Integrating Context and Intention for Optimal Semantic Web Service Composition Using AI Planning

This paper deals with an approach for the dynamic composition of web service (WS) based on the user’s intention and context. Our idea to implement this approach is based on the combination of artificial intelligence (AI) planning techniques and the semantic aspect describing the web service that will be composed. To this end, we will transform the web service composition problem into an AI planning problem. First, conversion of web service semantically described in OWL and OWL-S to an alternative problem and domain description specified in the planning domain description language (PDDL) is made. Then, a planner (AI Planner) using the Genetic Algorithm (GA) features will be invoked, by adapting the PDDL planning language actions to the GA rules, in order to generate the most optimal composition plan (sequence of PDDL actions), which solves the conceived service composition problem, having a goal to achieve (the intention) according to the problem domain description (context).