Predicting Online Game Disorder in Youngsters Using K-Nearest Neighbors and Artificial Neural Networks

Abstract

Online game disorder has emerged as a significant mental health concern, particularly among youngsters. With increasing engagement in online gaming, early detection of gaming addiction is crucial to prevent long-term psychological and social consequences. This study proposes a predictive framework using K-Nearest Neighbors (KNN) and Artificial Neural Networks (ANN) to classify online game disorders based on behavioral patterns extracted from social media and gaming activity. The dataset comprises user interactions, engagement frequency, and linguistic cues indicating addiction tendencies. Experimental results demonstrate that ANN outperforms KNN in terms of accuracy, achieving an improved classification rate of 98.4%. The findings indicate that AI-driven models can effectively identify high-risk individuals, providing a foundation for real-time intervention strategies. This research contributes to mental health analytics by integrating machine learning techniques to predict and mitigate the impact of online game disorder.