The dynamics of information fusion: synthetic versus misassociation

Abstract

Fusion entails both the association and synthesis of information. If missassociations occur, they obviously undermine the gains won by synthesis, compromising the fusion product. An analytic framework is presented here to study the competition between the negative effect of missassociation

and the positive effect of synthesis, to demonstrate

and analyze their interplay quantitatively. Here the quality of information being enhanced or degraded is taken to be the extent to which the information correctly determines a decision or action inference. To say that the uncertainty injected by missassociation may overwhelm the uncertainty reduction won

by synthesis, for instance, would mean that this inference-determining quality of information falls

in fusion below that of the best information source working independently. This is ultimately a study in uncertainty dynamics: the beneficial reduction of uncertainty by synthesis in fusion versus the detrimental increase of uncertainty due to association, which are both always present in fusion to some degree.