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

Area of Science:

• Library and Information Science
• Artificial Intelligence
• Computer Science

Background:

• Libraries are evolving into digital hubs, necessitating efficient processing of vast book collections.
• Traditional manual book classification methods are inefficient and lack standardization.
• Information technology advancements drive the need for automated library services.

Purpose of the Study:

• To develop an automatic book classification algorithm for improved accuracy and efficiency.
• To address the limitations of manual classification in digital library environments.
• To leverage artificial intelligence for next-generation knowledge organization systems.

Main Methods:

• Implementation of an automatic book classification algorithm.
• Utilizing the DeepSeek-R1-Distill model for classification tasks.
• Evaluation of the algorithm on a 21-category Chinese book dataset.

Main Results:

• The proposed algorithm achieved an average F1-score exceeding 87% in Chinese book classification.
• Demonstrated significant improvements in classification accuracy and efficiency.
• Validated the effectiveness of the AI-driven approach for library science.

Conclusions:

• The AI-based algorithm offers a novel technical pathway for intelligent book classification.
• Large language models show significant potential in library and information science.
• The research provides theoretical insights and practical value for modern knowledge organization.