Mapping User Behaviors: A Machine Learning Perspective on the NAVER Entry Programming Activity Community

Abstract

Understanding human behavior, particularly in digital realms where anonymity often dominates, is imperative for fostering positive online communities and facilitating effective communication. Deeply rooted in innate tendencies, this behavior becomes particularly complex to interpret in such online environments. Despite these complexities, we demonstrate that the collective behavior within digital communities can be deciphered using machine learning techniques. In our study, we analyzed data from a prominent online community dedicated to enhancing the coding skills of its members. We categorized users into four distinct classes using five different machine learning techniques, achieving an accuracy rate exceeding 95%. These high-precision findings not only reveal intricate patterns of interaction but also set a benchmark for future studies. By uncovering and understanding these behavioral dynamics, our research holds significant potential to shape online community management strategies, inform digital education platforms, and enhance user experience in similar online settings.