POLIBATTLE: A RAG-BASED INTELLIGENT TUTORING SYSTEM WITH GAMIFICATION ELEMENTS FOR INNOVATING UNIVERSITY EDUCATION

Teaching complex subjects like Software Engineering requires innovative methodologies to maintain high student engagement. While integrating third-party gamification tools has proven effective in improving classroom participation, these solutions often present technical limitations, driving the need for advanced proprietary platforms. To address this, this thesis presents the design and development of PoliBattle, an interactive web application for students and professors at the Polytechnic University of Bari. PoliBattle synergizes the motivational dynamics of Gamification with the adaptive capabilities of Generative AI. Teachers can create virtual classrooms and administer real-time quizzes, while students are incentivized through leaderboards, unlockable achievements, and avatar customization. The core of the platform is an integrated Intelligent Tutoring System (ITS). Leveraging the Gemini API and Retrieval-Augmented Generation (RAG) architecture, the system processes the official course materials to provide real-time, context-aware explanations for both correct and incorrect quiz answers, explicitly citing its sources. To assess the platform's effectiveness, a validation phase was conducted with Software Engineering students using standardized metrics: the System Usability Scale (SUS), the Technology Acceptance Model (TAM), the User Experience Questionnaire (UEQ-S), alongside a custom evaluation for the AI tutor. The results demonstrated excellent usability, strong user acceptance, and a highly positive overall experience, with the ITS proving to be accurate and contextually relevant. Ultimately, PoliBattle provides a scalable educational ecosystem that offers students a tailored, interactive, and frustration-free learning experience, demonstrating the high effectiveness of combining gamified elements and AI in modern higher education.