An Immersive Mobile Learning Environment and Cross-Cultural Interactive Teaching Model for the Transmission of Ethnic Music

Abstract

The living transmission of ethnic music is challenged by a fundamental disconnect between traditional oral and demonstration-based transmission practices and the cognitive paradigms of learners in the digital era. To address this challenge, a dual-engine immersive mobile learning framework was proposed, in which a context-aware dynamic graph attention recommendation engine and a multimodal sensor fusion-based interaction mechanism were integrated through a closed-loop data flow. In the recommendation engine, cultural background, real-time context, and skill progression were jointly modeled as a dynamic heterogeneous graph, and threedimensional personalized learning-path recommendations were generated through dual-channel graph embeddings and dynamic edge attribute updating. In the interaction mechanism, multimodal data acquired from accelerometers, gyroscopes, microphones, and cameras were fused, while embodied interaction, audio analysis, and augmented reality contexts were integrated through task-adaptive cross-attention to create an immersive learning experience. A shared optimization objective was adopted by the two modules, and synergistic enhancement of content adaptation and experiential immersion was achieved through joint training. Evaluation on a mobile ethnic music learning dataset comprising 200 cross-cultural learners demonstrated that the proposed model significantly outperformed the baseline methods in recommendation accuracy, immersion level, and cultural identity gain. These findings provide a feasible technological paradigm for mobile education aimed at the transmission of intangible cultural heritage.