A model-based approach to information fusion

Military situations are typically highly complex, rapidly changing and involve adversaries who endeavour to make it appear more complicated and dynamic to ultimately confuse our understanding of it. Furthermore, crucial evidence may be missing and that which is available can be sparse, incomplete, irrelevant, incorrect, imprecise or vague, late by some hours, and it can be subjective. The paper describes some further work on our model-based method to fuse that kind of information. This method supports model-based reasoning, which enables complex hypotheses to be developed from small amounts of information of poor quality and to make predictions about possible future situations.