The use of collaborative networks services in general, and web based social networks (WBSN) services in particular, is today increasing and, therefore, the protection of the resources shared by network participants
is becoming a crucial need. In a collaborative network, one of the main
parameters on which access control relies is represented by trust and reputation, since access to a resource may or may not be granted on the basis of the trust/reputation of the requesting node. Therefore, the calculation of the trust of the nodes becomes a very important issue, mainly in business to business (BtoB) social networks, where trustworthy nodes can increase their benefits taking profit of their good reputation in the network. In order to address this point, in this paper we propose a mechanism to dynamically compute nodes trust, based on their past behavior. The key
characteristic of our proposal is that trust is computed in a private way.
This is obtained by anonymizing the local log files storing information about nodes actions.