This study designs and implements a multimodal data fusion-oriented distribution system around the intelligent distribution needs of cultural and educational resources. The system comprehensively handles multimodal educational resources such as text, image, audio, etc., integrates content features and user behavioral profiles, and constructs a set of end-to-end personalized recommendation mechanisms. Through the introduction of multimodal feature extraction and fusion technology, combined with deep learning recommendation algorithms, it realizes accurate pushing and dynamic updating of educational resources. The system architecture covers core modules such as data acquisition, preprocessing, feature fusion, recommendation engine and user interface, and has good scalability and adaptability. In the actual application test, the system performs well in terms of resource matching accuracy, response speed and user satisfaction, which verifies its potential and practical value in the field of digital education and cultural communication.
Feng et al. (Sun,) studied this question.