Research on the Construction of Digital Teaching Resource Library for Photoshop Course under Collaborative Filtering Technology

Aiming at the problem of personalized recommendation of teaching resources for Photoshop courses, a digital resource library based on collaborative filtering technology is designed. Using the method of matrix decomposition, the latent factor model in the network environment is constructed by linking the user's behavior and network popularity. This project intends to study a similarity calculation method that integrates users and resources, on the basis of which, accurate matching of multi-level personalized needs is carried out in order to improve the recommendation accuracy and retrieval efficiency. Experiments demonstrate that the method has good user satisfaction and diversity, with an accuracy rate of more than 80% for highly rated categorized resources. It can also improve the coverage in various recommendation environments and provide strong support for intelligent recommendation of educational resources.