Abstract: A loss control system for an insurance classification plan has a policy holder database, a predictive apparatus and a derived actual loss ratio generator. The policy holder database generates signals indicative of the premium, actual loss, a one a plurality of classification plan variable values and an actual loss ratio for each policy holder. The predictive apparatus generates a plurality of predicted loss ratio signals indicative of predicted loss ratios of the policy holders. The derived actual loss ratio generator generates signals which are indicative of a difference between the predicted loss ratio and the actual loss ratio of the policy holders.

Abstract: A problem solving expert system is provided which is particularly useful in managing the health care of individual patients. A description of a problem (e.g., medical condition) and a proposed solution therefor (e.g., medical procedure) is entered via a user interface. A topical library is searched to identify information relevant to the problem and proposed solution. Access to the identified information is available in either a full text or synopsis format, to assist a user in assessing the appropriateness of the proposed solution. An inference engine provides a recommendation to the user as to the appropriateness of the proposed solution based on information entered via the user interface and rules associated with the inference engine. A user can interact with the inference engine in either a structured or guided mode.