Evaluating Confidence in Context for Context-Aware Security

We propose a software framework that augments context data with a range of assorted confidence/reputation metadata for dimensions such as security, privacy, safety, reliability, or precision, defined according to a generic context confidence ontology. These metadata are processed through the network of federated distributed software services that support the acquisition, aggregation/fusion and interpretation of context, up to its exploitation by context-aware applications. This solution for qualifying and gauging context data makes possible its use in more critical applications of context awareness, such as adaptation of security mechanisms. We show how to implement with our framework a quality-critical application like contextual adaptation of security services, where security is tailored to the protection requirements of the current situation as captured by relevant context data.

• Type: Conference
• Year: 2007
• Where: EUSAI
• Authors: Marc Lacoste, Gilles Privat, Fano Ramparany
• Keywords: Artificial Intelligence, Assorted Confidence/reputation Metadata, Context Confidence Ontology, Context Data, EUSAI 2007