Semantic Web service composition framework based on abstract template

template was proposed,and analysis of the template retrieval,data pre-processing and template adaptation modules of this framework were introduced in detail. In order to improve the accuracy of the template retrieval,the framework adopted the feature matching and graph matching methods in the similarity computation between the abstract template and the user's request. Besides,it constructed the service dependency graph through the data pre-processing,deleted the useless services and verified the satisfiability of the user's request. According to the service dependency graph,it implemented the automatic template adaptation,by using the Heuristic Plan Adaptation Algorithm (HPAA),and improved the adaptability and scalability of HPAA by providing the function of the flexible configuration of the heuristic strategies. Finally,the experiments on the different heuristic strategies show the effectiveness and reliability of HPAA when dealing with the large-scale problems.