An evolutionary fuzzy color emotion model for coloring support systems

In this study, we improved the coloring support system that was proposed in our previous research. The previous system could advise color combinations for clothing. The user enters some emotional keywords, such as “casual” and “pretty,” which are impressions put forth by colors, and the system retrieves color combinations that seem to suit the keyword from a color database. The proposed system has many fuzzy color emotion models, which are controlled by parameters for defining membership functions or fuzzy rules. These models are evolved using interactive genetic algorithms, to adapt them to the user’s particular emotional responses, because human reactions vary for each individual. This paper proposes a method for the advanced evolution of the fuzzy color emotion model. This method supports co-evaluation of color combinations, which exhibits a better performance in evolving models compared to the previous method.