The increasing shift toward online learning and digital examinations has introduced a range of challenges, particularly in maintaining academic integrity. This paper presents the design and implementation of an E-Exam Cheating Control System aimed at detecting and reducing cheating during online examinations. The system integrates multiple technologies, including webcam-based facial recognition, screen activity monitoring, keyboard and mouse behavior logging, and network traffic inspection, to ensure exam fairness. A rule-based engine and AI-powered behavior detection model are used to identify suspicious patterns such as multiple face presence, gaze switching, use of unauthorized devices, and switching away from the exam window. The system was developed using PHP, JavaScript, and OpenCV for facial recognition and tested across different environments and exam scenarios. Results show the system's effectiveness in identifying cheating attempts with a detection accuracy of over 90% in controlled settings. A comparative analysis with other systems demonstrates its robustness and adaptability. The research contributes to the field of e-learning security and proposes improvements for future systems, including privacy-preserving mechanisms, offline examination support, and multi-factor identity verification. The system has potential for real-world deployment in educational institutions to promote integrity in digital assessments.
Derrick et al. (Tue,) studied this question.