In the research of Aulakh, Roul, & Kaushal (2023) named E-learning enhancement through educational data mining with COVID-19 outbreak period in backdrop: A review explains how Educational Data Mining (EDM) has been used to improve e-learning systems, during and after COVID-19 pandemic. This research was made when online learning was the primary mode of education.

There are awesome things that EDM can do which includes:

Performance Analysis- identifying if the students succeeded or failed

Student’s engagement and Behavior- Attendance tracking as well as times when student logs in

Tailor fit Approach- Lessons that are handpicked for different types of learners

EDM has extracted patterns that are not currently explored from student data like LMS logs, scores from quizzes, and other data present.

On the other hand, EDM has also limitations. That includes the following:

The “Why”- it measures the score but what it’s not recording is the reason why the student is doing it or the emotion behind it. The stress or the anxiety was not also captured and understood.

Quality of learning-the ability to login multiple times in a certain platform is not an indication that he has a deeper understanding of the topic. There is no way that creativity or real-life applications are measured and understood.

Bias- algorithms may have a different interpretation of some patterns that can have an effect on those students who didn’t have an access to technology during the pandemic.

To conclude, EDM can be a great tool to predict and recognize pattern in e-learning but it cannot understand and define emotions. It will never ever replace human’s judgment that uses all the aspects of critical thinking combined with emotions and understanding.

Source: https://pmc.ncbi.nlm.nih.gov/articles/PMC10196156/