Semantic Evaluation in Possibilistic Logic, Application to Min-Max Discrete Optimisation Problems

Jérôme Lang

Abstract: Propositional possibilistic logic is a logic of uncertainty in which the notion of inconsistency is gradual, each interpretation having a compatibility degree with the uncertain available knowledge. We present here an algorithm for the search of the best interpretation of a set of uncertain clauses (i.e., the most compatible with it), which is an extension to possibilistic logic of semantic evaluation (based on the Davis and Putnam procedure). Possibilistic logic is also a general framework for translating discrete "min-max" optimisation problems (some examples of such problems are discussed).