Abstract:
Ensuring academic integrity in online assessments is crucial for upholding fairness and credibility, especially with
the widespread adoption of remote learning. This research addresses key vulnerabilities in preventing cheating and
unauthorized collaboration, common in online assessments lacking direct supervision. To address these challenges, an
intelligent proctoring system was developed and tested on BlockchainStudy.kz — an educational platform that offers online
courses and issues blockchain-based certificates. This system establishes a controlled examination environment through
facial recognition, user activity monitoring, and browser behavior tracking, effectively deterring dishonest practices. The
study adopted a phased methodology, starting with pilot testing for feasibility, followed by large-scale deployment to assess
scalability and effectiveness. The approach combines machine learning-based facial recognition for identity verification, user
action logging, and browser monitoring to detect suspicious behaviors indicative of academic dishonesty. Findings
demonstrated a marked decrease in cheating incidents, enhanced examination credibility, and improved perceptions of
fairness among both students and instructors. By encouraging accountability, the system fostered a culture of honesty within
the online education environment. Ethical concerns regarding privacy were addressed through robust safeguards in
compliance with General Data Protection Regulation (GDPR), building student trust in the proctoring system. This research
contributes to the field of e-learning by providing a scalable, effective solution for maintaining academic integrity in online
assessments. It facilitates informed decision-making for educators, reduces dishonest behavior, and promotes a culture of
integrity within digital education. Overall, this work enriches the body of e-learning knowledge by presenting a practical,
adaptable strategy for overcoming the complex challenges of academic integrity in remote learning environments.
