Research On Web Service Recommendation Method And Application Integrating Common Invocation And Data Fluctuation

With the continuous innovation and development of Internet technology,Web service recommendation and selection has gradually become an important research content of common concern in industry and academia.Quality of Service(QoS)is the key factor affecting Web service recommendation.However,there are some problems in the existing theory and application of Web service recommendation based on QoS.In order to solve these problems,the following work has been completed.Firstly,the advantages and disadvantages of different similarity calculation methods and the different data environments based on users' collaborative filtering algorithm are analyzed and summarized.A similarity calculation method based on collaborative filtering is proposed,which integrates common invocation and data fluctuation.This method takes into account the factors of common invocation and data fluctuation,and quantifies them into Pearson similarity calculation method,which improves the accuracy of similarity calculation.Furthermore,it improves the accuracy of QoS prediction.Secondly,user-based and service-based QoS prediction methods are taken into account in the process of QoS prediction.Self-confidence is decomposed by similarity degree of similar neighbors,and the weight of prediction results is balanced by human participation.A hybrid collaborative prediction of user-based and service-based QoS is carried out,which improves the comprehensiveness and accuracy of the prediction.Thirdly,a platform of omni-directional Web service research and recommendation based on QoS is designed and implemented hierarchically.The platform covers a variety of user roles and many functional models in Web service research and recommendation.It provides a better Web service recommendation for ordinary users and a good research environment for researchers in this field.Finally,several examples are given to demonstrate and analyze the application of the core model in the platform of omni-directional Web service research and recommendation based on QoS.By applying the corresponding functions of the core model of the platform in real environment,the good practicability and applicability of each core model in the platform of omni-directional Web service research and recommendation based on QoS are verified.