Efficient Construction of Dynamic Meta-Networks

Our objective is to enable military analysts to more rapidly produce intelligence results rearding adversarial groups and also diverse types of covert networks, and to improve the quality of the underlying analyses. We focus, initially, on the visual analytics and decision aids needed to improve multi-source data collection, target identification, and impact analysis and to reason about and stay attuned to the level and source of uncertainty and so its impact on analysis, and to assess the incremental value of new intelligence. Our approach utilizes Bayesian analysis for uncertainty assessment, machine learning and statistical techniques for coding and analyzing data, dynamic network agent based models and statistical reasoning to identify points where additional data collection are needed. We leverage the existing AutoMap and *ORA, and supplement these with GUI"s designed for the warfighter and military analyst and based on workflows identified by working directly with military intelligence officers. To reduce time from data collection to decision we employ mission driven model refinement, uncertainty tracking, DIME/PMESII and adversarial ontologies, visualization and menu driven data entry, and make use of novel data collection and information collection technology such as crowd sourcing and cell-phones.