Research On Resource Management And Agent Technology For The Semantic Web

Web, with a rapid development in the last decade, has become a huge global information space nowadays. Due to the syntactic and structural heterogeneity of information resources on the Web, its difficult to effectively manage, retrieve or automatically and intelligently process the Web resources with existing methods and technologies, such as the traditional databases and search engines which mainly based on statistics. The Web is facing a big challenge in this sense. Semantic Web is an emerging area to this aim. Its basic idea is to enrich the available Web resources with machine processable semantic metadata, so as to enable software agents to automatically process the Web resources, and eventually provide intelligent semantic Web services for human users. The key to the success of the Semantic Web vision is to share and interoperate the heterogeneous resources at a semantic level, in practice, which will heavily rely on the efficient organization and management of the Web resources in a semantic way.This thesis proposes a model to organize and management the Web resources for the Semantic Web, after an introduction of the origin of the Semantic Web and a survey of the state of the art of the domain. And agent, as an important user of the Semantic Web, is also addressed. The main researches and contributions are fourfold:Propose a knowledge-based data model. The basic idea is to organize heterogeneous resources according to the knowledge structure of domain ontology, in order to semantically structure and interrelate the heterogeneous resources. First, we analyze the requirements of organizing and managing of the Web resources and discuss the most important three elements i.e., semantics, metadata and ontology. With this respect, we point out the insufficiencies of the existing data models such as the relational model and the XML model, etc. Then, the knowledge-data model is formally defined which is composed of knowledge schema, resource instance base and integrity constraint. The knowledge schema is used to capture the agreed domain knowledge, and the resource instance base is an abstraction of the real Web resources, while the integrity constraint provides to capture the domain rule knowledge that is absent in the knowledge schema. Finally, an RDF Schema for the model is also defined.Propose a Description Logic based resource calculus and an algebra as formal languages for querying the knowledge-based data model. The resource calculus and its formal semantics is defined, in order to provide a fundamental logic framework for later design of user-oriented query language, as well as a formal declarative semantics. Query algebra...