The efficiency and accuracy of digital library resource retrieval directly impact the quality of teaching and research in universities. This paper addresses issues such as low retrieval efficiency, insufficient personalized services, and low resource utilization in a university library’s resource retrieval system. An intelligent retrieval system based on a hybrid recommendation algorithm is designed. A layered architecture is used to design the system framework, implementing key functional modules such as user profile construction, resource feature extraction, and hybrid recommendation engine. System deployment and testing verify that the optimized system significantly improves retrieval accuracy, response time, and concurrent processing capability, with a noticeable enhancement in resource utilization. The system’s usability meets the design expectations, providing an effective solution to enhance digital library service levels.