采用数据挖掘技术中的关联分析和聚类方法,重点研究Web日志兴趣发现的理论和方法,指出普通日志记录方法的局限性,提出过滤用户偏好的定制Web日志方法,实验结果验证通过该方法采集的数据,可以发现隐藏在日志数据中的关联规则,同时找到相似用户的兴趣和偏好,并且能够提高过滤用户兴趣偏好的精度。 By using the method of association analysis and clustering in data mining, this paper focuses on the theories and methods of Web log interest discovery, pointing out the limitations of ordinary log recording methods, and proposes a customized web log method of filtering user preferences. The experimental results are verified by this method The data can be found hidden in the log data association rules, similar users find the same interests and preferences, and can improve the accuracy of filtering user interest preferences.