Web User Action Mining Based on Fuzzy Clustering
An approach for mining a collection of user access paths with a new fuzzy clustering algorithm is introduced to discover clusters of similar paths. In this paper a new approach of computing similarity coefficient about two user access paths is proposed, and it is obtained to represent the similarity of the user access patterns. Then the paper presents a new clustering algorithm-transmission shut package fuzzy clustering algorithm in order to capture user groups which have the common access pattern by mining web usage data. The improved algorithm is superior to the general transmission shut package algorithm in saving memory space, and improving clustering speed through reducing the complex operands. Finally through performing web usage mining on web logs, the user groups with common access patterns are discovered. The experimental results indicate that the techniques discussed here are promising, and bear further investigation and development.
data mining web usage mining fuzzy clustering web log
Dajiang Lei Haidong Fu
School of Computer Science and Technology, Wuhan University of Science and Technology, China
国际会议
2006现代科技国际研讨会(The International Workshop on Modern Science and Technology in 2006)
北京
英文
276-281
2006-04-01（万方平台首次上网日期，不代表论文的发表时间）