Improving dynamic skills searching in virtual social communities using agent’s network self-organization

Our paper will focus on the use of a multi-agent approach for the introduction of a member in a social network. As for MultiAgents Systems, a global view of a social network is not available and furthermore undesirable. To join a network, each member must build a local network representation according to its needs. Our contribution focuses on the detection of thematic areas to assist the introduction of a new user in the social network. To do so, we must define and identify the parts of the system (agents, groups of agents) that represent the interesting nodes for this user. Our approach is based on a decentralized interrogation of the network and on organizational structures detection.