SURVEY AND CONCEPTUAL DESIGN OF DATABASE-CENTRIC ADAPTIVE INTELLIGENT TUTORING SYSTEM
Ana Morača · Emerging Trends in Industrial Engineering · 2025
This study analyzes the relational-database architectures of five leading intelligent tutoring platforms—CTAT's Cognitive Tutor, CARLA's context-aware recommender, TutorGen's generative framework, the OpenEdX LMS, and a lightweight commercial ITS—and distills from them a minimal prototype. By identifying shared schema elements, the paper proposes a design based on three core tables (Learner, Content, Interaction) and a single parametric ranking function that recommends exercises. The prototype is then contextualized using industry case studies from Carnegie Learning, Khan Academy, and Duolingo, each reporting measurable learning gains. Essential data-privacy measures are addressed, and it is demonstrated how this research-oriented design can scale to full-production ITS deployments. Finally, the study outlines key privacy safeguards and illustrates how this lean, research-driven design can be adapted for large-scale production environments.