A new approach to introducing semantics and subjectivity in approximate reasoning

Abstract

It is by now taken for granted (and perhaps recognized) that the concept of possi bility via fuzzy sets could be of interest to build up computerized models of approxi mate reasoning, and most works in this way proceed via the min-max operator and projections onto subspaces.

The purpose of the present paper is two-fold: first, to explicitly introduce subjectivity in this framework, and second, to re-define the entropy of fuzzy sets by using the concept of entropy of degree d of deterministic maps which we previously derived, and to apply the corresponding model so obtained to approximate reasoning.

Basically, the fuzzy entropy so derived involves the parameter d which depicts how does the observer scan the fuzzy set under consideration, therefore more flexibility in the modelling of the interactions between observer and observable.