Service Relation Mining And Network Analysis

As Web services proliferate and being widely used, the function of a single Web service can not meet the complex requirements of users. How to find service relation and use the analysis of the characteristics of web services cluster to promote service discovery have become the focus of research. In response to these problems, the proposition of a Web service relation mining algorithm and the analysis of the mining results give the answer.The proposition of Web service relation mining algorithm based on Service Network (SN) model. With the introduction of parameter layer, this algorithm makes use of the association rules in parameter layer and association rules between parameter layer and service layer to generate sets of candidate services, then get service relation in SN. Service relation mining algorithm can be applied off-line mining process, but also for dynamic SN construction process.Network modeling is based on service relation. The analysis of model uses complex network technology. By using statistical methods, we draw geometric properties and structure of network characteristics, which can provide a basis for the improvement of SN stability and the efficiency of service discovery and composition.In summary, Web service ecosystem is established when cooperation and competition relations were gained. With it and learning the way of its evolution can promote service discovery and reduce the difficulty of service composition, make SN play infrastructure facilities role better.