一种新的Web频繁访问模式挖掘算法

A new algorithm of Web Frequent Access Patterns Mining

提出了一种基于有向图的从Web日志中挖掘用户频繁访问模式的新算法,与传统使用基于关联规则挖掘的序列模式挖掘技术相比,本算法采用有向图来记录Web访问序列和它的计数,在挖掘过程中只需要扫描数据库一次,不产生数量庞大的候选模式,即可直接挖掘出所有的Web频繁访问路径,大大提高了Web访问模式的发现效率。

In this article, It propose a algorithm of Frequent Access Patterns Mining from Web logs Based on Directed Graph. Compare with the previous sequential patterns mining algorithm based on Association Rule, the algorithm makes Directed Graph to register Web access sequence and corresponding counts, and the algorithm can min all Web Frequent Access Patterns through scanning database only once, does not create candidate patterns during the mining, The efficiency is distinctly improved in discovers Web Access Patterns.