基于保险条款文本的知识图谱构建研究  

Research on Knowledge Graph Construction Based on Insurance Clause Texts

作　　者：王浩畅[1] 宗杨 WANG Haochang;ZONG Yang(School of Computer and Information Technology,Northeast Petroleum University,Daqing 163318)

机构地区：[1]东北石油大学计算机与信息技术学院,大庆163318

出　　处：《计算机与数字工程》2024年第9期2759-2763,2836,共6页Computer & Digital Engineering

基　　金：国家自然科学基金项目(编号:61402099,61702093)资助。

摘　　要：为了方便人们在投保时能够更快速地对比不同产品的优缺点,论文提出一种基于保险产品条款文本的知识图谱构建方法。首先结合保险领域专家对保险产品设计相关的研究,分析条款中的要素并构建Schema层,接着将保险条款文件作为数据基础,使用BERT-BiLSTM-CRF模型抽取出其中的保险公司、保险产品、保险保障等实体,按照Schema中定义的实体关系和实体属性构造关系和属性的提取模板,最后将抽取出来的保险三元组存储至图数据库Neo4j中构成知识图谱。该方法能够快速有效地将条款文本转换为结构化的知识图谱,有利于推动保险领域智能化升级。In order to facilitate policyholders to compare the advantages and disadvantages of different products more quickly when purchasing insurance products,this paper proposes a knowledge graph construction method based on insurance product clause texts.Firstly,combined with the research of insurance experts on insurance product design,this paper analyzes the elements in the terms and build the schema layer,then uses the insurance term files as the data basis,and uses the BERT-BiLSTM-CRF model to extract the insurance companies,insurance products,insurance guarantee and other entities.According to the entity relationship and entity attribute defined in the schema,the extraction template of relationship and attribute is constructed.Finally,the extracted entity,attribute and relationship are stored in the graph database Neo4j to form a knowledge map.This method can quickly and effectively convert clause text into a structured knowledge graph,which is conducive to promoting the research of intelligent upgrading of insurance.

关 键 词：保险 知识图谱 命名实体识别 Neo4j 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]