| The primary goal of the Semantic Web is to provide a web of data that are accessible and processable by machines. Different communities have started to publish their resources using Resource Description Framework (RDF), RDF Schema (RDFS), or Web Ontology Language (OWL), based on the level of expressiveness suitable for representing the knowledge. This approach inherits the existing Web's Representational State Transfer (REST) architecture, whose essential principle lies in the state-less publication of information based on a global and persistent URI.; The Semantic Web will go beyond just providing common formats for data sharing. Semantic Web-based applications are expected to realize the intelligent functionality enabled by the knowledge representation models and techniques. State-full communication and concurrent data mediation are critical among those applications. The tuple space coordination model provides an open, distributed and sealable framework, and is a suitable middleware for Semantic Web. This dissertation discusses the necessary extensions to the original tuple space and presents a semantic tuple space coordination model. An implementation of the semantic tuple space, YourSpace, is used for biomedical data integration and reasoning. With the implementation of microarray clustering services, we also show how YourSpace enables the intelligent discovery and invocation of Semantic Web services. |
