Research On The Discovery And Evolution Method Of Service Patterns

With the rapid development of the Internet,web services are in more ways presented in daily life in the context of big service research in the big data era.Single service forms large-scale service program through a reasonable combination to meet the increasing personalized needs of more users.On the one hand,massive service requirements have a very high demand for the service composition efficiency.On the other hand,users' expectations for the service not only require fast and convenient but also want to get the most satisfactory results.Based on the above background,the concept of service pattern has been proposed and introduced into the service composition process.The core idea is to use high-frequency effective small-scale service composition results to form a service pattern.Taking the service pattern as the unit of the composition to finish the fast and efficient process of the servie scheme construction,which improves the composition efficiency and customer satisfaction.In the relevant research of service pattern,the design,discovery and evolution technology of service pattern are the foundation and core work.Taking the theortical research as the basis,this paper studies the concept,description,core nature and usage of service pattern,and puts forward the discovery algorithm and evolution methods of service pattern.Finally,all these achievements have applied in the field of health care service.In the part of theoretical research,this paper first studies the concept and characteristics of the service pattern,gives the formal expression and analyzes its three domain characteristics.Then,based on the feature of its characteristics,the paper abstracts the service pattern into the form of graphs.According to the frequent subgraph mining technology,an algorithm to discover service pattern based on graph mining is proposed.The performance of the algorithm is evaluated by a lot of detailed experiments.The relationship between the parameters of the algorithm and the influence on the algorithm are analyzed.Pattern-based Greedy Service Composition Algorithm is used to consider the merits of the service pattern discoverd,which provides the guidance for the practical application of the algorithm.And then the paper makes a research on the evolution law and methods of service patterns.A set of evaluation indexes and their calculation methods are firstly set up for service pattern and single service.The eight evolution types are analyzed and the evolution strategy for each type is given,which makes the service pattern constantly updated and optimized to the maximum degree of availability.In the part of the application,the proposed algorithm and technology are applied to the system design.A service pattern management tool is designed and developed which implements effective discovery,management,evolution and other operations for service patterns.The whole program becomes the validation of the proposed algorithm and carries out in the field of health care services,which verifies the correctness and effectiveness of the proposed methods.