Addressing the knowledge acquisition issue for model based level 2/3 fusion

By observing various battlefield situations as they unfold and by interacting with peers and with battlefield artifacts, such as sensors and information processing systems, intelligence analysts form internal, mental models of things that they are observing and with which they are interacting. These models provide predictive and explanatory power for understanding a specific situation at hand, for which there are no algorithmic solutions. This implies that one needs to capture the mental model of an analyst to "automate" the situation understanding process, which is critical in the presence of huge volumes of data continuously generated by many battlefield sensors. One way to partially address the knowledge acquisition issue is to employ an effective automated means for discovering useful situation information in the form of computational models from the thousands of multi-source events generated every minute in the theatre of operations. We discuss two approaches based on spatiotemporal clustering.