Weigh your Preferences! Towards using hierarchies for building personal libraries

The existence of a vast amount of document collections both on-line or off-line leads to an overwhelming experience when users attempt to query them. This emphasises the need to develop collaborative approaches for acquiring data/knowledge from distributed and semantically heterogeneous sources of knowledge/information. Work on user preferences has recently acquired more attention in order to retrieve and rank documents which are relevant to the user’s interests. In this paper, we present a hierarchical modeling of a user’s interests/preferences. The proposed hierarchical structure results in a better semantic handling of the users’ interests and proper matching of documents over these nodes representing the user preferences. Our preliminary experiments show the merits of such an approach and open new avenues of research in this area