Quantitative Discrimination and Its Analyzing Method among Nuance of KANSEI Words on Estimation Scale

Abstract

As many teaching materials have been developed on multimedia, KANSEI (so called hearty, favor, feeling, or aesthetic sense) words have been frequently used just even in retrieval or processing of the educational information. But, until now in studies of KANSEI words, fuzziness of KANSEI words is slighted even though those are used in a fuzzy way. In this paper, forcusing our attention on fuzziness of KANSEI words, the authors formulate the mathematical model to understand quantitatively how nuance of KANSEI words are slightly different from person to person, and discuss researching method to determine numerical values of the quantitatively fuzzy nuance of KANSEI words. The formula is described with membership functions in the fuzzy set theory, and forms of the membership functions are calculated by analyzing the examinees' response data. In order to understand the practical meaning of the methematical model, some applications of the nuance function are shown in the last part of this paper.