Extraction of KANSEI from Expert and Sensory Evaluation System based on KANSEI

Human brain mainly performs two kinds of information processes, intelligent process and KANSEI process. KANSEI is used for the sensory test or evaluation in various fields. In this paper, we propose an extraction algorithm of evaluation standard inching KANSEI from evaluation results of person. At first, input variables for evaluation indices are selected by multivariate analysis from available physical parameters. Then, we build KANSEI model of the examinee using fuzzy model with spline function on the antecedent part. In this algorithm it is possible to find the parameters that mainly affect the evaluation results, and to make the examinee decision process clear. The spline function in the antecedent part has an advantage of continuously forecasting evaluation result interpolating measured sample- To show the effectiveness, we applied this algorithm to evaluation of persimmons and eggplants.