Leveraging Ontology-aided AI Planning for Automatic Composition of Semantic Web Services

Composing existing web services into more advanced and complex new services may promote rapid application development, service reuse, and cross-enterprise collaboration. Instead of pre-established relationship between component services in the composition, service oriented computing advocates dynamic discovery, composition and execution of those services according to user requirements. In this paper, the issue of leveraging AI planning technique to automate web service composition is investigated. We propose a flexible framework for dynamic composition of semantic web services which can facilitate automatic composition schema planning and optimized service selection based on non-functional information. Then we present an ontology-aided composition schema planning approach for semantic web services. A planning-graph heuristic partial-order planning algorithm is given to implement the planner.