International Journal of
Information and Education Technology

Abstract

Advances in educational technology are reshaping learning, with Intelligent Tutoring Systems (ITS) offering personalized education. However, creating high-quality, adaptive content remains a significant challenge for educators, requiring substantial time and effort. This research explores the potential of Large Language Models (LLMs), such as GPT-4, to automate content generation, addressing these challenges and enhancing educational efficiency. LLMs, leveraging deep learning and transformer architectures, are capable of generating human-like, contextually relevant text. By fine-tuning these models, this study investigates their application in producing diverse educational materials, including lesson plans, quizzes, and study guides. The system employs prompt engineering to ensure adaptability and alignment with learner needs. Evaluation results demonstrate promising outcomes. Content generated by the system achieved a 96% accuracy rate, overcoming common issues like hallucination, while surveys indicate an 85% likelihood of educator adoption. These findings underscore the potential of AI-powered tools to reduce the workload for educators, enabling them to focus on meaningful student engagement and tailored teaching strategies.