Intelligent Learning Behavior Analysis for Student Based on Fuzzy Agent Model

Learning style plays a significant role in improving a student's educational performance and finding strategies to adapt to new concepts. It is the process of evaluating the way of learning of individuals. Several factors, such as interaction with fellow students and teachers, emotional and social influences, interaction with the environment, teaching style, and personal characteristics, play vital roles in determining a student's learning behavior. This study introduces a methodology to determine student's learning styles automatically using a fuzzy agent model. Fuzzy logic is a good solution for handling uncertainty and vagueness and is therefore suitable for use in an intelligent learning behavior model. However, in a model that contains human behavior, complexities arise in the interaction of the factors in the system. Agent-oriented modeling can handle system complexities that are not considered in fuzzy logic. In this paper, we analyze students' behavior, present the concept of fuzzy logic and agent-based modeling, and finally develop a fuzzy agent model to implement an intelligent model. This implementation will help students increase interaction with education and knowledge, expand awareness of the student about their suitable way to learn, and enlarge student performance. The main contribution of this paper is to develop a fuzzy agent model and determine an individual's learning style automatically and accurately, which supports the student in acquiring knowledge appropriately and the teacher in suitably teaching the students. For this purpose, a real dataset is extracted from final-year CSE students at the University of Barishal, Bangladesh, that contains 35 students' required information. The proposed model predicts learning styles with 91% accuracy, which is satisfactory for contributing to the study of learners.