AUTOMATED EXAMINATION MISCONDUCT MONITORING USING IMAGETRAINING AMONG COMPUTER SCIENCE STUDENTS

Abstract

Examination misconduct refers to any immoral or unethical behavior occurring before, during, or after an examination. It constitutes a disorganized practice that undermines the quality of education and devalues the credibility of certificates awarded to students across different levels of learning. The rising incidence of examination malpractice poses a serious challenge to educational systems and demands innovative solutions for effective management. The objective of this research is to automate examination misconduct monitoring using image training techniques capable of detecting and tracking students engaged in malpractice, with a focus on Computer Science students. The system was developed using OpenCV and PyAudio in Python, integrated with a web-based interface, and supported by a MySQL database. The Fisher Linear Discriminant algorithm was employed for face recognition, alongside dataflow and flowchart modeling to guide system implementation. The system provides proof-based monitoring, enabling institutions to capture and recognize the faces of students involved in examination misconduct. Experimental results demonstrated that the approach is effective in identifying suspicious activities and can serve as a technological framework for universities to adopt in strengthening examination integrity.