Maximal paths recipe for constructing Web user sessions

This paper introduces a new method for the session construction problem, which is the first main step of the Web usage mining process. The proposed method defines user sessions as a set of navigation paths in the Web graph and produces a complete set of all possible maximal paths. Our new method is capable of generating navigation paths which cannot be extracted by using previous greedy approaches. Through experiments performed on real data, it is shown that when our new technique is used, it outperforms previous approaches in Web usage mining applications such as next-page prediction. Our analysis on Web user sessions exposes an important observation: Web users sessions contain navigation graphs that has small number of nodes where users branch out their navigation into multiple paths.