A New Evolutionary Coclustering Approach for Web User Profiling

Coclustering is a two-way clustering approach involving simultaneous clustering along two dimensions of the data matrix. Extraction of coclusters comprises of web objects (i.e., web users and web pages) is an emerging research topic in the context of web usage mining. It overcomes some of the problems associated with traditional clustering methods by allowing automatic discovery of browsing pattern based on a subset of attributes. A correlated cocluster of clickstream data is a local pattern such that users in cocluster exhibit correlated browsing pattern through a subset of pages of a web site. This study aims to provide a new correlated coclustering framework using genetic algorithm to identify overlapping correlated cocluster from clickstream data. Experiment is conducted on the benchmark dataset. Results demonstrate the efficiency and beneficial outcome of the proposed method by correlating the users and pages of a web site in high degree. It also outperforms the existing traditional clustering of web usage data.