A research team led by Canadian University Dubai (CUD) has identified a new technology-based approach to tackling academic integrity using machine learning techniques. The research is the subject of a recent article, “Machine learning based approach to exam cheating detection”, published in PLOS ONE, a prestigious academic journal ranked 30th worldwide across all disciplines by Google Scholar.
The academic study was initiated as the COVID-19 pandemic forced education institutions around the world to adopt remote teaching and assessment methods. The project was led by Dr. Firuz Kamalov, Associate Professor, Department of Electrical Engineering at CUD, in collaboration with, Dr. Hana Sulieman, Professor, Department of Mathematics and Statistics at American University of Sharjah (AUS), and Dr. David Santandreu Calonge, Associate Professor and Director, Corporate Training, at CUD.
Explaining the rationale behind the study, Dr. Kamalov said, “One of the greatest challenges in online education is preserving the academic integrity of student assessments. The lack of direct supervision by instructors during final examinations poses a significant risk of academic misconduct. We set out to explore new approaches to detecting potential cases of cheating on the final exam, using machine learning techniques.”
The new method uses an algorithm to identify cases of cheating using a post-exam analysis of student grades. The inputs to the algorithm are sequences of grades – from quizzes, midterm exams, and the final exam – of an entire class. The method takes account of the grades prior to the final exam, grades on the final exam, and the overall performance of the class, in order to make a judgement.
It is anticipated that the new technique can supplement the work of commercial plagiarism detection software and potentially provide an alternative, non-intrusive deterrent to remotely-invigilated exams. Speaking about the potential impact of the research, Dr. Santandreu Calonge
said, “The insights gained from this study may be of assistance to academics and administrators interested in preserving the academic integrity of course assessments. It also lays the groundwork for future educational research into outlier detection techniques.”
This project adds to the University’s growing body of collaborative research that is supporting the growth and advancement of the UAE knowledge economy. Dr. Kamalov concluded, “Our joint work with AUS shows that the UAE is becoming a hotbed for high-quality research with real-life impact that has the potential to change practices around the world. At CUD, we remain committed to building further relationships in the region that will create more opportunities for students and faculty members to grow and prosper in an environment conducive to creativity and innovation.”