Modelling geographical effect of user neighbourhood on collaborative web service QoS prediction
by Zhen Chen; Limin Shen; Dianlong You; Huihui Jia; Chuan Ma; Feng Li
International Journal of High Performance Computing and Networking (IJHPCN), Vol. 14, No. 2, 2019

Abstract: QoS prediction is a task to predict the unknown QoS value of an active user to a web service that he/she has not accessed previously for supporting appropriate web service recommendation. Existing studies adopt collaborative filtering methods for QoS prediction, while the inherent issues of data sparsity and cold-start in collaborative filtering have not been resolved satisfactorily and the role of geographical context is also underestimated. Through data analysis on a public real-world dataset, we observe that there exists a positive correlation between a user's QoS values and geographical neighbourhood's ratings. Based on the observation, we model the geographical effect of user neighbourhood on QoS prediction and propose a unified matrix factorisation model by capitalising the advantages of geographical neighbourhood and latent factor approaches. Experimental results exhibit the significance of geographical context on modelling user features and demonstrate the feasibility and effectiveness of our approach on improving QoS prediction performance.

Online publication date: Tue, 30-Jul-2019

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