The scope of this work is knowledge management in the Semantic Web. Its contribution is the design, development, and characterisation of an application node architecture for the Semantic Web, called Swan. The core of the Swan architecture consists of an OWL knowledge base which is supplemented by a hybrid reasoning mechanism. Hybrid reasoning in Swan combines F-Logic and Description Logic reasoning. Update operations for the manipulation of the knowledge base are provided with a well-defined semantics. Intensional updates are possible and allow to define updates with respect to implicit knowledge. A unique trigger mechanism is used for the completion of intensional updates, but also for the maintenance of knowledge base integrity. By its ability to process abstract action definitions, the application node can be integrated directly into the event-driven Semantic Web architecture Mars. A protoype of this architecture has been implemented, which shows the flexibility and applicability of its concepts.
Keywords:Semantic Web, Event-Driven Architecture, OWL, Active Knowledge Base, Intensional Updates, Hybrid Reasoning, Description Logic, FLogic.
ContentsAbstract 3