Implementation of an adaptive DSS generator for senior managers

Providing active, interventionist support is advocated as a means of decision support that goes beyond the traditional passive approaches. A differential study conducted as the first stage of this project showed that senior decision-makers with different personality dispositions display distinct preferences when making decisions. This paper presents a decision support architecture that attempts to provide the basis for systems that are capable of adapting to their users based on their personality preferences. An implementation of the architecture is also discussed. Repeated use of a decision support system is expected to give the system the knowledge required to incrementally adapt to the decision preferences of the individual. The capabilities of adaptation are facilitated through the construction of profiles of decision-makers, decision situations and decision domains. The profiles are maintained as criteria preference models. An inference mechanism that utilises neural networks is used to build these profiles. The approximations provided by the system are expected to improve progressively. A mature system may be capable of identifying inconsistencies in an individual's decision-making and provide active decision support. Presented at: 4th Conference of the International Society for Decision Support Systems; 1997 Jul 21-22; Lausanne, Switzerland. 18 leaves