As web services proliferate, finding a service that can perform a given request becomes more difficult. In some cases, a composition of several services may be required. In semantic web service composition, a repository of services is given in which input and output parameters of each service are annotated with a concept from ontology. Given request is defined by a set of known input concepts and a set of wanted output concepts. Semantic composer should find a sequence of services, called composition that satisfies the wanted output concept. In this paper, a semantic composer based on ant colony optimization method is proposed. This semantic composer is tested using the challenge set of Web Service Challenges (WSC) 2008. The proposed composer can find optimal composition length in each challenge set in a reasonable time. The result of the suggested composition has been compared with the best result reported in WSC 2008 and in, most cases; it performs similar or better than the other systems.

Keywords