System Theoretic Formalization of Social Group Processes and Optimal Evidence-Indicator Discovery

Modelers using varying levels of human driven event domain expertise and computational techniques have worked to expand computational sciences into human driven event analysis. While computational modeling efforts continue to draw new insight from statistical decision theory, control systems and machine learning, a large gap remains between what we have seen to be mostly academic endeavors and practical end-user systems. Here we describe a formal framework for treating a class of computational social scientific problems and describe an end-user driven system developed based on a two part segmentation of the social scientific evidence acquisition and modeling stages. The system is conducive to validation of results against real world data with intuitive handles for decision makers from nonacademic world to respond to emerging issues in their own processes. Furthermore, the formal framework gives access to analytical handles needed for system enhancement via integration of advanced statistical inference and reasoning concepts.