Ontology Driven Community Access Control

In this paper we present RelBAC (for Relation Based Access Control), a model and a logic for access control which models communities, possibly nested, and resources, possibly organized inside complex file systems, as lightweight ontologies, and permissions as relations between subjects and objects. RelBAC allows us to represent expressive access control rules beyond the current state of the art, and to deal with the strong dynamics of subjects, objects and permissions which arise in Web 2.0 applications (e.g. social networks). Finally, as shown in the paper, using RelBAC, it becomes possible to reason about access control policies and, in particular to compute candidate permissions by matching subject ontologies (representing their interests) with resource ontologies (describing their characteristics).