Collaborative Prediction Of QoS And Web Service Recommendation Considering The Similarity Ratio

With the rapid development of Internet technology, the number of Web services increases rapidly, Web services with the same or similar functions are also increasing. Help the user to select a service with a certain function, we need take into account the non functional attributes of the service, namely of service quality(QoS), such as response time, availability, throughput, etc. Therefore, it has become an important research hotspot in the field of service computing to recommend and select the best Web service for the user. However, due to the increase of the candidate services, the incomplete degree of QoS information is becoming more and more serious. Therefore, to study the prediction of QoS is of great significance to the service recommendation system considering the quality of service quality. In this paper, we study the problem of QoS prediction based on collaborative filtering and the recommendation of Web service.Firstly, in the process of collaborative filtering approach, the influence on similarity calculation of the relationship between the number of common calls and the number of services that each called is taken into consideration. By using the minimum common call number threshold, the standard for the calculation of the similarity of different data sets is adjusted. In the selection of similar neighbor, the traditional Top-K algorithm is replaced by the minimum similarity threshold.Secondly, to deal with the prediction error caused by ignoring the ratio relation between Web services QoS in the method of collaborative prediction approaches used the PCC, a prediction approach considering the similarity ratio SRPre was proposed. Based on the historical QoS data, the combine of user-based approach and item-based approach was used. Computing the similarity ratio according to the QoS average of similar neighbors was proposed. The similarity ratio was added to the process of QoS collaborative prediction. At the same time, the SRPre service recommendation platform is proposed. The difficulty of collecting QoS data is resolved by the collaboration of collector and data center. The recommendation module is implemented by SRPre method.Finally, according to the proposed SRPre service recommendation platform and the experimental needs, the experimental platform is established. The SRPre method is compared with the other prediction methods on the real data set and the experiments and analysis were taken.