Probabilistic and fuzzy relational semantic systems in propositional approximate reasoning

摘要

The notion of an RSS (relational semantic system) as a set of atomic propositions evaluated in the unit interval and functions representing complementation (negation n) and logical intersection (junctor j) is introduced. This concept is used to determine under what circumstances the models of fuzzy approximate reasoning can be embedded in a probability space. This situation includes most of the classical fuzzy-set-theoretical approaches. A straightforward way of representing the concept of conditioning for most fuzzy measures is obtained.