Predictive Model for the Impact of Online Exams on the Final Semester Grade

Abstract

Countries around the world closed educational institutions in an attempt to stop the spread of COVID-19. Saudi Arabia announced educational quarantine as part of efforts to stop the epidemic. This study provides a case study in extracting educational data from, Majmaah university, College of Science in Al Zulfi, Department. Computer and Information Sciences (Saudi Arabia). The university relied on the online education system using the blackboard system to continue and complete the educational period (2019-2020), which affected the performance of students. This study used data mining methods to investigate the impact of online learning due to quarantine, by creating a predictive model using decision tree algorithms. J48, random tree and RepTree algorithms were used, this model was created using weka tools, and it was found that the random tree is the best algorithm because it has the highest accuracy than J48 algorithm and RepTree algorithm.