Study On Extraction And Storage Of OWL Ontology Based On Relational Database

With the extensive use of information system, more and more information is stored in RDB(Relational Database). How to integrate data in different formats to satisfy the need of information integration and analysis has become a research hotspot in information domain. Ontology can describe the concept as a modeling tool on semantic and knowledge level, which can solve the problems of information integration and share in information domain effectively and provide a solution to multi-source data integration. How to transform data in RDB into ontology model for deeper integration and how to store ontology duration for further analysis have become more urgent problems to be solved. This paper is based on the above contents, and its main work and contributions are:1. Analyze the problems of ontology extraction and storage after introducing the main research status in this domain, and then expatiate and synthesize the key technologies including ontology description language, ontology semantic expression to solve these problems.2. After analyzing the disadvantages of current mapping methods from RDB model to OWL ontology, an universal abstract extraction model from RDB model to ontology is proposed based on the formalized definitions of RDB and ontology. Then, a set of mapping rules from RDB model to OWL ontology is designed, which can extract semantic information from RDB, parse RDB model into OWL ontology model, and generate OWL ontology files automatic. The test results show that this approach is universal and effective.3. The universality and particularity principles are raised after intensive study of limitations of current ontology storage patterns based on RDB. Then, storage pattern and transformation algorithm are proposed for OWL ontology. The experiment results show that this storage pattern is clear in structure and efficient in query, which is more appropriate for macro scale ontology files storage.