The NOMO Zone: A Web-Based System for Predicting Electronic Gadget Addiction and Stress Using Ensemble Machine Learning

Abstract The rising dependency on electronic gadgets among students has led to significant concerns regarding gadget addiction and associated mental health issues such as anxiety, sleep disruption, and academic decline. Traditional assessment methods rely heavily on subjective self-reporting or clinical evaluations, which are often inaccessible, stigmatized, and lack scalability [3,5]. This paper presents THE NOMO ZONE, a web-based intelligent system that predicts gadget addiction levels and detects psychological stress by integrating behavioural analytics with sentiment analysis. The system employs a structured 10-question behavioural survey and user-generated text data to train and evaluate multiple supervised machine learning classifiers, including Support Vector Machine (SVM) [8], Random Forest, Decision Tree, and the proposed ExtraTrees Classifier [19]. Experimental results demonstrate that the ExtraTrees model achieves 100% test accuracy and 99.84% training accuracy, outperforming all other classifiers. The system is deployed as a modular Flask web application with a responsive frontend, secure admin dashboard, and real-time prediction capabilities. The NOMO Zone provides actionable classification into five distinct addiction levels---from "No Impact" to "Severe Dependency"---offering a scalable, private, and data-driven tool for early intervention and digital wellness promotion [6,14]. Keywords: Gadget addiction, stress detection, machine learning, ExtraTrees classifier, sentiment analysis, Flask, mental health screening, digital wellness.