Ontology and rule based knowledge representation for situation management and decision support

Supporting human decision making in tasks such as disaster planning and time-sensitive targeting is challenging because of the breadth and depth of knowledge that goes into the decision making process and the need to reason with this knowledge within tight time constraints. Ontologies are well suited for representing the concepts that humans use in describing the domain of interest. However, ontologies can be costly to develop and ,by themselves, are inadequate to capture the kinds of decision making knowledge that arise in practice―for instance, those that refer to multiple ontologies or to established precedent. Such decision making knowledge can be represented by using a knowledge representation formalism that we call decision rules. These decision rules are similar to the rules used in rule based systems but can (a) include primitives from multiple ontologies and primitives that are defined by algorithms that run outside of the rule framework (b) be time dependent and (c) incorporate default assumptions. We report on our ongoing experience in using such a combination of ontologies and decision rules in building a decision support application for time sensitive targeting.