Predicting Learning Styles Using an Adaptive Hierarchical Questionnaire with Machine Learning Techniques

Abstract

An effective teaching-learning process is one in which instructional style and content are based on the learner’s preference and learning style. The same holds true in the case of online Learning Management Systems and especially in Adaptive Education Hypermedia Systems(AEHS) which adapt themselves accordingly to the learning style of the user. Felder- Silverman's Learning Style Model is a commonly used model that provides a framework to classify the learners with the help of the Index of Learning Styles Questionnaire. The main challenge in predicting the learning style in AEHS using the Felder ILS questionnaire is the number of questions in the questionnaire that a learner has to respond to. Here due to a large number of questions in the questionnaire, there is a probability of users skipping certain questions. Such misinterpreted or unattempted questions can lead to inaccuracy in the evaluation stage of the questionnaire. The present study aims to bridge this gap by developing an adaptive ILS questionnaire. Decision Tree Algorithm J48 was used to identify the minimum number of questions that can be used to predict the learning style of the learners across the 4 dimensions of Felder Silverman's Learning Style Model.