Mining Moodle Data to Detect the Inactive and Low-performance Students during the Moodle Course

Abstract

In web-based learning systems such as massive open online course (MOOC) and modular object-oriented developmental learning environment (Moodle), monitoring the student's activities as well as predict the low-performance students is an important task because it enables the instructors to award the students when their activities level drops from normal activities levels as well as having lower grades. We used several machine learning (ML) classification and clustering techniques to extract the pattern from student data during completing the Moodle course; which enables the instructor to detect the low-performance student in advance before the examination. The experimental result shows that the fuzzy unordered rule induction algorithm (FURIA) classification technique achieves high accuracy in detecting inactive students as well as predicts the different categories of the student during the Moodle course. The K-means clustering is also able to group the inactive and active users and poorly performed users. The result demonstrates that our proposed system will be easily integrated to Moodle system to send alert to inactive and low- performance students while completing the course and build efficient education environment for the students.