Desenvolvimento de uma plataforma gamificada para estudos com extração automatizada de questões do ENEM e integração com inteligência artificial

Abstract

This work presents the development of an educational gamified web platform, designed for studying for the Brazilian National High School Exam (ENEM). The platform automates the extraction of questions, alternatives and answer keys from the PDF the ENEM, using text processing techniques with regular expressions (regex) in Python. The system consists of a backend developed with Django and Django REST Framework, which provides a RESTful API for managing users, questions, and the game logic, including secure JWT authentication. Data persistence is handled by a PostgreSQL database. The frontend, built with React.js, offers an interactive interface where users progress through a system of ten phases, consume energy to access questions, and accumulate crystals to request contextual hints generated by integration with the ChatGPT API. Gamification elements such as a global ranking and achievements were implemented to increase engagement. The results demonstrated the effectiveness of regex extraction for the specific format of the ENEM, with all 95 questions and 475 alternatives being correctly categorized. However, the approach proved to be dependent on the PDF's structure and unfeasible for the robust extraction of images, pointing to the need for AI-based solutions in future work. The functional platform serves as a solid foundation for evolution in learning personalization and intelligent tutoring systems, highlighting the potential of the intersection between gamification, automation, and artificial intelligence in education.