On an information compression view of possibility theory
Jörg Gebhardt, Rudolf Kruse · 1994
In this paper we consider possibility distributions as information-compressed representations of imperfect characterizations of object states. With the corresponding formalized semantic background, we investigate how to operate on possibility distributions and how to introduce appropriate uncertainty measures for this framework. It turns out that the well-known concept of a possibility measure and an additional certainty measure, which is not its dual necessity measure, are justified as reasonable choices.>