ChatGPT-based Recommendations for Personalized Content Creation and Instructional Design with a Tailored Prompt Generator
Akrivi Krouska, Christos Troussas, Ioannis Voyiatzis, Phivos Mylonas, Cleo Sgouropoulou · 2024
Recent advancements in Artificial Intelligence (AI) have significantly influenced educational innovations, making learning environments more sophisticated with personalized learning, tailored feedback, and learning analytics. Among these advancements, Generative AI, namely ChatGPT, has emerged as a powerful tool in education. ChatGPT's ability to generate human-like content in real-time can assist educators with course design and educational material creation, allowing for enhanced instructional strategies without starting from scratch. Moreover, integrating ChatGPT into the recommendation process offers superior understanding of user needs and delivery of context-aware recommendations. In view of the above, this paper explores the integration of ChatGPT with a tailored prompt generator within an intelligent authoring system to offer effective recommendations for personalized content creation and instructional design. The system was evaluated and compared to its conversional version where instructors manually use ChatGPT. The findings indicate significant improvements in system usability, content quality, personalization, and time and effort savings, demonstrating the effectiveness of the proposed approach.