<![CDATA[Advances in information technology have transformed the way adolescents communicate through social media, yet excessive use negatively impacts psychological well-being. This study aims to develop a Naïve Bayes classification model to identify adolescent depression risks and evaluate the effectiveness of the SMOTE method in addressing data imbalance challenges. This research utilizes the "Social Media Impact on Teen Mental Health" dataset from Kaggle, consisting of 1,200 rows of data, with the entire modeling process validated using 10-fold Cross Validation. The results indicate that the application of the SMOTE method was successful in optimizing detection for the minority class or at risk groups. Through the integration of SMOTE, this classification system achieved a Recall value of 96.77% and an F1-Score of 74.07%, with false negative errors reduced to just 1 case. The combination of the Naïve Bayes algorithm and SMOTE has proven to produce a classification system that is fairer, more sensitive, and relevant for early mental health detection, enabling faster and more accurate identification of depression risks in the digital era.]]>
Copyrights © 2026
