Hybrid Intelligent Tutoring System With Didactic Transposition Of The Subjects Guided By Expert Knowlegdment And Self Organizing Maps Neural Network

The development of Intelligent Tutoring Systems has been studied for decades and requires multi and interdisciplinary knowledge. In this context, this article presents a hybrid intelligent tutoring system in which the teaching strategies are based on neural networks, Self-Organizing Maps (SOM), and knowledge of a specialist teacher. This model has a reactive and adaptive characteristic that provides a personalized and individual instruction for the student, by promoting the necessary guidance on the transfer of didactic contents. This paper presents the development process of the proposed hybrid model, including the system with specialist guidance that ultimately provides the data used in the training stage of the neural networks. The results show the behavior of the system when using specialist guidance and when using the hybrid decision model. The results indicate that the application of the hybrid intelligent tutoring systems model is feasible because it includes both the teacher's knowledge and the student's behavior for establishing teaching strategies, which allows for a greater proximity of the tutor's and the student's actions. The performance of the proposed model was satisfactory compared to other systems proposed in the literature that use connectionism for establishing the teaching strategy.