Developing means of secure and trustful communications for Vehicular Social Networks (VSN) is essential to enable active information sharing among vehicles. Reputation-based trust management system is a popular security mechanism used in vehicular networks and proven to be robust in many other applications. In this paper, we propose a similarity-based bootstrapping method using analytic hierarchy process (AHP) for trust management in VSN. Then we introduce a reputation mechanism that takes into account, the user behaviors as well as historic features such as the total distance driven. Our simulation results show that the proposed system is robust against high density of malicious nodes (up to 60%). Furthermore, we also prove the accuracy of the proposed system in different conditions.