Model rangkaian neural bagi penentuan gaya pembelajaran pelajar berasaskan model Felder-Silverman

Abstract

Student learning style is the tendency of a student's approach to learn. Most of the conventional methods to determine students’ learning style are by inquiry and survey methods that require students to answer several questions. This conventional method has several disadvantages such as requiring an amount of time to answer questions, questions are answered casually, low quality of survey questions and a different way each student interprets the questions given. With the development of today's e-learning technology such as the digital learning management system (LMS), records of student activities and interactions with the learning system can be stored online. This study aims to examine the relationship between students’ activities in a LMS in determining their learning style. Based on the Felder-Silverman learning style theory, an artificial intelligence model has been developed to automatically determine the learning style dimensions by using student activity records in the LMS. Artificial intelligence models are built based on neural network algorithms and the supervised machine learning technique. These neural network models have been trained using records of how often a student visits learning content such as lecture notes, learning videos, teaching slides and online exercises. As a result, four neural network models have been produced, each of which represents the dimensions of the Felder-Silverman learning style, namely information reception, information delivery, information processing and information organisation. The accuracy of the model obtained is between 77.8% to 100%.