一个简单的Web日志挖掘系统  

A Simple Web Log Mining System

作　　者：杨怡玲[1] 管旭东[1] 陆丽娜[2] 尤晋元[1] 

机构地区：[1]上海交通大学计算机科学与工程系,上海200030 [2]西安交通大学计算机科学与工程系,西安710049

出　　处：《上海交通大学学报》2000年第7期932-935,共4页Journal of Shanghai Jiaotong University

摘　　要：在分析 Web日志挖掘的困难及对策的基础上 ,给出了一个简单的 Web日志挖掘系统( SWLMS)的体系结构 .具体介绍了 SWLMS中日志的预处理过程 ,包括数据净化、用户识别、会话识别、路径补充的主要任务及其实现 ,并着重介绍了预处理之后的序列模式识别过程和算法 ,包括最大向前路径的识别和频繁遍历路径的发现 。This paper mainly discussed Web log mining, the application of date mining to log data generated by Web servers, which could assist the webmaster to optimize site architecture and increase visiting efficiency. Based on the analysis of difficulties and the corresponding solutions of Web log mining, the architecture of SWLMS, our sample Web log mining system was addressed. The data preprocessing phase in SWMLS, including data cleaning, user recognition, session identification and path filling was discussed in detail. Then, the sequential pattern recognition phase and its algorithms were presented, including the recognition of maximum forward paths and frequent traversal paths, with some experimental results presented.

关 键 词：数据挖掘 WEB日志挖掘 序列模式识别 SWLMS 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]