AI-powered knowledge organization: a next-generation approach to library classification using DeepSeek-R1

With the advancement of information technology, libraries have shifted from traditional physical services to an integrated "offline + online" model, becoming digital hubs in the national public cultural service system. However, current book classification still relies primarily on manual efforts, which suffers from inefficiency and inconsistent standards, making it difficult to meet the growing demand for processing massive volumes of books. Leveraging the latest developments in artificial intelligence, this paper proposes an automatic book classification algorithm based on the DeepSeek-R1-Distill model to improve classification accuracy and efficiency. Experimental results demonstrate that the algorithm achieves an average F1-score of over 87% in a 21-category Chinese book classification task, validating its effectiveness. Future work could explore the integration of more advanced large language models and domain-adaptive pre-training strategies to further advance classification capabilities. This research contributes a novel technical pathway toward intelligent book classification and underscores the significant potential of large language models in library and information science, offering both theoretical insights and practical value for the evolution of next-generation knowledge organization systems.