Determine Felder Silverman Learning Style Model using Literature Based and K-Means Clustering

Abstract

The student learning process is influenced by several factors, one of which is student learning styles. Learning style is one of the most important factors in the E-learning environment because it can help the system to effectively personalize the learning process of students according to their learning style. Previously, to detect student learning styles by asking students to fill out questionnaires. However, there are problems with this static technique. One of these problems is the lack of students' self-awareness of their learning preferences. In addition, almost all students feel bored when asked to fill out a questionnaire. This research determined the learning style based on the Felder and Silverman Learning Style. This determination process is carried out using student activity data on a pure Moodle learning management system (LMS). The process begins with processing based on the literature to get a vector combination of learning styles. Student activity data is processed to produce data that only contains activities that are included in the selected features. The results of both are combined as input to the clustering process. This research applies the modified K-Means Clustering algorithm. Modifications were made using the learning style combination vector as the initial centroid. The k value used in this study was 8 which came from 8 combinations of learning styles from 3 dimensions used in this study. This is different from the combination of learning styles in FSLSM which has 16 combinations of learning styles originating from 4 dimensions of learning styles. This difference is caused by student activity data that only supports 3 dimensions of learning style.