T6-O/LT1-2 - RPOIA: A Method of selecting Learning Objects using Petri Nets.

1. Innovative Practice Full Paper
Joethe Moraes de Carvalho1 , Josias Gomes Lima1, José Francisco de Magalhães Netto1, Ruiter Braga Caldas1
1 Federal University of Amazonas

This Research Full Paper presents a method to select Learning Objects based on students' cognitive profiles, aiming to facilitate understanding of explained content and improve the programming skills. The STEM approach has favored the development of programming skills in Computer Science courses, as well as in technical computer courses. The discipline of Programming Logic has been the main means of applying these techniques, trying to disseminate Computational Thinking. Due to complexity of initial concepts, many students find it difficult to understand the proposals of this discipline, causing demotivation or course abandonment. The method we propose uses Petri Nets formalism to select a Learning Object, which are digital resources designed to assist in teaching and can be a video, an audio or a game, which addresses the subject being studied. Petri nets are formal description techniques used to specify competing systems through graphical and mathematical modeling. Intelligent Agents are computer programs created to automate and perform a certain task on a network and can even monitor web pages, databases, discussion forums and others. They can perform a direct interaction with the student and propose solutions considered appropriate, based on the knowledge obtained during interactions. To carry out this research, the Petri Net was used to create an Intelligent Agent model, which chooses the Learning Object from the answers obtained by applying a questionnaire. The approach evaluation was applied in a class with 40 students from a programming technical course. The questions were based on the VAK (Visual, Auditive and Kinesthetic) learning model, where each question presents three answer options, corresponding to one of three VAK learning modes. At end, the item with highest score represented the student's dominant learning form and a Learning Object was appointed according to detected profile. Each student used proposed artifact and then gave an opinion on method functionality. The results obtained showed the method's effectiveness, because, beyond to being well evaluated by the students, the system selected the object compatible with 72.5% of initial participants opinion.