Hangeul Font Classification System Proposal of a Network Structure for Improved Font Search Result Similarity

Abstract

Background Typefaces are the fonts born out of the motive and context of their creation. Once formatted, however, it is not easy to specify the background, motive, and proposed usage of the fonts, because a typeface acquires a new context once it is incorporated into the context of the user's purpose of use and environment. The font search service is meaningful in that it helps users to find the optimal font for their intended use out of many fonts marked by complicated and diverse impressions. However, font classification systems that are applied in currently available Korean font search services use different font classification standards, making it difficult for users to judge objectively. In addition, it is difficult to update detailed specifications such as new types, shapes, and uses of the numerous fonts that are continuously created and available over time, and it is difficult for users to obtain accurate information about the font that they are looking for. In an ever-expanding font use environment and range, font search services should be changed to accommodate a large amount of font-related information so that users can organically obtain necessary information. Realizing this problem, this study reexamines the problems of font classification systems currently used in font search services and proposes a new direction and standards for the font classification system with a network structure that can organically combine and use a range of complex font classification information. Methods This study was conducted through theoretical examination and case analysis. In the case analysis, the font classification system selected as the subject of this study, was analyzed to find out how it is utilized in the current font search services and to identify content problems that appear when it is applied to those services. The results of the examination and case analysis revealed that these problems originated from the classification structure designed to visually present various fonts. Based on these results, this study focused on the information delivery diagram type of “hierarchical structure” and “network structure” of the font classification system. The former is defined as the current font classification system, and in the case of the latter, the structure is marked by the mixture and intersection of different groups of information that are subject to different standards. With this difference in mind, this study compared and identified the difference between the hierarchical and network structures and proposed the possibility of using the network structure font classification system as an alternative to the current one, as well as its direction of design, standards, and examples. Results The hierarchical classification structure of the current font classification system has a contradiction, because it generalizes a lot of information associated with a font, such as font type, development period, copyright, and format, by classifying it hierarchically based on specific criteria. On the other hand, the network structure emphasizing symmetrical propositions places the complex information of letters on the same line and interconnects the relationships without special class distinctions such as “superior-inferior”. Convinced that this font classification method can overcome the limitation of the current font classification system which offers only the font information, this study identified basic font classification standards that are based on the collection and classification of information used to distinguish font types. To this end, the standard categories of existing font classification systems, as well as various font legibility-related problems were reexamined, and comprehensive re-editing was attempted based on the reexamination. In this way, font classification criteria and keywords were newly proposed, and based on the results, a network font classification system model was created, and an example of utilization was proposed. Conclusions The Korean font classification systems are presented only with specific criteria such as font shape, expression, and use. The problem is that the Korean writing system Hangul has a grouping structure, and it is difficult to distinguish fonts based only on these standards. In addition, it is not easy to design a font search service that can meet the various aesthetic demands of users with the current hierarchical classification systems. For this reason, it is believed that the network-type font classification system proposed by this study can provide an alternative. It is judged that the network font classification system proposed by the results of this study can be meaningfully used in follow-up studies related to user font selection, such as definition of common terms related to fonts and analysis of font structure and shape information through deep learning. © This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License