment.The approach to map relational database to RDF(S), which is proposed in this thesis, can extract more information from relational database and takes into account the migration of tuples. Furthermore, this mapping method builds RDF(S) ontologies which are independent of relational database, and it will help the expansion of the inner-ontologies, but also conducive to the integration of inter-ontologies.