基于知识图谱的标准知识管理研究

摘要：

行业标准化体系构建过程中，随着标准的种类与数量不断更新，人工加载以及查询的方式已经难以满足标准查新跟进，知识服务手段较为单一。知识图谱技术为整合标准知识提供了一种全新的知识互联思路，为完善标准化建设路径、标准文件结构化查询提供了全新的方向。本文分析了标准体系构建现存的问题，以食品安全国家标准为例搭建了以食品产品标准为中心的标准引用知识图谱，基于该图谱展示了其可视化检索、标准重要性等应用，进一步分析知识图谱这一知识管理形式在标准体系建设上的优势，促进标准知识智能服务与发展。标准领域知识图谱强调对标准知识管理，有效集成各类标准文本知识、梳理标准信息，同时结构化知识有益于精确标准查询和关联标准挖掘从而助力推动标准的数字化发展。

Abstract:

The types and quantities of standards are constantly increasing in the process of industry standardization system construction, and manual loading and querying have been difficult to meet the demands of rapid checking and following up of standards. The means of knowledge information service is relatively single. Knowledge graph technology provides a new knowledge interconnection idea for integrating standards knowledge, and provides a new direction for improving the standardization construction path and the structured query of standards documents. This paper analyzes the existing problems in the construction of the standards system, builds a standards citation knowledge graph centered on food product standards as an example, demonstrates its applications such as visual retrieval and standard importance based on this graph, further analyzes the advantages of knowledge management form of knowledge graph in the construction of the standards system, and promotes the standardsknowledge intelligent service and development. The standards domain knowledge graph emphasizes the intelligent management of the guiding normative text, effectively integrating various types of standard text knowledge and sorting out standards information, while the structured form can effectively improve the accuracy and efficiency of retrieval for the subsequent provision of accurate queries and intelligent mining of associated standards, thus helping to promote the digital development of standards.

基金项目：本文为浙江省大学生科研创新活动计划资助项目“基于知识图谱的食品安全标准智能问答机器人研究”（项目编号：2021R409047）的研究成果。

作者简介：杨德相，硕士研究生，研究方向为知识图谱、知识管理、金融欺诈防范。李剑锋，副教授，研究生导师，研究方向为商务数据分析、人工智能（机器学习）、大数据技术、管理信息系统。