Dynamic Service Matching Method Based On Virtual WEB Service

With the increasing number of Web services with similar functions and different qualities on the Internet,how to choose the best service for users becomes a problem that needs to be solved urgently.Because clients are difficult to sense the changes of Web services timely in the dynamic environment of Internet,it causes problems such as the invalidation of call.Therefore,we need to use virtualization technology to complete the dynamic call of Web services by providing a uniform interface,so as to better meet the personalized needs of users,and this interface is implemented by virtual Web service system.As for which real Web service is selected in this system,it is necessary to focus on the research hotspot of service matching.Currently,there are data sparse and cold-start conditions for various matching methods based on syntax and semantic to a certain extent.In order to solve above problems,this paper combines the related research of virtual Web service and semantic Web service to carry out following work.Based on the concept of existing virtual Web service,this paper proposes a new implementation scheme of virtual Web service matching engine firstly.The matching engine consists of static matching and dynamic matching.Static matching engine mainly solves matching problem between request service and publishing service,and the function of dynamic matching engine is to provide a unified interface for target user to realize the dynamic call of Web service.Secondly,this paper proposes a static matching engine of virtual Web service which includes function matching module,QoS prediction module and non-functional matching module.Among them,function matching module retains original matching algorithm and filters same or similar Web services for service requester.QoS prediction module first adopts the spectrum clustering of community discovery algorithm to divide users into communities,and then uses the geographic location information of Web services for secondary clustering,which saves a lot of time for subsequent calculations.The hybrid collaborative filtering method is used to match optimal service for target user at last.In the non-functional matching module,due to the influence of semantic information of Qos attributes on service matching,this paper proposes a service matching method that combines Qos semantics and Qos values.In addition,this paper combines virtual Web service unified interface,QoS ordering module and load balancing module to implement a dynamic service matching method based on virtual Web service.This method first calculates the Qos integrated values for candidate Web services,and simply sorts services according to these Qos comprehensive values.Then,load balancing module performs user quantity detection and response time detection for Web services which passed by Qos sorting module in turn,so that filters out services which have invalid or poor performance for service requesters.The virtual Web service unified interface adopts the call of polymorphic to provide transparent service matching for target user after that.At the same time,this paper proposes a dynamic matching evolution mechanism to help virtual Web service system continuously optimizing dynamic matching results,thereby improving the user experience.Finally,this article not only lists the pseudocode of the core algorithm in detail,but also presents its expected blueprint in the background of e-commerce applications through a prototype of virtual Web service system.And it uses a data set which named WSDream from real world to conduct a large number of comparative experiments and related parameter analysis.The experimental results show that the QoS prediction algorithm proposed in this paper can not only ensure matching accuracy in the case of sparse data,but also performs well in time complexity.At the same time,the feasibility verification and comparison experiments of Qos-based Web service matching algorithm proposed in this paper are also carried out.The results show that this algorithm not only meets the personalized requirements of service requester,but also improves the recall of service matching to some extent.