Extraction of relationship among Kansei words by expert system using rough set analysis

In this paper, we propose the extended technique from the usual rough set analysis by considering the dominance relations to the quantitative data, and propose the criterion to select the rule. Proposed expert system using rough set analysis applies a most likelihood to estimate mapping from attributes of product to Kansei word. Usual estimation of interval efficiency has been developed by linear regression, however, our system can estimate interval efficiency by nonlinear mapping. As the experimental result, we analyze about the influence to user's psychology from the front design of a car and extraction of relationship among Kansei words by expert system using rough set analysis.