AI Tutor Enhanced with Prompt Engineering and Deep Knowledge Tracing
Radhika Makharia, Yeoun Chan Kim, Su Bin Jo, Min Ah Kim, Aagam Jain, Piyush K. Agarwal, Anish Srivastava, Anant Vikram Agarwal, Pankaj B. Agarwal · 2024
The evolving educational landscape necessitates creative solutions to address the demand for immediate and personalized academic support. This study explores the integration of prompt engineering of the OpenAI’s Generative Pre-trained Transformer (GPT) and Deep Knowledge Tracing (DKT) to develop an AI tutor capable of shaping responses to students’ knowledge levels, promoting a dynamic and adaptive learning experience. By leveraging Large Language Models (LLMs) like GPT-3.5 and integrating DKT, our AI tutor addresses the need for real-time, tailored academic assistance. LLMs serve as virtual instructors, explaining concepts and providing detailed solutions, while DKT ensures responses align with the student’s knowledge level, optimizing challenge and engagement. Our research introduces an AI tutor that revolutionizes personalized learning experiences. Students can interact with the AI tutor by shaking their device during quizzes, initiating customized assistance and encouraging a deeper understanding of concepts, ultimately enhancing academic performance through individualized learning experiences.