Integrating a Customized AI Chatbot in Database Design and Programming Course

Abstract

This research-to-practice full paper describes our initiative to integrate a customized AI conversational agent in our Intro to Data Management course in Fall 2024 at Purdue University in the United States to support students with concept comprehension and in-class activities. This work uses a Retrieval-Augmented Generation (RAG) based customized chatbot using the Mindjoy application to provide students with precise and relevant responses. We customized the chatbot's capabilities by incorporating domain-specific knowledge and adding conversational constraints. This tailored chatbot was implemented in the CIT 21400 (Intro to Data Management) course to function as a tutor, enabling students to ask questions and receive customized answers within the context of their introductory database management curriculum. To evaluate its effectiveness, we conducted anonymous surveys based on the Technology Acceptance Model (TAM) and Self-Determination Theory (SDT). TAM was used to assess the perceived usefulness and ease of use with respect to the adoption of this new technology by students, while SDT offered insights into how the chatbot supported students' choices in learning (autonomy), their confidence in understanding the concepts of the database (competence) and their participation in the course material (relatedness). The surveys aimed to evaluate the effectiveness of the AI tutor in the undergraduate course and the overall value of the comprehensive AI experience for students. Results indicated positive outcomes across all evaluation criteria, reinforcing TAM's hypothesis that perceived usefulness enhances the adoption of this technology. Students also provided constructive feedback for further improving the chatbot, highlighting key areas such as enhanced contextual awareness and greater adaptability to complex queries, though findings are exploratory given the limited sample.

Ähnliche Arbeiten