A DSS devoted to the evaluation of qualitative criteria

Nowadays, the modelling of observable problems can be carried out in a combined or algebraic way. Expert systems appeared in the 70s introducing the heuristic research and belonging to the combined trend. The introduction of decision support systems (DSS) based on this modelling brings to the decision maker (DM) a wider range of decisions. The role of the DSS, faced with an expressed problem, is to capture a set of decisions using a combined modelling of this problem. In most of the cases, decisions are expressed by a specific processing. Therefore the evolution of the system behaviour is a consequence of the qualitative or quantitative user´s choice. Consequently, the DSS appear to be more adaptable to the problems that are likely to be met in the human sciences. In the particular domain that concerns the authors´ research their Multicriteria Choice Decision Aid (MCDA) model deals with qualitative criteria. In the first part of this paper, the authors introduce the model called Satisfaction-Regret and present its rules. These rules represent in fact a method for synthesizing qualitative criteria. The rules use fuzzy logic operators. This method allows the DM or the expert to evaluate abstracted notions (such as the picturesque features of an environment). The different stages of the method called Satisfaction-Regret are described and an example for a better understanding is also proposed. In the second part, the authors give the architecture of the DSS by briefly presenting its components before explaining in the third part how this system works