With the increasing number of Web services, recommending and selecting the optimal Web services for consumers has become one of the most important challenges in the field of service computing. The goal of consumers is to discover and use services that lead to their experiencing the highest quality. The quality of service (QoS) performance of Web services is highly related to invocation time since the service status and the network environment change over time. Invoking a huge number of Web services for consumers to predict the quality is time-consuming, resource-consuming, and sometimes even impractical. To address the challenge, this paper proposes a time-aware QoS prediction approach for Web services and designs a prediction framework. In our experiment, we collect QoS information with timestamps from geographically distributed service consumers through the framework. Based on the information, we predict the quality of services; in addition, the relationship between their expectations and the level of the services is considered. As a result, we can obtain a list of recommended services for selection. Finally, the experiment shows that the approach achieves better prediction.
展开▼
