International Communication of Hainan Folk-Culture Terminology within a Cross-Modal AI Translation Framework

Abstract

This study examines the dual challenges facing the international communication of Hainan's folk culture in the context of globalization-namely, cultural discount and terminology distortion-as well as the inherent limitations of traditional translation paradigms. At the theoretical level, it reconstructs the methodological foundation by integrating terminology studies with intercultural communication theory. At the technical level, it proposes an innovative cross-modal AI translation framework that incorporates neural machine translation optimization, multilingual semantic alignment model fusion, and a dynamic annotation-incremental learning mechanism. Empirical validation focuses on the cultural events "Sanyuesan" and the "Junpo Festival." A multimodal corpus was constructed, and models were trained and evaluated, ultimately supporting the formulation of three international standards. The research outcomes have already been applied in real-world settings and have demonstrated significant effectiveness. Future research will focus on federated learning, blockchain, and metaverse applications to further deepen the technological framework, expand the application ecosystem, and enhance the international dissemination of Hainan's folk culture.