A new method of QoS prediction based on probabilistic latent feature analysis and cloud similarity
by Weina Lu; Xiaohui Hu; Xiaotao Li; Yuan Wei
International Journal of High Performance Computing and Networking (IJHPCN), Vol. 9, No. 1/2, 2016

Abstract: With the increasing requirements of service mode in cloud computing, predicting accurate quality of service (QoS) is greatly significant in the recommender or composition system to avoid expensive and time-consuming invocations. Unlike previous research approaches which generally stay on the explicit values of QoS data, in this paper, we propose a new prediction method based on probabilistic latent feature analysis and cloud similarity. As user experience quality of service is influenced by the implicit factors, such as network performance, user context and user preference, we first consider these factors as the latent features of user and relate it to the QoS data using pLSA model. Then, the users or services are clustered based on the similar latent features. Finally, after mining the similarity of users in the same cluster by cloud model, the personalised QoS values are predicted by the experience quality of the similar users with the similar services. Experiment results with a real QoS dataset show that the proposed approach can effectively achieve an accurate QoS prediction.

Online publication date: Thu, 11-Feb-2016

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of High Performance Computing and Networking (IJHPCN):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?


Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com