Interleaving Semantic Web Reasoning and Service Discovery to Enforce Context-Sensitive Security and Privacy Policies

Abstract : Enforcing rich policies in open environments will increasingly require the ability to dynamically identify external sources of information necessary to enforce different policies (e.g. finding an appropriate source of location information to enforce a location-sensitive access control policy). In this paper, we introduce a semantic web framework and a meta-control model for dynamically interleaving policy reasoning and external service discovery and access. Within this framework, external sources of information are wrapped as web services with rich semantic profiles allowing for the dynamic discovery and comparison of relevant sources of information. Each entity (e.g. user, sensor, application, or organization) relies on one or more Policy Enforcing Agents responsible for enforcing relevant privacy and security policies in response to incoming requests. These agents implement meta-control strategies to dynamically interleave semantic web reasoning and service discovery and access. The paper also presents preliminary empirical results. This research has been conducted in the context of myCampus, a pervasive computing environment aimed at enhancing everyday campus life at Carnegie Mellon University. The framework presented can be extended to a range of other applications requiring the enforcement of context-sensitive policies (e.g. virtual enterprises, coalition forces, homeland security, etc.).