Research And Application Of Emotional Imageries Of Typefaces Based On Kansei Engineering

As the most important carrier of information,Chinese characters are full of every corner in daily life,and the typeface is the most intuitive expression of the Chinese characters which directly affects the reader’s first impression.Most of the domestic research on Chinese characters typefaces focuses on the design.There are relatively few studies on the emotional feelings conveyed by the typefaces to the audience.Meanwhile the currently existing research on typefaces emotions is mainly on the basis of theoretical narrative and qualitative research.Based on Kansei engineering,this paper did quantitative studies on the emotional tendency of Chinese characters typefaces.The main work was divided into three aspects.:First,establish the semantic space for evaluating typefaces.A large number of words used to describe fonts were collected.After preliminary screening by experts,card classification experiments were carried out,and then representative words were extracted by two cluster analysis.According to the semantic differential method,8 pairs of emotional words were finally obtained.Second,modify the semantic space.20 kinds of font samples and emotional word pairs were composed into a subscale which was issued as questionnaires.According to the data recovered from the subjects,one pair of poorly performing words was removed,and the remaining 7 pairs were used as emotional imageries for evaluating font emotions in this paper..Third,establish a mathematical model between the font structure and the emotional word pairs.The basic structural attributes of the font were analyzed and divided into different categories according to their characteristics,so that an independent variable reaction matrix was obtained.Then the method of Quantification Theory Type I was used to establish a linear prediction equation between the characteristics of the font attributes and their sentimental tendencies and the equation’s accuracy was verified later.Based on the results of the experiments and analysis,the mathematical model can calculate the scores of the 7 pairs of emotional words when the structural attribute characteristics of a certain font are known,and predict the emotional tendency and its strength that font conveys to readers.The emotional words in semantic space are relatively objective,and these values can be used for classification and recommendation of fonts,so that users can quickly and efficiently filter out appropriate fonts.On the other hand,in a predictive equation of a emotional word pair,the coefficient of each feature indicates the direction and extent of its contribution to the words.When designing a series of new typeface,whose emotions that need to be expressed are determined,Designers can focus on the characteristics of the font attributes that are highly contributed to this emotional imagery.It plays a guiding and auxiliary role to a certain extent.The main innovation of this topic is the introduction of the method of Kansei engineering to quantitatively study the emotions of the typefaces.According to the obtained mathematical model and related parameters,the emotional imagery prediction and classification recommendation function of the font can be realized,also these data can assist the design of new typefaces.