The Research About Recommendation Algorithm Of Web Services Based On QoS Historical Record

Web services as a new Web application model develop very rapidly recent years. And with the rapid development of the Internet and related technologies, the type and the number of Web services is growing quickly. More and more Web services provide the same or similar functionality, which leads to difficult for users to quickly select their own services needed; At the same time, for the service provider concerned, there have been problems of information overload, which leads consumers to continue to drain. Above all. this makes how to improve the quality of Web services recommendation is a key issue in related fields.QoS (Quality of Service) as an important factor in determining the success of the service provider, which describes the non-functional properties of the service, is recommended for successful Web services critical factor. Currently, the study of Web services recommendation based on QoS is limited, and the existing research methods assume a static QoS Web services known. But. most services QoS attribute values are changing, and in close contact with the status of the service is running. Therefore, the existed studies still exist the problem of low recommendation accuracy.In this paper, in the background of Web network environment, we research traditional Web services based on QoS for the use of assumed QoS values and all of the user’s QoS historical record, leading to the results of Web service recommendation accuracy is not high and low efficiency.Firstly, we introduce the concept of user interest degree, and divide the old users into a cluster by Euclidean distance, whose interest degree is similar with the new user. Then, use the QoS historical record of users in the same cluster to determine various QoS attribute values for the same Web service. The QoS attributes as arguments, at the same time, the corresponding quality of service QoS values as the dependent variable. We use multiple linear regression to solve each Web service QoS attributes weights predictive value, and then get the QoS prediction function of each Web service, whereby recommends Web services for new users. The recommendation algorithm of Web services based on QoS historical record(WSRAHQ). which is described in this paper, gives a Web service recommendation process. Finally, the simulation demonstrates the effectiveness of the algorithm in improving the accuracy of the recommendation and reducing the uptime of recommendation.The innovation of this paper are two points, one is the use of the QoS historical record, these values reflect the real Web service operation, which helps to improve the accuracy of the recommended service; the second is divide users with similar interest into a cluster, no longer use QoS historical records of all users during Web services recommendation, which reduces the computational complexity and greatly improves the efficiency of recommendation.The research on Web service recommendation has become a hot issue in service recommendation region. In the Web environment, an effective service recommendation algorithm can consider service quality and user preferences to ensure user to select the satisfactory service. Ensure the quality of service for the user, at the same time, meet their individual needs, has a certain theoretical significance and practical value.