Assessment of E-learning Activity During COVID-19 Pandemic using Data Science Technique

Abstract

The emerging of COVID-19 pandemic has become a threat to humanity, many activities of higher education forced to use Learning Management Systempad. This sudden transition significantly changed traditional face-to-face learning into fully online or blended learning environments, requiring both lecturers and students to quickly adapt to digital platforms and new methods of interaction.It provides various tools such as online quizzes, discussion forums, assignment submissions, and learning resources that can be accessed anytime and anywhere. Through these features, lecturers are able to distribute materials, monitor student participation, and evaluate learning outcomes more efficiently.The log records include information such as login frequency, access to learning materials, participation in discussion forums, quiz attempts, and assignment submissions.By applying data mining, statistical analysis, and data visualization methods, complex and unstructured log data can be transformed into meaningful insights. These visual representations help management identify trends, monitor student engagement, evaluate learning effectiveness, and support strategic decision-making in improving the quality of education.Processing log large data was optimized by the use of Graphics Processing Unit (GPU) and python programming language to extract, transform and load data (ETL) then convert the information to specific chart. By analyzing the result, we found some information regarded to student total activities by date, day, hour and also heatmap chart which represent total student activities by hour and day. Finally, the whole series of the processes are proposed as the assessment of e-learning activity on higher education.