An Efficient Approach to Web Services Discovery and Composition when Large Scale Services are Available

Web services composition and discovery are two most important research issues in the study of Web services. Based on the study on previous known conferences it is pointed out that the Web service composition tree (WSCT) is the key that affects the performance of Web services compositions when a large number of services are available. Then the mutual search operations among Web service operations, inputs and outputs are studied, and a novel data structure called double parameter inverted file (DuoParalnvertedFile) is proposed to implement these operations. An algorithm to build DuoParalnvertedFile is provided as well. Based on above results an efficient algorithm for constructing the WSCT is given. Furthermore, the QoS (quality of service) of Web service composition results (WSCRs) is defined, and then an algorithm used to get the best WSCR in terms of QoS is proposed. Experiments show that this approach not only ensures the success of the performance of Web services composition but also ensures the QoS of Web services composition