Online Examination Proctoring for Quality Assurance in the Post COVID-19 Pandemic New Normal

Abstract

There has been a steady growth and usage of E-learning platforms over the past decade. However, with the Covid-19 pandemic, whereby lecturers and instructors in Institutions of learning could not have physical contact with the students, E-learning is witnessing astronomical growth and usage. Most of the available E-learning platforms in use have no proper solutions to academic examinations. Online Examination Proctoring System is an integrated system that uses live proctor(s) with an AI-enabled support. In the heat of the covid-19 pandemic, semester examinations conducted using online learning platforms by most learning institutions were marred with irregularity and academic fraud by the students. Some institutions gave out short essay questions in which the students copied from the internet, students were free to make calls and find answers from friends, and Instructors were not able to adequately create and author questions with graphics as would reflect their needs. At the same time, some implemented inefficient remote proctoring systems where a human invigilator keeps monitoring the students' activities. The lack of effective and efficient solutions to online examination proctoring by the e-learning platforms has led to cheating in the online examinations. With the covid-19 pandemic still very much with us and e-learning becoming the new normal, there is a need for researchers and developers to develop efficient and effective automated and online examination proctoring systems. This paper presents an Online Examination Proctoring System that detects and reports cheating in an online examinations. This ensures the integrity of the test which provides in overall, the desired quality assurance. The system was developed using Visual Studio Code installed on the system, and Anaconda environment was used to run the machine learning algorithms. The test results obtained from deploying the system showed significant improvement over the existing ones with 92% cheating detection rate.