Automatic detection of learning styles in learning management system by using literature-based method and support vector machine

Abstract

Each learner has their own preferences in the learning process. Differences in preferences are closely related to the learning style of each learner. Personalization of e-learning is an overview of online learning that has been customized content based on learning styles of each learner. Detecting learning style needs a technique that is effective and accurate. This study combines literature based method with Support Vector Machine (SVM) to detect students' learning styles. The data used is learning log data of Data Structures and Algorithms class at the Faculty of Computer Science, Universitas Indonesia. The test results showed that SVM has better accuracy compared to Naive Bayes.