<![CDATA[The development of information and communication technology has encouraged the increasing use of social media among teenagers, especially high school students (SMA). Social media is not only used as a means of communication and entertainment, but also has the potential to influence students' learning behavior. Excessive use can reduce concentration and learning effectiveness, so analysis is needed to systematically identify patterns of social media usage behavior. This study aims to group the social media usage behavior of high school students using the K-Means Clustering algorithm. Data were obtained through a closed questionnaire with a numerical scale that includes three variables, namely the frequency of social media use, duration of use, and the number of social media platforms used. A total of 20 sample data were used to illustrate the calculation process. The analysis stages include data normalization using the Min-Max method, determining the number of clusters as many as three clusters (K = 3), calculating distances using Euclidean Distance, and iteratively updating the entroids. Based on the results of the K-Means algorithm calculations until reaching a convergent condition, the social media usage behavior of high school students was successfully grouped into three clusters, namely low usage with centroids (3.50; 1.25; 2.75), moderate usage with centroids (7.25; 3.25; 5.88), and high usage with centroids (12.25; 5.75; 8.88). These results indicate that the K-Means algorithm is effective in grouping students' social media usage behavior based on differences in access frequency, duration of use, and level of social media interaction.]]>
Copyrights © 2026
