Modeling Of Recommender Systems Through Resource Description Framework

Faced with problems of informational over-load on a dynamic, distributed and heterogeneous web, current research aims to design and develop recom-mender systems that are mainly based on techniques of information filtering. In this paper, we propose a hybrid modeling of recommendation systems by formalizes re-sources description framework (RDF), while based on the integration of elements of the Dublin Core (DC) de-scribing resources and the vocabulary Friend of A Friend (FOAF) describing the users. A hybridization procedure was introduced into the function of similarity calculation. The empirical tests on various real data sets (Book-Crossing, FoafPub) showed satisfactory perfor-mances in relevance and precision.