Web日志挖掘中数据预处理方法的研究

Data Preprocessing Method Research for Web Log Mining

Web日志挖掘是目前网上智能信息检索和电子商务的主要研究课题之一。而数据预处理在Web日志挖掘中起着很重要的作用,直接影响日志挖掘的质量和结果。介绍了Web日志挖掘数据预处理过程,综述了国际上的研究现状,及流行的处理方法。针对预处理步骤中的用户会话识别和路径填充进行了相应的改进。根据评估会话构造方法的标准,通过实验对给出的新方法与其他方法进行了分析比较。

Web log mining is one of the main research domains in intelligent information retrieval system and electronic commerce. While data preprocessing has played an important part in Web log mining, directly influenced the quality of the Web log mining and its results. This paper introduces Web log mining data preprocessing process. Summarizes studies status and prevalent methods did in this area. Then improves the user session identification and path identification which are the processes of the data preprocessing. Finally according to measures for session construction methods, compares the performance of the new method to other session construction methods by means of experimental data.