Sistema inteligente de apoio ao ensino de Dedução Natural

Abstract

This paper proposes the development of a tool to support the teaching-learning process of Mathematical Logic contents, taught in Computer courses. Formal proof systems are worked throughout these disciplines, with an emphasis on Natural Deduction. The for malization of sentences and the construction of models in the studied logics are not con sidered difficult by the students, but the introduction of formal proof content causes the level of difficulty to increase considerably. Thus, it was identified the opportunity to de velop a multiplatform tool that chooses a learning path for each student, depending on their level of knowledge, and assists in solving the proposed exercises. The tool was ap plied to the context of natural deduction in propositional logic. The methodology of the work consists of a systematic literature review, mainly addressing the concepts related to intelligent tutoring systems, followed by the proposal and development of the tool, which includes the survey of system requirements, use case definition, architecture modeling, implementation, testing and system validation. The implemented system achieved a sat isfactory result, with its essential requisites beint totally met, integrating the pedagogical concepts to create an intelligent tutoring system with the characteristics provided for this type of tool. Through testing, it is observed that the learning sequence is tailored to the student’s performance. Through the assessment performed by a professional in the area covered by the system, it was considered fit to be used as a support for teaching in the subjects in which it is proposed, meeting its main objective.