以泸县宋代石刻纹样为载体，挖掘宋代石刻纹样知识体系，并构建了宋代石刻纹样知识图谱。在划分泸县宋代石刻种类，解析35类可移动文物核心元数据及结合专家知识的基础上，引入纹样元数据和纹样知识体系，对泸县宋代石刻纹样进行了深度挖掘，并搭建了一套基于泸县宋代石刻纹样的知识图谱构建路径：知识图谱设计、标签化（关键词提取）、结构化（命名实体识别、关系提取）、知识图谱存储、可视化，为传统文化数据的标签化和结构化提供解决方案。

This paper takes the Song Dynasty stone carving patterns in Luxian County as the carrier, excavates the knowledge system of Song Dynasty stone carving patterns, and constructs the knowledge map of Song Dynasty stone carving patterns. On the basis of the classification of the stone carvings of the Song Dynasty in Luxian County, the analysis of the core metadata of 35 types of movable cultural relics and the combination of expert knowledge, pattern metadata and pattern knowledge system are introduced to deeply excavate the stone carvings of the Song Dynasty in Luxian county. Furthermore, a set of knowledge graph construction paths based on the Song Dynasty stone carvings in Luxian county are built in this paper: knowledge graph design, tagging (keyword extraction), structuring (named entity recognition, relationship extraction), knowledge graph storage, visualization, which provide solutions for labeling and structuring traditional culture data.