• Journal of Korean Library and Information Science Society
• /
• v.41 no.4
• /
• pp.151-170
• /
• 2010

There is a growing interest in user-centered classification or reader-interest classification, as questions have arisen from the meanings and the effects of traditional library classification. American public libraries have used fiction genre classification called bookstore model as an alternative to the traditional classification schemes. As a result, accessibility to the collection was promoted and library service for their users was improved. This study intends to make a comprehensive inquiry about the philosophical background and functional features of genre classification. To the end, literature survey and interviews or e-mails with librarians in American public libraries were conducted.

• Journal of the Korean Society for information Management
• /
• v.28 no.1
• /
• pp.171-193
• /
• 2011

This study aims to analyse the necessity of the subject heading construction for the word based classification and to suggest a methodology that uses various knowledge organization systems(KOS). For this purpose, six kinds of KOS were collected for the 20 selected works in each subject. The collected subjects were analysed in terms of constructing a subject heading for the word based classification. The result of the analysis shows that there is a noticeable difference between the library oriented KOS and commercial oriented KOS. In addition, user oriented tags are more similar to the commercial sector's concerning subject categorization than the library oriented ones. However, there is no noticeable difference among the library oriented KOS, commercial sector oriented KOS, and user oriented tags regarding the subject vocabulary. Some practical implications were suggested for the application to the Korean libraries based on the findings of this study.

• Journal of the Korean Society for Library and Information Science
• /
• v.52 no.4
• /
• pp.51-69
• /
• 2018

The purpose of this study is to analyze the potential use of the theme collection as a new classification method that reflects the interest of users in terms of classification and categorization. For this purpose, the background of the theme collection was identified based on the discussion of the library resource organization and the introduction of the curation service of bookstore. In addition, based on case analysis, which is building the theme collection, concrete concepts and characteristics of theme collection are derived. Based on the above discussion, the classification and categorization characteristics of public library themes collections were analyzed, and the characteristics and functions as a classification were compared with other categories relatively. Finally, the utility and applicability of the theme collection is presented and it is based on the discussions about the user-centered classification system design of the library in the future.

• Journal of the Korean Society for information Management
• /
• v.30 no.3
• /
• pp.157-181
• /
• 2013

Using user satisfaction surveys and LibQUAL+ instruments, libraries are increasingly gathering qualitative data such as verbatim user comments as well as quantitative data. Such qualitative data can be utilized as clues in establishing library service strategies: to better understand user issues, to identify areas for service improvement, and to prioritize user needs. For this, it is necessary to analyze user comments data and to apply results to the delivery of service and the library policies. This study is an attempt to investigate ways in which user comments data can be made useful in libraries. It identifies different methods of analyzing user comments data from LibQUAL+ surveys and compares qualitative data analysis software programs and taxonomies. It also presents the results of applying these tools to a subset of actual user comments data gathered from a recent LibQUAL+ survey at a major university library in Korea.

• Proceedings of the Korean Society for Information Management Conference
• /
• 2006.08a
• /
• pp.63-70
• /
• 2006

이용자중심의 정보기술이 발전하면서 웹 2.0이라는 개념이 등장하고 이용자가 정보를 조직하는 사회적 태깅 방식의 웹사이트가 확산되고 있다. Connotea와 CiteULike는 웹 2.0이라는 개념의 일부인 사회적 태깅 방식을 학술영역에 본격적으로 도입한 최초의 서비스로서, 과거의 통제어휘 위주의 디렉토리 분류체계를 벗어나 이용자들이 참여하여 제공한 정보와 협력하여 부여한 키워드를 중심으로 자유롭게 콘텐트를 분류하고 관리하는 정보공유 포털서비스이다. 이 글에서는 이와 같은 이용자 참여를 유도하는 정보서비스동향을 개관하고 Connotea와 CiteULike의 두 서비스를 비교해보았다.

• Journal of the Korean Society for information Management
• /
• v.25 no.4
• /
• pp.25-42
• /
• 2008

The research was conducted to investigate factors that facilitate users' finding books by analyzing classifications in public library and bookstore. This research was based on the assumption that the users' needs and information behaviors are similar in both public library and bookstore. The main purpose of this study is not to recommend classifying public library collections the way a bookstore does, but to figure out what makes the users' book finding more convenient by analyzing the classifications. To carry out the research, users' book finding in public library and classifications of public library and bookstore are analyzed. Then, a survey was conducted to investigate users' book finding behaviors, degree of convenience in finding books according to different classifications and the causes of the convenience. The results of the research showed that bookstore's classification was more convenient for the users in finding books.

• Journal of the Korean Society for Library and Information Science
• /
• v.43 no.1
• /
• pp.5-26
• /
• 2009

The purpose of this study is to improve the classification focused on KDC 4th edition for children's materials. For the study, literature study, KDC application and user survey in 51 libraries for children were performed. Based on the results, classification for children's materials is proposed to improve in two ways: classification system itself and arrangement in the shelf. First of all, it is required to scatter the subjects that are centered on specific numbers and to allow various types, characteristics and literature genre of children's materials to be classified in KDC. KDC also should provide specific notes and classification guidelines for interdisciplinary subjects. In the shelf arrangement, it is proposed to consider the introduction of user-friendly call number and arrangement principles by ages or titles. In the end, it is necessary to compose various collections that represent users' needs, and the user education program is needed.

• Journal of Korea Game Society
• /
• v.7 no.4
• /
• pp.23-31
• /
• 2007

The study is to explore the ways to design game characters according to the tendency of game users by classifying game users and analyzing the relation between user classes and their preferences towards game characters. The study examines various user classifications based on users' engagement levels, and designs a user questionnaire from them. Based on the result of questionnaire analysis, the study redefines user classes and applies the formal elements of character design to draw on the relationships between user classes and their preferences.

• Journal of Korean Society of Archives and Records Management
• /
• v.14 no.4
• /
• pp.127-151
• /
• 2014

To search based on the functional classification or provenance is not easy for users, and the key word-based information retrieval presents only simple words matching with the title of the records. The Presidential Archive of Korea developed a subject classification scheme to improve the convenience of searching for various records and came up with a subject thesaurus based on the scheme that utilizes the terms appearing on the title of the records and the terms used by the users who searched the portal or requested information disclosure. This research presents the development process of subject thesaurus. It also presents the utilization methods for records management work and services.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.529-532
• /
• 2023

본 논문은 소셜 네트워크 이용자의 텍스트 데이터를 대상으로, 트랜스포머 계열의 언어모델을 전이학습해 이용자의 MBTI 성격 유형을 분류한 국내 첫 연구이다. Kaggle MBTI Dataset을 대상으로 RoBERTa Distill, DeBERTa-V3 등의 사전 학습모델로 전이학습을 해, MBTI E/I, N/S, T/F, J/P 네 유형에 대한 분류의 평균 정확도는 87.9181, 평균 F-1 Score는 87.58를 도출했다. 해외 연구의 State-of-the-art보다 네 유형에 대한 F1-Score 표준편차를 50.1% 낮춰, 유형별 더 고른 분류 성과를 보였다. 또, Twitter, Reddit과 같은 글로벌 소셜 네트워크 서비스의 텍스트 데이터를 추가로 분류, 트랜스포머 기반의 MBTI 분류 방법론을 확장했다.