Abstract: Due to the complexity and inaccuracy of user identification and session identification in the traditional Web access pattern mining system, this paper proposes the filter based on Web access pattern mining system, which can identify a user and a session accurately, and provides good data for the mining algorithms. It presents the implementation and deployment of the log filter, and proposes the Web access pattern mining algorithm. The method is widely used in the scientific database.

Key words: Data mining, Web log, Access pattern, Frequent set

摘要： 针对传统Web访问模式挖掘系统中用户识别和会话识别的复杂性和不准确性，该文提出了基于过滤器的Web访问模式挖掘系统。它能够准确地识别用户和会话，为挖掘算法提供优质的数据。给出了日志过滤器的实现和部署，提出了Web访问模式的挖掘算法。目前该方法已经广泛地应用于科学数据库系统中。

关键词: 数据挖掘, Web日志, 访问模式, 频集

CLC Number: 

• TP311.13