Abstract
In a highly dynamic and decentralized environment, where data are uncertain, Trust has become a key factor in the process of decision making. Trust-based recommendation is based on Trust between users. It was the main subject of several studies such as: Haydar (2014), Simon et al (2012), Fabiana et al (2011), Golbeck (2005), Josang and Pope (2005). In fact, for relevant recommendation, it is very important to define the adequate techniques for modeling and evaluating trust between agents. In this paper, we give a state of the art of modeling trust in recommender systems. Furthermore, we make a comparative study between various existing methods.