Data-Driven Model-Based Decision-Making - Otober 2015

Public Audience
Purpose: To highlight project progress. Information is generally at a higher level which is accessible to the interested public. All information contained in the report (regions 1-3) is a Government Deliverable/CDRL.

PI(s): William H. Sanders, Masooda Bashir, David Nicol, and Aad Van Moorsel*

Co-PI(s): Ken Keefe, Mohamad Noureddine, Charles Morriset* and Rob Cain* (*Newcastle University, UK)

HARD PROBLEM(S) ADDRESSED
This refers to Hard Problems, released November 2012.

• Predictive Security Metrics - System security analysis requires a holistic approach that considers the behavior of non-human subsystem, bad actors or adversaries, and expected human participants such as users and system administrators. We are developing the HITOP modeling formalism to formally describe the behavior of human participants and how their decisions affect overall system performance and security. With this modeling methodology and the tool support we are developing, we will produce quantitative security metrics for cyber-human systems.
• Human Behavior - Modeling and evaluating human behavior is challenging, but it is an imperative component in security analysis. Stochastic modeling serves as a good approximation of human behavior, but we intend to do more with the HITOP method, which considers a task based process modeling language that evaluates a human's opportunity, willingness, and capability to perform individual tasks in their daily behavior. Partnered with an effective data collection strategy to validate model parameters, we are working to provide a sound model of human behavior.

PUBLICATIONS
Papers published in this quarter as a result of this research. Include title, author(s), venue published/presented, and a short description or abstract. Identify which hard problem(s) the publication addressed. Papers that have not yet been published should be reported in region 2 below.

[1] "Impact of Policy Design on Workflow Resiliency Computation Time", presented at Quantitative Evaluation of Systems (QEST) Sept. 1-3, 2015.

[2] "Resiliency Variance in Workflows with Choice", presented at the International Workshop on Software Engineering for Resilient Systems (SERENE) Sept. 7-8, 2015.

ACCOMPLISHMENT HIGHLIGHTS

In this quarter, we are currently in the process of writing a paper that summarizes our findings for the past year. In the summer of 2015, we were joined by an undergraduate intern that performed a literature review on the subject of human behavior in cyber-security from a psychological perspective. Such an approach will help enrich our work with basic social and cultural theories that we used to enrich the presentation and the discussion of our case study results.

We have also continued to explore the provision of quantitative measures predicting the resiliency of workflows whose tasks are assigned to users who may become unavailable at runtime whilst satisfying security constraints. Workflow resiliency measures are useful as they indicate the likely risk of having to violate security constraints in order to reassign tasks and complete a workflow. Previous work has formalized assigning workflow tasks optimally as a decision problem meaning the resiliency of a workflow can be computed as its probabilistic satisfiability using probabilistic models, e.g., Markov Decision Processes (MDPs), which in this case are highly data dependent in terms of user availability predictions. Workflow resiliency may need to be computed at runtime meaning the collection of user availability data to input into the model can itself impact the completion time of a workflow. This makes computing resiliency a good candidate to explore optimal data collection strategies and form the basis of strategy tool support. Details of the early stages of a design of data collection strategy tool support for Mobius are given below.