From Representation to Relational Mentorship: User Perceptions of Virtual Heritage Mentors in AI-Mediated Intangible Cultural Heritage Learning

This study presents a convergent mixed-methods empirical investigation of user perceptions of Virtual Heritage Mentors (VHMs) in AI-mediated intangible cultural heritage learning. The study draws on an online survey of 370 participants and semi-structured interviews with 20 independently recruited participants, all of whom were ethnically Chinese. The results indicated that participants prioritized interactive explanation, clarification, and guided support over visual realism. Trust was associated with cultural accuracy, contextual grounding, transparency, and respect for heritage communities. Emotional resonance was valued less as anthropomorphic simulation than as pedagogical care expressed through patience, encouragement, re-explanation, and responsiveness. Participants also envisioned VHMs as adaptive learning companions while emphasizing that AI should complement rather than replace human practitioners. These findings empirically support and refine the proposed VHM framework by highlighting relational mediation, relational mentorship, co-transmission, and human–AI complementarity as key design priorities. The study argues for shifting digital heritage design from representation alone toward relational, culturally accountable support for heritage learning and transmission.