Enrichissement du profil utilisateur à partir de son réseau social dans un contexte dynamique : application d'une méthode de pondération temporelle

The development of users’ profiles is central for adaptive system to propose information that corresponds to user specific needs. Online social networks provide valuable information sources to collect and identify user’s information and user’s interests. This work focuses on extracting user’s interests from his social egocentric network to build the social dimension of his user profile. To improve this social dimension, we take into account the dynamic of the social network by applying a time-weight strategy in order to drift the user interests. The time-weight of an interest is computed by combining the accuracy of individuals in his egocentric network (computed by taking into account the freshness of their ties) with the information accuracy (computed by taking into account its freshness). The experiments on scientific publications networks (DBLP/Mendeley) allowed us to demonstrate the effectiveness of our proposition comparing to the time-agnostic user profiling approach. MOTS-CLES : Profil utilisateur, Réseau social, Réseau égocentrique, Pondération temporelle.