LEVERAGING LARGE LANGUAGE MODELS FOR THE CREATION OF AN AI LECTURE CHATBOT AVATAR

Abstract

This work aimed to design an AI lecture chatbot by combining large language models (LLMs) and learning management systems (LMS). LLMs like ChatGPT are trained on vast data, enabling advanced natural language processing. A survey of 36 students from Edo State University Uzairue revealed that 80.6% were from the Faculty of Engineering, 40% were familiar with AI chatbots, and 85% supported AI in education. The chatbot uses a Retrieval Augmented Generation (RAG) pipeline where users upload files, which are processed into vector embeddings for generating accurate responses. Despite the small sample size and focus on engineering students, the study highlights the positive reception of AI in education. Future research should include a larger, more diverse sample to confirm these findings. This work shows the potential for AI chatbots to enhance education.