Abstract: 【Purpose/significance】In the context of the new era, to explore the knowledge graph construction process of traditional cul⁃ ture reading promotion in college libraries based on the big language model, to optimize the reading promotion path through the visual graph, to promote the deep excavation and systematic organization of traditional culture resources, and to innovate the reading promo⁃ tion mode in college libraries.【Method/process】This study proposes a knowledge graph construction process of traditional culture reading promotion content based on large language model. After collecting data, we first carry out data cleaning and processing, use BERTopic topic model to extract entity categories and relationship labels, and then use deepseek-r1 large language model for entity extraction and relationship recognition, and finally import the entity relationship triples into Neo4j graph database, which can be used to optimize the reading promotion path through visual graph. Finally, the entity-relationship triples are imported into the Neo4j graph database to achieve the visualization of the knowledge graph. In addition, we use the specific content of Baidu search and WeChat search platform to construct the knowledge map of traditional culture reading promotion in college libraries, and explore its innovative path mode in traditional culture reading promotion.【Result/conclusion】The results of the study show that the knowledge mapping method based on the deepseek-r1 model can clearly present the organization, activities and promotion strategies in traditional culture reading promotion, forming a structured knowledge expression system; further analysis of the knowledge content of traditional culture reading promotion can assist in the construction of the traditional culture reading case base, help build an intelligent Q&A system, as⁃ sist in the mining of traditional culture reading promotion paths, and assist in the construction of the traditional culture reading case base, and assist in the construction of the traditional culture reading case base. mining traditional culture reading promotion paths, in order to play the role of the position of cultural heritage in college libraries.【Innovation/limitation】A knowledge graph construction method of traditional culture reading promotion content based on big language model is proposed, and the knowledge elements are mined for traditional culture promotion in college libraries, but due to the limitation of computer resources, the big model interface of the third-party platform can be used in the future research.