The exploitation of trust in virtual environments, on-line services, intelligent pervasive environments and many other scenarios still deserve researchers’ interest. In this work we investigate on how to assess the local (direct) trust a node receives from each of his neighbors. Our proposal is social-based and takes into account both positive and negative experiences as well as the history of past feedbacks and their aging, ensuring good stability also when a node receives hundreds of positive feedbacks briefly followed by few negative feedbacks. Our results are illustrated by performing several simulations on different networks.