Volume 19, No. 1, 2022
Cheating Detection in Online Exams during Covid-19 Pandemic Using Data Mining Techniques
Ali M. Duhaim, Safaa O. Al-mamory and Mohammed Salih Mahdi
Abstract
Face-to-face learning has been replaced by E-learning due to the closing of academic institutions in the world during the covid-19 pandemic. Educational institutions faced many challenges in the online platforms and the most important of which was assessing students' performance, which resulted in the general problem of cheating detection in the online exams. E-learning has grown significantly every day over the last decade with the growth of the internet and technology. Therefore, an online examination can be beneficial for people to take the exam, but cheating in tests is a common phenomenon around the world. As a consequence, the prevention of cheating can no longer be completely effective. Many researchers discussed online examination cheating without addressing an important point, which is analyzing students' answers to find similar responses between them. This paper proposed a recommendation system for evaluating students' answers and detecting cheating during an online exam utilizing statistical methods, similarity measures, and clustering algorithms by presenting a set of features derived from an online exam based on the Moodle platform. The results showed that the suggested online examination system effectively reduces cheating and provides a reliable online exam. In conclusion, presenting an effective and fair system that maintains academic integrity, which is the most important aspect of education.
Pages: 341-366
DOI: 10.14704/WEB/V19I1/WEB19026
Keywords: E-learning, Moodle, Online Exam, Cheating Detection, Similarity Measures, Clustering Algorithms.