Bioevent detection via fusion of disparate data

The biosurveillance domain presents a deluge of disparate data overloading analysts to the point where critical information can be missed. The abundance of data heightens the appeal of information fusion solutions. In general, sensor fusion approaches can be viewed as bottom-up and driven by technical capabilities. Often neglected are the information needs that can help human operators optimize performance. We assert improved fusion techniques can arise by implementing top-down processes derived from operator information needs, task demands, and work environment context. Numerous cognitive engineering methods are available to construct such processes. Thus we posit a bidirectional approach is needed to achieve higher-level fusion and support the human operator with decision tasks.