Clustering Social Media Users Based on Digital Behavior Patterns Using Machine Learning Approaches for Understanding Online Engagement and Addiction Tendencies

The rapid growth of social media has significantly influenced patterns of human behavior, communication, and productivity. While social platforms offer opportunities for connection and self-expression, excessive or unstructured use can lead to declining satisfaction and reduced wellbeing. This study aims to identify distinct patterns of social media user behavior by applying machine learning clustering techniques to examine differences in engagement, self-control, satisfaction, and productivity. Using a dataset of 1,000 participants containing demographic and behavioral variables, the study employed the K-Means algorithm and Principal Component Analysis (PCA) to classify users into groups based on their digital activity. The results revealed two distinct behavioral clusters: the first, labeled as Engaged but Controlled Users, consists of individuals who demonstrate balanced and intentional use characterized by high satisfaction and low productivity loss; the second, labeled as Addicted and Unproductive Users, consists of individuals with higher self-control but lower satisfaction and greater productivity loss, suggesting a pattern of habitual or emotionally unfulfilling engagement. The findings indicate that digital wellbeing is determined not only by the amount of time spent online but also by the quality and purpose of engagement. This study highlights the importance of mindful and goal-oriented social media use, suggesting that interventions should focus on promoting purposeful engagement rather than merely limiting screen time. The application of clustering in behavioral analysis provides a data-driven framework that can assist policymakers, educators, and platform developers in designing strategies that support healthier and more meaningful digital interactions.