Behavior instance extraction for risk aware control in mission centric systems

摘要

In the pursuit of behavior modeling for dynamic policy management and usage control, mission centric systems are the most important to examine, not only because situation dynamics can dramatically alter the collaboration environment, but also since behavior in these systems is constrained by mission objectives or workflows. Traditional mission decomposition into tasks and objectives has led to static policy deployment and role based views of access control and, but we propose Risk-Adaptive Mission Policy (RAMP) to enable commanders to automate decisions to constrain or proliferate access depending on behavioral context and risk assessment for successful mission outcomes. As the critical source of knowledge to enable this new methodology, we formulate the behavior instance extraction problem (BIEP) to pre-process unstructured activity data and mine interesting behaviors for modeling applications. Finally, we develop the workflow behavior instance extraction algorithm to tailor a solution to BIEP specifically for RAMP. Our evaluation weighs performance against new functionality to show that our work supports risk aware security decisions for dynamic situation management in mission centric environments.