Learning Analytics Approach Using Clustering Data Mining for Learners Profiling to Extrapolate E-...

Posted by AACE Conferences on November 3 2021 at 8:38 a.m.

  • Hi colleagues, Much of my research and teaching over the years has focused on different pedagogical approaches and educational technology that can be used to deliver learning experiences in a variety of contexts and settings. At present, my research focuses on data analytics in education, educational data mining, learning analytics and Artificial Intelligence in education. This study aims to gain insights into the patterns related to the diversity of learners and evaluate the relationship between learners’ factors and academic performance.


  • Good morning :)

    Thank you for your interesting research.

    Three questions came to my mind while going through your paper and presentation:

    1. What is the reason for using 4 clusters instead of 3 or 5?

    2. Building different clusters of students to individualize learning support is a nice thing. However, to be able to seperate different students into different clusters, you need variables, that really differ between the groups. With regard to age and gender (variables a teacher can easily observe) cluster 1-3 seem fairly similar to me. Do you think this might be a problem, when trying to individualize learning support?

    3. How would you image the learning support to differ between the clusters 1 to 4? How could fairness-considerations be taken into account (e.g., 1 student receives additional treatment, others not)?

    Thank a lot in advance and kind regards, Michael


  • Hi Michael, Thank you for your questions. The reason to have 4 clusters is because there are 4 categories of their final status: (1) withdrawn, (2) distinction, (3) fail and (4) pass. Besides demographical features (eg: gender, age band), there are also academic record features (eg: highest education, number of previous attempts, studied credits, assessment score) and e-learning interactions feature (eg: sum click). For cluster 1, it is noted that the learning supports will need to be provided to improve the learners’ assessment score and for cluster 4, the learners’ may need to be encouraged or reminded to interact with the online resources more often. With this paper, we are laying the foundation for future work where we will explore how to further value add to a student’s learning experience and success. We hope the above clarifies, have a great week ahead!


  • Your study seems to be quite fascinating and relevant to the present situation of schooling. Data analytics, educational data mining, learning analytics, and artificial intelligence have the ability to give useful insights into how students learn and what variables influence their academic achievement papa's pizzeria. I'm interested in learning more about your unique research topics and approach.


Log in to post a a comment in this discussion.