Two characterizations of a minimum-information principle for possibilistic reasoning

AbstractWe show that the principle of maximum U-uncertainty for ampliative possibilistic reasoning can be characterized as uniquely satisfying a small set of normative axioms. Two proofs are given—one each for Hisdal's and Dempster's definitions of conditional possibility. These results complement a similar characterization of maximum entropy for ampliative probabilistic inference, given by Paris and Vencovska