An Efficient Hybrid AI-Based Privacy-Preserving System for Remote Work Activity Classification

Author

Smrita Sachdeva, Rupali Chaure, and Dr Mohit Singh Tomar


Abstract

 The rapid adoption of remote and hybrid work has increased the need for effective productivity monitoring while raising concerns regarding employee privacy and intrusive surveillance. This study presents a privacy-preserving remote work monitoring framework that combines lightweight activity monitoring with a hybrid rule-based and artificial intelligence (AI)-based activity classification approach. The proposed system collects essential activity information, including application usage, window titles, timestamps, and idle status, and classifies activities into Work, Private, and Idle categories. The system identifies known activities using predefined rules, while it evaluates unknown or ambiguous activities using an AI-based classification mechanism. A secure backend based on FastAPI and PostgreSQL manages activity data, authentication, and processing, while a React-based dashboard provides productivity summaries and visual insights. The comparative evaluation reported 99.2% classification accuracy, with approximately 99% precision, recall, and F1-score, while idle-time detection achieved approximately 97% accuracy. The system processed activity logs in approximately 1.9 s, with a dashboard response time of 1.6 s and database query latency of approximately 280 ms. The usability evaluation indicated approximately 91% task-completion efficiency. Overall, the results demonstrate the feasibility of integrating AI-assisted activity classification with privacy-oriented monitoring and secure system architecture for remote-work environments.



Keywords

Remote Work Monitoring; Privacy-Preserving Computing; Artificial Intelligence; Hybrid Classification; Activity Recognition; Productivity Monitoring; FastAPI; PostgreSQL; React Dashboard; Secure Monitoring.



Full Text:

Download Paper PDF


References


  • G. Ainsworth and J. B. Carver, “Employee Productivity Monitoring in Remote Work Environment,” Journal of Organizational Computing and Electronic Commerce, vol. 34, no. 1, pp. 45–60, Jan. 2024.
  • [M. A. Rahman, S. Hossain, and R. Islam, “AI-Based Workplace Activity Monitoring Systems for remote workforce management,” IEEE Access, vol. 11, pp. 112345–112356, 2023.
  • Kumar and R. Sharma, “Privacy-Preserving Employee Monitoring Using AI-Based Activity Classification,” in Proc. IEEE Int. Conf. on Artificial Intelligence and Data Engineering, 2024, pp. 210–216.
  • Chen and H. Zhang, “A Hybrid AI Approach for User Activity Classification in Workplace Productivity Systems,” Expert Systems with Applications, vol. 225, 2024.
  • Patel, K. Shah, and R. Mehta, “Design and Implementation of Remote Work Monitoring Systems,” International Journal of Computer Applications, vol. 185, no. 12, pp. 21–27, 2023.
  • Brown et al., “Artificial Intelligence for Workplace Productivity Analytics,” in Proc. ACM Int. Conf. on Human Factors in Computing Systems, 2023.
  • B. Patil and A. Deshpande, “Design of Secure Web-Based Monitoring Systems Using RESTful APIs,” International Journal of Advanced Computer Science and Applications, vol. 14, no. 4, pp. 512–518, 2023.
  • Chollet, “Xception: Deep Learning with Depthwise Separable Convolutions,” in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 1251–1258.
  • He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2016.
  • Tan and Q. Le, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” in Proc. Int. Conf. on Machine Learning (ICML), 2019.
  • Vaswani et al., “Attention Is All You Need,” in Proc. Advances in Neural Information Processing Systems (NeurIPS), 2017.
  • PostgreSQL Global Development Group, “PostgreSQL Documentation,” 2024. Available: https://www.postgresql.org/docs/

 

 

Share your valuable work from Social Media Buttons