Dynamic Web Service Composition Based on Service Integration and HTN Planning

Web Service Composition is an important but challenging issue because of the large number and dynamicity of Web services. This paper proposed a dynamic Web service composition framework which based on OWL-S and HTN planning. It uses OWL-S as Web service description language and HTN planning engine to calculate service execution sequence dynamically. Under this framework, we firstly develop a service integration algorithm based on an improved K-Means algorithm. Based on the proposed service integration algorithm, we design a two-stage service match algorithm, which reduce the time cost of service match making with a large number of services. In particular, our HTN planning approach is based on several service types, not all the service instances, which greatly improve the efficiency of the planning procedure.