A Formal Context Representation Framework for Network-Enabled Cognition

Network-accessible resources are inherently contextual with respect to the specific situations (e.g., location and default assumptions) in which they are used. Therefore, the explicit conceptualization and representation of contexts is required to address a number of problems in Network- Enabled Cognition (NEC). We propose a context representation framework to address the computational specification of contexts. Our focus is on developing a formal model of context for the unambiguous and effective delivery of data and knowledge, in particular, for enabling forms of automated inference that address contextual differences between agents in a distributed network environment. We identify several components for the conceptualization of contexts within the context representation framework. These include jurisdictions (which can be used to interpret contextual data), semantic assumptions (which highlight the meaning of data), provenance information and inter-context relationships. Finally, we demonstrate the application of the context representation framework in a collaborative military coalition planning scenario. We show how the framework can be used to support the representation of plan-relevant contextual information.