DigiAddix: Smart Phone Addiction Predictor

Abstract

Smart Phone addiction is becoming a serious issue among school and college students. In today’s digital world, early prediction of addiction is important to identify risks before they develop into serious mental health or academic problems. Many existing methods detect addiction only after it becomes severe. DigiAddix: Smart Phone Addiction Predictor is a system that identifies addiction risk using psychological and behavioral responses collected through questionnaires and converted into a machine-readable format. The system utilizes the CatBoost model to classify users into Low, Medium, or High-risk levels. It also calculates key behavioral indicators such as Attention Stability Index (ASI), Behavioral Drift Score (BDS), Control Impact Score (CIS), and Reliability Score (RS) to analyze user behavior. Based on these insights, users are further classified as Functional or Problematic. The system incorporates adaptive intervention strategies, including personalized reminders, notifications, screen dimming, and focus mode, to help users regulate their smartphone usage over time. Comparative analysis indicates that traditional models such as Decision Tree and Random Forest achieved accuracies of 74% and 82% respectively, whereas the proposed model outperformed them with an improved accuracy of 92%.

Introduction

DigiAddix is an AI-powered smartphone addiction prediction and intervention system designed to address the growing problem of excessive smartphone usage among students and young adults. While smartphones are essential for communication, education, and entertainment, overuse can negatively impact concentration, academic performance, sleep quality, and mental health. Existing solutions mainly focus on screen-time tracking and lack the ability to identify behavioral addiction patterns or provide personalized interventions.

The proposed system combines behavioral analysis, machine learning, and adaptive intervention mechanisms to assess and manage smartphone addiction. User data is collected through structured questionnaires based on established behavioral frameworks and supplemented with real-time usage information such as screen time, notification frequency, and device interactions. The data is preprocessed and transformed into structured features for analysis.

DigiAddix evaluates user behavior using several metrics, including the Control Impact Score (CIS) to measure self-control, the Attention Stability Index (ASI) to assess focus levels, the Behavioral Drift Score (BDS) to detect compulsive usage patterns, and a Reliability Score (RS) to verify response consistency. These behavioral indicators provide deeper insights into smartphone dependency.

For addiction prediction, the system employs the CatBoost machine learning algorithm, which classifies users into Low, Medium, or High-risk categories. CatBoost improves prediction accuracy through gradient boosting and ordered boosting techniques that reduce overfitting. Based on the predicted risk level and behavioral scores, users are further categorized as Functional or Problematic users.

The system’s key feature is its Adaptive Intervention Module, which delivers personalized actions according to addiction severity. Low-risk users receive motivational notifications, medium-risk users receive reminders and suggestions, while high-risk users receive warning alerts, focus mode activation, screen dimming, and other usage-restriction measures. These interventions are continuously adjusted using real-time monitoring and behavioral analysis.

Experimental results demonstrate strong performance, achieving approximately 91–93% accuracy with balanced precision, recall, and F1-scores across all risk categories. Comparisons with Random Forest and Decision Tree models show that CatBoost consistently delivers superior classification performance. Overall, DigiAddix provides an effective and intelligent solution for smartphone addiction prediction, behavioral monitoring, and digital wellness promotion through personalized interventions.

Conclusion

The DigiAddix system provides an effective solution for predicting and managing smartphone addiction using behavioural analysis and the CatBoost machine learning algorithm. The system classifies users into different addiction risk levels and provides adaptive interventions such as reminders, alerts, and focus mode activation to encourage healthier smartphone usage habits. With reliable prediction performance and behavioural monitoring capabilities, DigiAddix supports digital well-being and offers a practical approach for reducing smartphone addiction.

References

[1] Kim, K., Yoon, Y., & Shin, S. (2024). Explainable prediction of problematic smartphone use among South Korea\'s children and adolescents using a Machine learning approach. International journal of medical informatics, 186, 105441. [2] Xiao, B., Parent, N., Rahal, L., & Shapka, J. (2023). Using machine learning to explore the risk factors of problematic smartphone use among Canadian adolescents during COVID-19: The important role of fear of missing out (FoMO). Applied Sciences, 13(8), 4970. [3] Osorio, J., Figueroa, M., & Wong, L. (2024). Predicting smartphone addiction in teenagers: an integrative model incorporating machine learning and big five personality traits. [4] Wu, R., Yu, C., Pan, X., Liu, Y., Zhang, N., Fu, Y., ... & Shi, Y. (2024, May). MindShift: leveraging large language models for mental-states-based problematic smartphone use intervention. In Proceedings of the 2024 CHI conference on human factors in computing systems (pp. 1-24). [5] Lee, J., & Kim, W. (2021). Prediction of problematic smartphone use: A machine learning approach. International journal of environmental research and public health, 18(12), 6458. [6] Kent, S., Masterson, C., Ali, R., Parsons, C. E., & Bewick, B. M. (2021). Digital intervention for problematic smartphone use. International Journal of Environmental Research and Public Health, 18(24), 13165. [7] Joseph, C., & Maheswari, P. U. (2025). Facial emotion-based smartphone addiction detection and prevention using deep learning and video-based learning. Scientific Reports, 15(1), 18025. [8] Li, Z., Long, Q., Hao, M., Liu, L., Weng, Y., Yang, C., ... & Huang, G. (2025). Discovery of factors associated with smartphone addiction among high school adolescents: Using machine learning and network analysis. Journal of Affective Disorders, 120096. [9] Duan, L., He, J., Li, M., Dai, J., Zhou, Y., Lai, F., & Zhu, G. (2021). Based on a decision tree model for exploring the risk factors of smartphone addiction among children and adolescents in China during the COVID-19 pandemic. Frontiers in Psychiatry, 12, 652356. [10] Wayahdi, M. R., & Ruziq, F. (2025). Predicting Smartphone Addiction Levels with K-Nearest Neighbors Using User Behavior Patterns. Jurnal Teknik Informatika (Jutif), 6(5), 3379-3391.

Copyright

Copyright © 2026 Akash P, Akshaya S, Akshaya S, Mr. N. Selvaganesh. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.