X-Tracking the Changes of Web Navigation Patterns

In this paper, we firstly investigate the possible changes on web usage behaviors and then propose an x-tracking method to detect these changes. The changes on web navigation patterns are depicted from microscopic and macroscopic levels: the former is for the "internal" and "external" changes which show the variations on the semantics and external physical features respectively, while the latter is modeled for the changes of the popularity on "local" and "global" time line. The x-tracking method we propose is to detect the newly emerged patterns (EP) based on "internal" feature, which is the premise to compute the changes on other features by tracking their internal unchanged patterns (IUP). Experiments show that the x-tracked changes are condensed, efficient and informative.