Research on Service Recommendation Method Based on Cloud Model Time Series Analysis
摘要：The problem of information overload is becoming increasingly prominent, and recommendation systems are developing rapidly in various fields. How to find the most user-friendly services has become the focus. Service recommendation based on QoS is an important technology to select appropriate services for users. In this paper, a service selection method based on time series analysis of cloud model is proposed. Firstly, the noise was removed by clustering algorithm, clustering was divided, and similar user sets were obtained.Then, the cloud model was established by using similar user history data in different periods, and the comprehensive cloud model was obtained by combining time decay function. Finally, the recommended service was obtained by comparing TOPSIS method with ideal cloud model. The experimental results on WS-Dream dataset show that the accuracy of recommendation is improved compared with the existing recommendation algorithms.