Preference modeling as a reasoning process: A new way to face uncertainty in Multiple Criteria Decision Support Systems

How is it possible to improve both multicriteria methodology and building of flexible and reliable Decision Support System? In this paper an attempt is made, through the use of Artificial Intelligence techniques, towards a declarative representation of decision aid activities and principally towards a new way to face problems of uncertainty in preference modeling. Particularly the use of nonmonotonic inference procedures is explored and a new formalization of the preference relations is proposed on a reason maintenance approach. A possible generalization is also discussed.