Adaptive Information Fusion in Asymmetric Sensemaking Environment
Paul Munya, Celestine A. Ntuen · 2007
The existing sensemaking models for traditional force-on-force battlefield information management rarely survive the kinds of information in asymmetric battlespace environments. By combining the abduction process and Bayesian probability network formalisms, we propose a Bayesian Abduction Models (BAM) to support in the sensemaking process of evaluating multiple hypotheses in the context of changing information. This paper describes a Bayesian network that captures abduction logic primitives from a kernel of disparate information sources. We use a genetic learning algorithm to solve BAM information fusion problems. We show how the model can be used in prospective and retrospective sensemaking conditions to simulate the ways commanders and the battle staffs process information.