Focusing on the needs of digital protection and inheritance of folk music in the new era, this paper constructs a technical framework that integrates multi-source data collection, feature modeling, knowledge organization, intelligent service and application verification, and puts forward an innovative mode for the transformation of folk music from static digital storage to dynamic digital inheritance. Based on 4570 multi-modal samples, the study comprehensively uses computer technologies such as audio analysis, multi-modal fusion, knowledge graph, semantic retrieval and intelligent recommendation to realize the structured expression and intelligent service of folk music resources. The results show that the average recognition accuracy of the system reaches 90.6%, the Top-10 retrieval hit rate reaches 93.1%, the number of resource reaching users increases by 180.3%, the dissemination coverage index reaches 66.3%, and the learner task accuracy increases from 62.8% to 85.6% after 8 weeks of platform operation. The research shows that digital technology can effectively improve the preservation quality, transmission efficiency and inheritance support ability of folk music resources, and provide an implementiable technical path for the live inheritance of folk music.
