Creating Semantic Data from Relational Database

The semantic web technology contributes to finding accurate information on the basis of the meaning of words or improving access capabilities to information through inference between meanings. However, the reason that the semantic service is not spreading is that the semantic technology is not yet settled for practical use. Another reason is that it is not easy to build ontology which is used as a knowledge representation model for providing the semantic service. Previous studies have concentrated on the process of automatically creating ontology in a RDB (Relational Database), or automatically converting it to instances. This study focuses on allocating a unique identifier to the existing DB data through entity identification to enable more accurate services to be provided in the process of converting the RDBMS data to ontology instances. Besides, this study extracts triple data including important entities and their relationships from unstructured texts stored in BLOB (Binary Large Object) format by using text mining technology, and makes them ontology instances. The DB-to-OWL converting system (Onto URI) uses a mapping rule between the DB schema and the ontology schema, an identification rule for identifying entities, text mining for extracting semantic triple, and the authority data in order to effectively support automatic creation of ontology instances with DB data. As a result, the proposed system contributes to creating semantic data automatically from relational database through URI allocation and identification.