Abstract: Data farming is a process that integrates design of experiments, high-performance computing, distillation model development, rapid prototyping of scenarios, and collaborative processes to automate and optimize the creation and execution of software-based experiments over a very large range of input factors and to extensively explore and gain insights from the experimental outcome space. Data farming is a cross-discipline capability that has been used, in particular, to address complex systems due to its potential for efficient examination of multi-dimensional input factor spaces. In physical complex systems, factors such as chaos effects, non-linearity, the non-closed nature of real-world systems, issues of availability and resolution of initial and ongoing state data, and limitations of resources prevent confidence in prediction from such experimentation. When systems involve social entities they become yet more complex with intangibles such as morale, trust and charisma coming into play along with adaptation and co-evolution. Social network analysis and graph theory are widely used techniques for the evaluation of social systems. It is proposed that the incorporation of these techniques into the data farming process will provide analysts examining complex systems with a powerful new suite of tools for more fully exploring and understanding the effect of interactions in complex systems. Further, this essential integration will provide modelers with the capability to gain insight into the effect of network attributes on the breadth of the model outcome space and the effect of model inputs on the resultant network statistics. The objectives of this study were to: examine existing data farming capabilities as well as a set of illustrative multi-discipline experiments to define a generalized data architecture; build tools to extract emergent network data structures from modeling systems; develop and integrate network analysis capabilities into existing data farming data infrastructure; and demonstrate the utility of these capabilities by applying the tools to a problems associated with the counter-improvised explosive device battle. This new integration of social network analysis techniques into data farming systems will be outlined in this thesis. The result of applying these capabilities to abstracted models of insurgency recruitment, clique creation and evolution, and initial population variation as well as an initial examination of the impact of female engagement teams in an Afghan village scenario are presented.
Keywords: Applied sciences, Data farming, Social networks