Implementing a Semantic Web Blackboard System using Jena

In this paper, we discuss the need for a hybrid reasoning approach to handing Semantic Web data and explain why we believe that the Blackboard Architecture is particularly suitable. We describe how we have utilised it for combining ontological inference, rules and constraint based reasoning within a Semantic Web context. After describing the metaphor on which the Blackboard Architecture is based we introduce the key components of the architecture: the blackboard Panels containing the solution space facts and problem related goals and sub-goals; the differing behaviours of the associated Knowledge Sources and how they interact with the blackboard; and, finally, the Controller and how it manages and focuses the problem solving effort. To help clarify, we use our test-bed system, the AKTive Workgroup Builder and Blackboard (AWB+B) to explain some of the issues and problems encountered when implementing a Semantic Web Blackboard System in Java, using Jena. We also discuss our reasons why we elected to use the Jena toolkit and explain its usage within several of the key components of our system.