Using SOM to clustering of web sessions extracted by techniques of web usage mining
IDEAL'12 Proceedings of the 13th international conference on Intelligent Data Engineering and Automated Learning
| Hi-index | 0.00 |
Web log mining is the application of data miningtechniques to web log data repositories, in which clusteringanalysis is one of important Web usage mining techniques.Recently various clustering approaches have been developedfor web pages and customer segments clustering. However,most of them take user access frequency or web page timeduration as measurement of user navigation interest withouttaking into consideration such important factors as userpreference, browsing context, etc. A novel clustering methodbased on fuzzy multisets is proposed to deal with the problemsin this paper. In proposed method, the fuzzy multiset isadopted to characterize user’s navigation behavior and toconstruct a multi fuzzy similar matrix to represent similaritybetween different users’ browsing behavior, which can reflectwholly the interest of web user with the web page-click rate,web page viewing time, user's preference and so on. And webpage clusters and customer segments are abstracted directlyfrom the corresponding multi fuzzy similar matrix. Anillustrative example is given to show how the algorithms work.