基于主观贝叶斯方法的数据预取技术

Data prefechting application research

随着信息技术的迅速发展,网络规模随着用户数呈指数级增长,也越来越庞大,要提高用户的访问速度,主要采用缓存和预取技术来减少网络延迟,提出了一种智能的预取方案。该方案使用模糊匹配来计算用户对页面的访问概率,用主观贝叶斯方法计算网页的重要性,同时控制预取的量和预取的时刻,以避免对网络的性能产生负面影响。

Currently, caching and prefetching are two common ways to decrease network delay. In this paper, a new and intelligent prefetching algorithm was designed. It computed the access probability of some Web pages for certain clients by fuzzy matching. Subjective Bayes theory was used to compute the importance of the Web pages. To avoid negative effect on network＇s performance, it must control how many Web pages one time and when to prefetch.