期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2013
卷号：55
期号：2
出版社：Journal of Theoretical and Applied
摘要：High bandwidth utilization, reduced load on the origin server, high access speed are possible by combining Web caching and pre-fetching techniques. Pre-fetching is the process of fetching few Web pages in advance which will be assumed to be needed by the user in near future and those pages are cached in the memory. Lots of work has been reported for caching and pre-fetching of Web pages in the literature. In this paper, pre-fetching using clustering technique is combined with SVM (Support Vector Machine) � LFU algorithm, a machine learning technique for Web proxy caching .By using real dataset it will be shown that the SVM technique will be better than clustering based pre-fetching technique using caching policy like LFU considering bandwidth utilization and access latency.
关键词：Classification; Least Frequently Used (LFU); pre-fetching; Support vector machine (SVM)
