A Multi-Agent Planning Architecture for Semantic Web Service Composition

With the rising popularity of Web services due to their potential to enable interoperability between applications implemented on different platforms, new issues far more complex and crucial than interoperability have appeared, related mainly to Web services composition. Indeed assuming that for each service request there exists a single Web service that perfectly satisfies it on its own is rather unrealistic, in many cases composing functionalities offered by different services may be needed to satisfy a client request. However, current processes of Web services composition tend to be static in the sense that such processes are normally generated off-line and manual as it requires human intervention to be managed. Therefore in order to give value added customized services to users, recognition is growing of the need for automation and dynamicity to be taking in consideration while designing systems supporting Web services integration. Since automated dynamic composition process allows to reduce the time spent in building composition schema, to eliminate human errors and to reduce the overall cost of the process. Thus our research contribution lies in the area of automated dynamic composition of Web services. We propose a conceptual approach to compose Web services by combining the Semantic Web technologies, Artificial Intelligent Planning methods, Multi-Agent architecture and Self-healing propriety of Autonomic Computing paradigm