Student Behaviour Analysis To Detect Learning Styles Using Decision Tree, Naïve Bayes, And K-Nearest Neighbor Method In Moodle Learning Management System

• Published in: Majalah IPTEK Vol. 33; no. 2; pp. 94 - 104
• Main Authors: Sianturi, Santi Tiodora, Yuhana, Umi Laili
• Format: Journal Article
• Language: English
• Published: Surabaya IPTEK, The Journal for Technology and Science 30.08.2022
• Subjects: Accuracy; Algorithms; Classification; Cognitive style; Data collection; Decision analysis; Decision trees; Distance learning; Educational materials; Learning management systems; Methods; Online instruction; Questionnaires; Reflective teaching; Student behavior; Teaching
• ISSN: 0853-4098; 2088-2033
• DOI: 10.12962/j20882033.v33i2.13665

A learning management system (LMS) manages online learning and facilitates interaction in the teaching and learning processes. Teachers can use LMS to determine student activities or interactions with their courses. Everyone learns uniquely. It is necessary to understand their learning style to apply it in students' learning activities. One factor contributing to learning success is the use of an appropriate learning style, which allows the information received to be appropriately conveyed and clearly understood. As a result, we require a mechanism to identify learning styles. This study develops a learning style detection system based on learning behavior at the LMS of Christian Vocational School Petra Surabaya for the subject of Network System Administration using the Decision Tree, Naive Bayes, and K-Nearest Neighbor. The results of the study showed that the Decision Tree method could better detect and predict learning styles, namely using the 80:20 train split test, which obtained an accuracy of 0.96 process time of 0.000998 seconds, while the K-Fold 10 Cross-Validation test obtained an accuracy of 0.98 and a processing time of 0.04033 seconds.