Ran LLM Driven Chatbot In Higher Education For Databases And Information Systems

Abstract

The integration of large language model (LLM) chatbots into Learning Management Systems (LMS) has the
potential to enhance the teaching and learning experience in higher education. This study investigates the
development, deployment, and evaluation of an LLM-based chatbot named MoodleBot designed for computer
science classroom environments. The chatbot aims to support self-regulated learning (SRL) and assist
students in seeking academic help through an interactive platform integrated with the Moodle LMS. Despite
challenges associated with artificial intelligence technologies, such as bias, hallucinations, and resistance from
educators toward adopting new tools, this research addresses two key questions: the level of student
acceptance of Moodle Bot as a learning support tool and the accuracy of its responses in relation to course
content. The chatbot was developed using a Retrieval-Augmented Generation (RAG) approach to ensure that
responses are generated based on relevant course materials. The Technology Acceptance Model (TAM) was
employed to evaluate user acceptance, focusing on perceived usefulness and ease of use. A total of 46
students participated in the study, with 30 completing the TAM questionnaire. The results indicate that
Moodle Bot achieved an accuracy rate of approximately 88% in providing course-related assistance. The
findings suggest that AI-driven educational chatbots can improve personalized learning experiences and
reduce instructors’ administrative workload, although further improvements in automated fact checking are
necessary to enhance reliability