Optimization of Intelligent Classification and Precise Recommendation Algorithms for Reading Resources in Libraries
Si Zhong, Xiaorong Pei · Procedia Computer Science · 2026
With the popularization of digital reading and the massive growth of library resources, the problems of low efficiency and insufficient targeted push services in traditional reading resource classification methods have become increasingly prominent, making it difficult to meet users’ personalized reading needs. In response to the above issues, this article aims to improve the efficiency of library reading resource management and the quality of push services, and conducts research on intelligent classification and precise push algorithm optimization. This article first reviews the current research status in relevant fields and clarifies the shortcomings of existing algorithms in feature extraction, cold start processing, and other aspects; Secondly, a multidimensional resource feature system is constructed, and a hybrid optimization algorithm that integrates deep learning and collaborative filtering is proposed to achieve accurate classification and personalized push of resources; Finally, experimental validation was conducted using real library resource datasets and user behavior data. The experimental results show that the improved Attention BiLSTM algorithm proposed in this paper achieves the best performance in all indicators, with an accuracy improvement of 5.3% compared to the basic LSTM and a macro average F1 improvement of 7.7%, effectively improving the effectiveness of push services and user satisfaction. This article provides feasible technical solutions and theoretical support for the intelligent management and efficient service of library reading resources.