Clustering web users based on browsing behavior

Abstract

It is critical to acquire web users behavior model in E-commerce community. In this paper, we propose to train the web user's browsing behavior, and clustering web users based on their browsing behavior. In particular, our method uses page-content information extracted from the user's click stream, then trains a behavior model that describes how web user locates useful information in the Internet. The classifier is trained on the data which describes how the user treats the information that she has visited, that is, browsing behavior. We find that some user groups based on the browsing behavior can get much higher accuracy for prediction. © 2010 Springer-Verlag.