Looking forward to designing Semantic Web services
Initially, the user completes the goal specification that defines the task the application needs to perform which is handled by a broker that accesses distributed digital libraries of reusable problem solving methods on the web. Then the broker selects, adapts and configures these components according to the specific request and provides a specification of the configuration of components that can be executed. Based on a competence modelling architecture- the Unified Problem-solving method Modelling Language (UPML)- competence specifications support the broker's reasoning. Among the various libraries of software components there are classification, document analysis, information filtering, case based reasoning, and scheduling. Additionally, several spin-off applications have been developed for testing purposes involving a number of classification applications, a Web Information Mediator, a health care service and a conference paper analysis and classification application. The IBROW project experimented with many brokering approaches and combined research on heterogeneous databases, interoperability and web technology with knowledge systems and ontologies. The key results derived include an elaborate descriptive framework, rich ontologies, distributed component libraries, competence specifications and context dependent brokering architectures and strategies are expected to constitute the sound basis on which future Semantic Web services will be developed.
