Research On The Modularization Of Ontologies In Semantic Web Services

The Semantic Web, which can provide enhanced information based on the exploitation of machine-processable meta-data, is the extension of the current web. Central to the vision of the Semantic Web are ontologies, which could represent shared domain descriptions that different applications and agents use.Ontologies grow dramatically in size and quantity while the difficulty of developing, maintaining and reusing large ontologies appears as the bottleneck of realizing the Semantic Web. The design and integration of ontologies is a tedious and costly challenge. While providing a shared, tested, and well-accepted vocabulary, many of the large standard resources are huge, often containing tens of thousands of concepts. However, many users need only a small fraction of the resource for their application.Modularization is an established principle in the Software Engineering. Modular representations are easier to understand, reason with, debug, extend and reuse. Recently, the benefit of modular representations of ontologies and modularizing ontologies in the sense of splitting up an existing ontology into smaller interconnected parts has been recognized and discussed in the Semantic Web community.This paper reviews the concepts of ontology and ontology modularization and proposes a framework of ontology modularization applications. As core of the paper, a method of Structure-based Ontology Partitioning is designed. In the method, the graph mode is constructed from querying results, simplified in the edges and transformed into the Dependence Graph after calculating the dependence between nodes. Finally ontology partitioning is done through decomposing the Dependence Graph where an Island Algorithm is used. Finally, a ontology modularizing system is developed.