Scholarlink

<p>In the digital era, Internet Gaming Disorder (IGD) and cyberbullying have emerged as significant risk factors for mental health issues. This study investigates the relationship between online gaming behaviors, cyberbullying incidents, and their psychological impacts using advanced machine learning and deep learning techniques. For IGD, various models including Logistic Regression, Random Forest, Ensemble Models, Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM) networks were evaluated. LSTM achieved the highest accuracy of 91.6%, demonstrating its robust capability in detecting patterns related to IGD. In cyberbullying detection, deep learning models, particularly LSTM and CNN, stood out with accuracies of up to 96% and 95%, respectively, underscoring their effectiveness in identifying complex patterns in cyberbullying interactions. Deep learning models, particularly LSTM and CNN, demonstrate high accuracy in cyberbullying detection and show promise for early AI-driven mental health intervention. This study contributes a robust framework for early risk detection, emphasizing the importance of temporal characteristics in modeling gaming behaviors and cyberbullying interactions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Learning-Powered Predictions of Mental Health Outcomes Related to Internet Gaming Disorder and Cyberbullying

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月12日

查看全文
  • Bryan Samuel James,
  • N. Krishnamoorthy,
  • V. Vinoth Kumar,
  • Iyappan Perumal

摘要

In the digital era, Internet Gaming Disorder (IGD) and cyberbullying have emerged as significant risk factors for mental health issues. This study investigates the relationship between online gaming behaviors, cyberbullying incidents, and their psychological impacts using advanced machine learning and deep learning techniques. For IGD, various models including Logistic Regression, Random Forest, Ensemble Models, Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM) networks were evaluated. LSTM achieved the highest accuracy of 91.6%, demonstrating its robust capability in detecting patterns related to IGD. In cyberbullying detection, deep learning models, particularly LSTM and CNN, stood out with accuracies of up to 96% and 95%, respectively, underscoring their effectiveness in identifying complex patterns in cyberbullying interactions. Deep learning models, particularly LSTM and CNN, demonstrate high accuracy in cyberbullying detection and show promise for early AI-driven mental health intervention. This study contributes a robust framework for early risk detection, emphasizing the importance of temporal characteristics in modeling gaming behaviors and cyberbullying interactions.