Reaserch on Monolingual Font Recommendation based on Target Perception
Xiangqi Kong, Zeyu Liu · 2020
Describing the target perceptual image is the most natural way to propose a request of font recommendation. For merchants, the target perceptual image of product is even more concerned than the actual ones. For professional designers, experience can be used to effectively convey the target perception through appropriate graphic design. And font selection is one of the important aspects. For ordinary users, it is relatively easy to judge or express the perceptual image of concrete products. However, it is difficult to judge the perceptual image of abstract fonts. To explore the solution of the problem, this paper proposes a monolingual font recommendation method based on the target perception. Specifically, the user assigns the target sense in the designated perceptual image space. The font library is divided into typical fonts and other fonts. Then the system will compare the target with the perceptual image space of the typical font library. And recommended fonts will be listed according to the perceptual similarity of target and the selected typical fonts. Furthermore, because different fonts of the same text in the same language often have similar appearance characteristics. On the basis of the typeface similarity of the typical fonts and others, further font recommendations can be made. In addition, a two-dimensional space word cloud is adopted to visualize the font similarity. In this way, users will be more intuitive when exploring and choosing recommended fonts.