Uncertainty-Sensitive Heterogeneous Information Fusion: Assessing Threat with Soft, Uncertain, and Conflicting Evidence

Abstract : Military and other government organizations put substantial effort into detecting and thwarting attacks such as those by suicide bombers or involving improvised explosive devices. Such attacks may be against military or government installations in the United States or abroad, civilian infrastructure, or any of many other targets. An element of thwarting such attacks is observing suspicious individuals over time with such diverse means as cameras, scanners, and other devices; travel records; behavioral observations; and intelligence sources. Such observations provide data that are often both complex and softi.e. qualitative, subjective, fuzzy, or ambiguousand also contradictory or even deceptive (as when humans lie). The problem, then, is how to fuse the heterogeneous data. This report summarizes our research on heterogeneous fusion methods. The context is military and civilian counterterrorism,