생성형 AI와 프로세스 마이닝을 활용한 기초 및 임상 치위생학 학습 패턴 비교 연구

A Comparative Study on Learning Patterns in Basic and Clinical Dental Hygiene Using Generative AI and Process Mining

Objectives: This study analyzed self-regulated learning (SRL) patterns in dental hygiene education using process mining and generative AI, comparing learning behaviors between basic and clinical courses to develop optimized educational approaches. Methods: Learning activities based on national examination questions were implemented with students (n = 10) participating in AI-assisted learning sessions structured on Zimmerman's SRL theory. Student-AI interactions were coded into four categories: definition (D.G), exploration (S.S), problem-solving (C.A), and reflection (E.RV). The data were analyzed using Alpha, Heuristic, Fuzzy, and Inductive Miner algorithms. Results: Problem-solving (D.G or C.A) emerged as the primary learning activity in both courses, with reflection (E.RV) commonly observed. Clinical dental hygiene showed balanced SRL phases, with effective use of exploration (S.S) and advanced problem-solving (C.A). Basic dental hygiene demonstrated insufficient exploration phases and prominent problem-solving/reflection loops. Process mining revealed exploration-reflection loops in clinical courses versus definition-reflection patterns in basic courses. Conclusions: Process mining provides educators with data-driven insights into students' SRL patterns, enabling personalized teaching method development. Educators can identify bottlenecks and intervention points to build effective feedback systems. Students using generative AI can recognize their learning patterns and develop metacognitive strategies, enhancing their integration of theory and practice, clinical problem-solving, and lifelong learning abilities.