Color and Font Recommendations based on Mental Images of Text(<Special Issue>Arts & Entertainment 3)

Abstract

We propose a method of recommending colors and fonts appropriate for texts based on the associations between words and colors. Colors and fonts were evaluated by SD scales using affective words in psychological experiments. First, our system estimates colors suitable for the mental representations of the inputted text. Then, the colors are described by affective words. Next, we obtain the degree of similarity between the colors and fonts, whose mental perceptions were evaluated by SD scales in the psychological experiments. Finally, our system recommends colors and fonts that are most appropriate for the inputted text. We verified the validity of our method for selecting appropriate text fonts and colors.