Title: Group student profiling in massive open online courses using educational data mining

Authors: Abdelghani Hmich; Abdelmajid Badri; Aicha Sahel

Addresses: Faculty of Sciences and Techniques Mohammedia, Laboratory of Electronics, Electrotechnics, Automation and Information Processing, University Hassan II Mohammedia Casablanca, BP 146, Mohammedia, Morocco ' Faculty of Sciences and Techniques Mohammedia, Laboratory of Electronics, Electrotechnics, Automation and Information Processing, University Hassan II Mohammedia Casablanca, BP 146, Mohammedia, Morocco ' Faculty of Sciences and Techniques Mohammedia, Laboratory of Electronics, Electrotechnics, Automation and Information Processing, University Hassan II Mohammedia Casablanca, BP 146, Mohammedia, Morocco

Abstract: Student profiling is one of the great accomplishments in the field of educational data mining (EDM). Student profile can mainly offer the most exact description of students in order to be able to offer students the most appropriate personalisation and recommendation. In this paper, we use a demarche for discovering group Massive Open Online Course (MOOC) student profile in relation to their quiz performance. This demarche is based on the combination of hierarchical density-based spatial clustering of applications with noise (HDBSCAN) and apriori algorithm, first the groups of students with similar learning characteristic are extracted by using the clustering algorithm HDBSCAN, then the extracted groups were analysed by using apriori algorithm in order to characterise the distinct profiles of group of students using data collected from a MOOC in Moodle platform. The results show that there are three groups of students organised by quiz performance and learning behaviours. First group is characterised by the low quiz performance, the second group is characterised by the good quiz performance and the third group is characterised by the excellent quiz performance.

Keywords: student profiling; EDM; educational data mining; E-learning; MOOCs; Massive Open Online Courses; clustering; HDBSCAN; association rules mining.

DOI: 10.1504/IJDS.2022.124358

International Journal of Data Science, 2022 Vol.7 No.1, pp.78 - 94

Received: 29 Sep 2021
Accepted: 03 Jan 2022
Published online: 25 Jul 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article