Qualification of fuzzy statements under fuzzy certainty

Abstract

In many problems the information can be imprecise and uncertain simultaneously. Linguistic terms can be then used to represent each one of these aspects. In some applications it is desirable to combine imprecision and uncertainty into a single value which appropriately describes the original information. We propose a method to combine imprecision and uncertainty when they are expressed as trapezoidal fuzzy numbers and the final goal is to obtain a normalized fuzzy number. This property is very useful in several applications like flexible querying processes, where the linguistic label used in the query is always normalized. © Springer-Verlag Berlin Heidelberg 2007.