The formalization of preferences and the representation of results in a group decision support system

The article describes input/output techniques for quantifying individual preferences and representing the results of a GDSS. Specifically, analysis is conducted of the features of numerical, spatial and linguistic representations of preference input in relation to the technological means currently available. The same analysis is then applied to the representation of the results. The paper argues that there is no optimal combination of I/O technologies and representation techniques, and multiple forms of human–machine interaction are required in order to respond effectively to different cognitive styles and to the different features of the tasks to be accomplished. The article also argues that the current GDSS make excessive use of numerical representation techniques which, although known to give unsatisfactory representation of uncertainty decision situations, for cultural and practical reasons are often preferred to the more natural spatial or linguistic techniques. The conclusions describe the principal theories concerning the way in which an imprecise input can be interpreted and handled by a calculation system or by a reasoning mechanism to produce output which effectively supports the decision-maker's activity.