Consequential intelligent agent systems

Automated decision support systems can provide vital information in a number of domains, ranging from business to medicine. In order to be effective, sophisticated reasoning structures must be built into the decision model. In the work described here, two methodologies are used to provide an added level of sophistication in the decision model. The first is the use of intelligent agents that permit a variety of reasoning methodologies to be used in the same system. The second is consequential reasoning in which consequences impact the decision. Consequential reasoning is incorporated into each agent as well as in the overall intelligent agent system. The methodology is illustrated in a medical decision support system.