Medical professionalism is fundamental to ethical and competent healthcare, with early lapses during medical school linked to future disciplinary actions in clinical practice. It is essential to identify and analyze the factors contributing to professionalism issues among medical students. However, existing assessment methods rely on subjective evaluations, making them susceptible to bias, time-consuming, and retrospective rather than proactive. To address these limitations, this study develops an AI-driven method to identify medical professionalism profiles and determine key factors of professionalism issues. First, we build a classifier to identify professionalism lapses. Subsequently, the XAI method is applied to extract the main features. Then, we use a clustering algorithm to identify professionalism profiles. Finally, expert interviews with medical educators and hospital administrators can guide the development of tailored interventions based on AI-generated professionalism profiles. This study introduces a transparent AI-driven framework for professionalism assessment, enabling tailored interventions to support the development of ethical and competent physicians.