Research On Evolving Model Of Web Service Network

With the development of the distributed computing technology, the network software application based web services has become the mainstream architectures. The calculated elements of the services software system are much larger and more complex than the tradition modules and classes in the particle size and structure. The services software system shows the characteristics of autonomy, heterogeneity and distribution. To meet the personalized and diversified needs of customers, the system generates some larger particle size components by combining and connecting web services. As a result, it is much more uncertain than before. In this situation, the problem must be addressed that how to understand and analyze the complex system.This paper presents the method of constructing service-oriented software evolution model combining with the researches on related complex network models, the features of service combined and the service software. This paper emphasizes on the global analysis of web services, being different from the previous researches on web services which pay attention on individual services and the relations of them. The study on the characteristics of the overall service system is conducive to understand the behavior pattern and evolution of the services system from different research dimensions.This paper includes three parts as follows:At first, the definition and models of web services network are proposed. And then, the paper will discuss the data-basic for the construction of web services which contains description files, interaction logs, process documents,Qos of services and the feedback of users.Secondly, the paper proposes an approach to build web services network mode based syntax and expounds how to collect and clean the data of web services. Four kinds of syntactic information matching schemes and the features of the model invoked before are given in this part. The model of web services network has the characteristics of power law distribution and small world.At last, the paper shows statistical properties of web services network by adopting complex network theory. The paper analyzes the behavior of interactive services. And then it provides Web service behavior evolving model based on preferential attachment and the mathematical reasoning. Finally the simulation generates Web service behavior evolving model and obtain the statistical properties.