Co-ordinating Multiple Agents for Diagnostic Decisions

This paper describes the development of an integrated intelligent (expert) system and conventional information system to support decision making of a diagnostic decision in a multi-agent environment. It argues that many a priori requirements for the information system can be elicited by accounting for the specific nature of the diagnostic decision and for the single/multiple-actor environment. This finding is generalized to a framework that indicates that a priori requirements take into account the basic theories on which the information system supporting the decision task is ultimately based. In particular, attention is focused on the nature of the task (e.g. diagnostic), whether the application is for a singleor multiple-agent situation and the nature of the multi-agent structure (e.g. product or functional form of organization).