Abstract
A problem of relevancy of fuzzy models (viz. models with ability to cope with fuzziness) is stated. Its importance in any identification scheme is underlined and discussed in light of some shortcomings existing in procedures of fuzzy model building. Two novel concepts (viz. equality index and confidence interval) are introduced and utilized for fuzzy-model evaluation. A flexible way of interpreting results with the aid of fuzzy models is formulated by means of interval-valued fuzzy sets (Φ-fuzzy sets). The clear importance of studies on the relevancy of fuzzy models for problem formulation and solution is pointed out.