열람실 로그데이터를 활용한 좌석 유형 분류 및 물리적 요인에 관한 연구: K대학교를 중심으로

Abstract

University libraries are transforming into dynamic learning environments supporting academic pursuits and personal development. Although recent studies have employed various methods, most have primarily depended on subjective assessments or basic usage statistics, which do not fully capture user behavior through data-driven clustering techniques. This study utilized machine learning techniques to examine 157,021 library reading room usage logs. By applying K-means, BIRCH, and GMM algorithms, the research classified different seat types and analyzed environmental factors. Unlike previous studies, this research addressed methodological gaps by adopting clustering-based machine learning methods, enabling a systematic exploration of the relationship between seat preferences and environmental influences. The findings introduced a novel seat classification framework and an improved operational strategy, offering valuable contributions to the field of library science and practical insights for knowledge management. By leveraging advanced clustering techniques, this study represents a meaningful step forward in library research methodology, effectively linking theoretical insights with practical solutions for optimizing library space management.