Research On Knowledge Base Grid Model

The emergence of the Semantic Web will bring an enormous amount of knowledge base resources onto the web. How to discover, organize, manage these resources, and provide extensible services with high reliability is a problem needs to be solved urgently. Grid technologies, using their mature distributed computing service architecture on internet, and promising the broad sharing and cooperative using of web resources, become the basic technical support to solve this problem. Knowledge base grid DartGrid/KG is a grid architecture, which focuses on the sharing, management and cooperative using of knowledge base resources in the Semantic Web. It covers the idea of the Semantic Web and Grid, and provides a series of interfaces in service form about management, manipulation, inquiring of knowledge bases. In this dissertation, deep research on its model is introduced surrounding the innovative concept of knowledge base grid, and the topology of DartGrid/KG and some key technologies about the model implementation are mainly described. The main achievements of this dissertation are listed hereinafter: The architecture of DartGrid/KG is discussed in the dissertation. Knowledge server is the basic element of knowledge base grid, which contains ontology knowledge server, rule server and case server. It is the information and knowledge source of knowledge base grid. Ontology server guides the knowledge classification, management and maintenance. Knowledge directory server enables the registration, discovery, and management of knowledge servers. And as the client of knowledge base grid, the semantic browser provides knowledge browsing, inquiring, simple reasoning, grid node management, and etc.The service model and core protocol suite are also defined in this dissertation. The service model of DartGrid/KG is extended from Open Grid Service Architecture, and the knowledge-base-grid-specific ontology service, rule inference service and case-based reasoning service are added in. Ontology service, semantic registration and discovery are the core parts of DartGrid/KG; therefore the corresponding Semantic Registration Protocol, Ontology Query Protocol, Ontology Verifying Protocol and Ontology Fusion Protocol are also the emphasis of this dissertation. The implementation of DartGrid/KG and its application in traditional Chinese medicine are introduced. As the semantic support of DartGrid, DartGrid/KG enables the dynamic semantic registry of resources, distributed semantic query and semantic browsing in knowledge level. The system is almost realized currently, and has fully applied in the research work about Chinese medicine.The emphases of the future work will be the compatibility of multiple ontology languages, consummating and realizing of distributed reasoning, and the design of knowledge level communication protocols.