Discovery of popular structural properties in a website for personalization and adaptation

The massive growth in the size and complexity of websites, lead to increased demand on personalization systems and tools which can help in providing users with what they want or need without them having to ask for it explicitly. In this paper, we present a novel approach towards the discovery of target pages for shortcuts. The approach is based on the Maximal Forward Reference algorithm. Few changes to this algorithm are suggested to make it more suitable for the discovery of popular paths, pages and individual user behaviors in relation to the structural design of the sit. The major impetus for the selection of Maximal Forward Reference approach in our research was driven by two of our own convictions. First, forward traversals more realistically represent the navigational intentions of the user. Second, the algorithm has already been proven to generate a complete set of maximum references from the processed log file. The proposed approach aims at limiting the consolidation process of the MFR to the level of individual users which should help in providing more detailed site adaptation, personalization, and visualization on the user level.