International Journal of Electrical and Computer Engineering (IJECE) Vol. 9, No. 5, October 2019, pp. 4408~4416 ISSN: 2088-8708, DOI: 10.11591/ijece.v9i5.pp4408-4416  4408 Journal homepage: http://iaescore.com/journals/index.php/IJECE Identifying learning style through eye tracking technology in adaptive learning systems Inssaf El Guabassi 1 , Zakaria Bousalem 2 , Mohammed Al Achhab 3 , Ismail jellouli 4 , Badr Eddine EL Mohajir 5 1,4,5 Faculty of Sciences, Abdelmalek Essaadi University, Moroco 2 Faculty of Science and Technologies, Hassan 1 st University, Morocco 3 National School of Applied Sciences, Abdelmalek Essaadi University, Moroco Article Info ABSTRACT Article history: Received Jul 1, 2018 Revised Apr 18, 2019 Accepted Apr 25, 2019 Learner learning style represents a key principle and core value of the adaptive learning systems (ALS). Moreover, understanding individual learner learning styles is a very good condition for having the best services of resource adaptation. However, the majority of the ALS, which consider learning styles, use questionnaires in order to detect it, whereas this method has a various disadvantages, For example, it is unsuitable for some kinds of respondents, time-consuming to complete, it may be misunderstood by respondent, etc. In the present paper, we propose an approach for automatically detecting learning styles in ALS based on eye tracking technology, because it represents one of the most informative characteristics of gaze behavior. The experimental results showed a high relationship among the Felder-Silverman Learning Style and the eye movements recorded whilst learning. Keywords: Adaptation Adaptive learning Eye tracking Learning style detection Copyright © 2019 Institute of Advanced Engineering and Science. All rights reserved. Corresponding Author: Inssaf El Guabassi, Faculty of Sciences, Abdelmalek Essaadi University, P.O.BOX 2121, Tetuan, 93000, Morocco. Email: elguabassi@gmail.com 1. INTRODUCTION Traditional education systems, which allow learner to learn independently without attending a classroom to meet the tutor, are unable to dynamically adapt to the learner’s needs. Subsequently they are unable to increase the output of the learners. In this respect, recently the concept of adaptation has become an important issue of research in learning area; indeed, providing adaptivity in learning systems helps learners to make the intended learning outcomes via a personalized way. This paper represents a continuation of our previous works carried out in the adaptive learning systems (ALS) [1-4], in which we provided adaptivity in ubiquitous learning systems based on learning styles of Felder-Silverman and learner context. However, a new adaptation problem has appeared, namely the automatic detection of learner learning styles. Learning styles are increasingly incorporated to enhance learning outcomes for learners, and therefore many research approaches are done in education area, in fact, most researchers agree that learning styles play a key role in this area. One of the most widely used models of learning styles is the Felder-Silverman Index of Learning Styles (ILS). This model is extremely simple and, furthermore, the results are easy to interpret. However, at the same time the questionnaires in general suffer from several disadvantages, both theoretical and practical. For example, it is extremely impossible to record changing of learning styles because it is impossible to repeatedly ask learners to complete the questionnaire for time and cost concern. Fortunately, this problem can be solved using automatic detection of learning styles.