A Linguistic Approach to Constructing Fuzzy Numerical Scales

Fuzzy set theory represents an attempt to construct a conceptual framework for a systematic treatment of fuzziness in linguistic variables that are represented in terms of words or sentences. These linguistic variables are interpreted as fuzzy sets characterized by membership functions. Membership function can capture the human's quantitative meaning of such variables in order to handle such meaning for any machinery. To capture human's true meaning of words or sentences, constructing their membership functions is important for the success of applications. A new scaling method, named as fuzzy scale, is proposed. Fuzzy scale, built upon the Hersh & Caramazza's work, provides a simple and heuristic scaling method to capture the meaning of natural language.