Research On Qos-based Web Service Prediction Recommendation Mechanism And Algorithm

With the development of a Web Services which is the Service-Oriented Architecture(SOA) framework,it is a challenge for inexperienced users to choose the appropriate service.Therefore,the framework needs to predict and recommend services based on quality of service(QoS),and the overall framework mechanism and algorithm research becomes more and more important.The service scheduling framework is established through the related technologies of Web services,and a set of prediction and recommendation mechanisms are formed by processing the QoS,so that the service system can operate completely and efficiently.Based on the sampling,prediction and other important aspects of the service prediction mechanism,this thesis solves the cold start problem and sparse problem in the Web service prediction process.Based on the analysis of the probability distribution of the Web QoS dataset,the Web service recommendation for the QoS of the service quality is proposed.Mechanisms and algorithms are being studied and experimented.This thesis proposes a new service sampling method based on probability distribution analysis to improve the information content of sampling data and improve the efficiency of service prediction from the sampling aspect.Secondly,in order to improve the service prediction method and improve the prediction accuracy,a new enhanced algorithm is proposed based on random walk to improve the similarity of neighbors in the service prediction.Finally,based on the detection of the probability distribution of the original and training data,the prediction results of various algorithms are combined to improve the prediction accuracy.Based on the proposed algorithm of service prediction mechanism,this thesis compares the prediction accuracy and operation effect of the same method with the simulation experiment of Web QoS dataset to verify the effectiveness of the method and improve the service prediction mechanism.The experimental results show that the innovation and improvement of related steps can improve the accuracy of the Web service prediction mechanism and improve the quality of the service system.