期刊
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
卷 27, 期 18, 页码 5826-5839出版社
WILEY
DOI: 10.1002/cpe.3639
关键词
cloud services; QoS; prediction; collaborative filtering; data smoothing
资金
- Natural Science Foundation of China [61100054, 61174152, 61272063, 61402167]
- Scientific Research Fund of Hunan Provincial Education Department of China [15K043]
- Open Foundation of State key Laboratory of Networking and Switching Technology (Beijing University of Posts and Telecommunications) [SKLNST-2013-1-21]
To assess the quality of services (QoS) in service selection, collaborative service QoS prediction has recently garnered increasing attention. They focus on exploring the historical QoS information generated by interactions between users and services. However, they may suffer from the data sparsity issue because interactions between users and services are usually sparse in real scenarios. They also seldom consider the network environments of users and services, which surely will affect cloud service QoS. To address the data sparsity issue and improve the QoS prediction accuracy, the following paper proposes a collaborative QoS prediction method with location-based data smoothing. The method first computes neighborhoods of users and services based on their locations which provide a basis for data smoothing. It then combines user-based and service-based collaborative filtering techniques to make QoS predictions. Experiments conducted using a real service invocation dataset validate the performance of the proposed QoS prediction method. Copyright (C) 2015 John Wiley & Sons, Ltd.\
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据