SQLBOOSTER: A TECHNOLOGYENHANCED LEARNING APPROACH FOR STUDENT MONITORING AND ADAPTIVE SQL LEARNING

Abstract

ABSTRACT This thesis presents SQLBooster which is a web-based learning management tool, that is designed to enhance SQL proficiency among the undergraduate students. The system addresses the limitations of traditional e-learning platforms, which often lack personalization and fail to monitor student engagement or psychological state. SQLBooster uses a multi-agent AI architecture, and uses technologies like Large Language Models, Retrieval-Augmented Generation, and FastAPI to enable a dynamic learning environment that is tailored according to the individual needs of the students. A standout innovation of SQLBooster is its emotion-based tension detection model, which uses the focus details obtained from head-pose detection model and attentiveness details obtained from the attentiveness model to categorize the detected student emotions into four distinct levels of tension. This provides critical insights into the psychological states of the students, and helps the lecturers to identify whether external factors like stress may affect the performance. SQLBooster also sets itself apart by offering comprehensive personalization features that are not collectively available in the existing tools like adaptive content delivery, personalized and progressive questioning, real-time feedback, and advanced partial grading for both DDL and DML SQL queries. It also provides the lecturers with details such as the common areas of difficulties among the students. SQLBooster was developed using a pragmatism-guided design science methodology, ensuring both theoretical innovation and practical applicability. The tension detection model demonstrated high accuracy, and chi-squared analysis from pilot evaluations confirmed a statistically significant improvement in students’ self-reported SQL confidence. Feedback from users was overwhelmingly positive, highlighting the system’s impact on engagement and learning. These findings suggest that SQLBooster effectively supports both learners and educators by combining adaptive learning with real-time affective monitoring. Keywords: personalization, partial grading, student monitoring, head-pose detection, emotionbased tension detection, attentiveness detection, learning analytics dashboard, feedback, SQLBooster, SQL education, learning management system.