基于用户访问树的分布式Web日志挖掘算法

Distributed Web Log Mining Algorithm based on User Access Tree

为了提高对分布式Web日志数据的准确挖掘能力,提出基于用户访问树的分布式Web日志挖掘算法。构建分布式Web日志的信息分布式检测模型,采用模糊信息粗糙集调度方法进行分布式Web日志信息的结构重组,提取分布式Web日志的统计特征量,采用用户访问树特征聚类方法进行分布式Web日志数据的空间分布式重组,结合粗糙集特征匹配方法进行分布式Web日志的离散融合处理,对多层分布式数据库中的主成分特征分量进行关联规则融合,结合信息融合结果进行分布式Web日志数据的特征参量聚集式调度,提取分布式Web日志的谱特征分量,采用空间信息聚类方法,实现分布式Web日志的用户访问树模型构造,结合决策树模型构建分布式Web日志挖掘的适应度参数,实现分布式Web日志挖掘。仿真结果表明,采用该方法进行分布式Web日志挖掘的准确性较高,抗干扰性较好,提高了分布式Web日志挖掘和用户信息访问能力。

In order to improve the accurate mining ability of distributed Web log data,a distributed Web log mining algorithm based on user access tree was proposed.An information distributed detection model of distributed Web logs was constructed,a fuzzy information rough set scheduling method was adopted to carry out structural reorganization of distributed Web log information,statistical feature quantities of distributed Web logs were extracted,a user access tree feature clustering method was adopted to carry out spatial distributed reorganization of distributed Web log data,a rough set feature matching method was combined to carry out discrete fusion processing of distributed Web logs,association rule fusion was carried out on the principal component characteristic components in the multi-layer distributed database,the characteristic parameter aggregation scheduling of the distributed Web log data was carried out in combination with the information fusion result,the spectral characteristic components of the distributed Web log were extracted,the spatial information clustering method was adopted to realize the construction of the user access tree model of the distributed Web log,and the fitness parameter of the distributed Web log mining was constructed in combination with the decision tree model to realize the distributed Web log mining.The simulation results show that the method has higher accuracy and better anti-interference performance in distributed Web log mining,and improves the ability of distributed Web log mining and user information access.