Mining Service Social Relations Based on Service Network Modeling and Analyzing

With the development of Web service technology, the quantity of the Web services published on the Internet is increasing rapidly and they are connected through input and output mapping. Although the properties of service network have been explored, the social relationships among services are not formally defined and there lacks practical approach to identify social relationships among real world services. Thus, we define three fundamental relations between Web services in a precise way: substitution, competition and collaboration. Moreover, based on 4 different granularity level Web Service Networks (WSN), an automatic service social relation annotation approach is proposed. We conduct experiments on a corpus crawled from the Internet. The experimental results show that the accuracy of single operation recommendation based on our method is up to 90% on average.