摘要
产业链知识图谱在金融领域应用较为广泛,但目前多数研究是面向单一产业的知识图谱构建或面向产业竞争情报服务,并没有把产业链和知识图谱有机地结合起来。本文从产业链应用角度出发,对产业链知识图谱的构建方法进行研究。首先提出了产业链知识图谱的构建流程和本体库,再基于领域语言模型,实现知识分类、抽取、融合等金融领域文本处理方法,对海量的领域文本进行知识抽取和融合,最终成功构建产业链知识图谱。根据本文方法构建的产业链知识图谱系统,覆盖产业链78个,细分行业7629个,已经应用到投融资、监管和产业规划等多个重要场景中。
Industry chain knowledge graphs are widely used in the financial field,but most of the current studies are based on single-industry knowledge graphs or industrial competitive intelligence services,and these have not organically com‐bined the industry chain and knowledge graph.From the perspective of the application,this paper examines the construc‐tion method of the industry chain knowledge graph.First,the construction process and ontology database are proposed.Based on the domain language model,the financial domain text processing methods such as knowledge classification,ex‐traction,and fusion are realized,massive domain texts are extracted and integrated,and the industrial chain knowledge graph is successfully constructed.The industry chain knowledge graph system constructed according to the method herein covers 78 industrial chains and 7629 subdivided industries,which is applied to many critical financial activities such as in‐vestment and financing,supervision,and industrial planning.
作者
毛瑞彬
朱菁
李爱文
周倚文
潘斌强
岳琳
Mao Ruibin;Zhu Jing;Li Aiwen;Zhou Yiwen;Pan Binqiang;Yue Lin(Center for Studies of Information Resources,Wuhan University,Wuhan 430072;Shenzhen Securities Information Co.,Ltd,Shenzhen 518022;Department of Management and Economics,Tianjin University,Tianjin 300110)
出处
《情报学报》
CSSCI
CSCD
北大核心
2022年第3期287-299,共13页
Journal of the China Society for Scientific and Technical Information
基金
国家重点研发计划项目“亿级节点时序图谱实时智能分析关键技术与系统”课题五“金融时序知识图谱查询与分析平台及应用验证”(2020AAA0108505)。
关键词
产业链
知识图谱
领域语言模型
实体和关系联合抽取
industry chain
knowledge graph
domain language model
entity and relation joint extraction
