面向生物危害风险评估的地理知识图谱构建  

Construction of Geographic Knowledge Graph for Biohazard Risk Assessment

作　　者：李辉[1] 邢帅[1] 郭松涛 陈丽[1] 戴莫凡 LI Hui;XING Shuai;GUO Songtao;CHEN Li;DAI Mofan(Institute of Geospatial Information,Information Engineering University,Zhengzhou 450001,China)

机构地区：[1]信息工程大学地理空间信息学院,河南郑州450001

出　　处：《测绘与空间地理信息》2025年第2期13-17,共5页Geomatics & Spatial Information Technology

基　　金：国家重点研发计划项目(2017YFC1200305);国家自然科学基金项目(41876105、41371436)资助。

摘　　要：在解析生物危害事件构成要素的基础上,本文采用自顶向下模式层构建和自底向上数据层构建相结合的方式进行了面向生物危害事件地理知识图谱的构建,其中模式层构建中进行了生物危害事件主要概念的分类分层和关系设计,数据层构建则针对不同数据设计了对应的知识抽取方法,并以图数据库进行结构化与非结构化数据的统一管理。最后,以郑州市新冠肺炎疫情事件为例,对所构建的知识图谱性能进行了实验分析。实验结果表明,本文构建的知识图谱可以清晰地表达生物危害事件中的对象与关系,预测事件发生的可能性和估算事件造成的危害程度,对于提升生物危害事件风险评估效能具有很好的支持作用。On the basis of analyzing the components of biohazard events,the knowledge graph was constructed by top-down pattern layer construction and bottom-up data layer construction.In the pattern layer construction,the classification,stratification,and relationship design of the main concepts of biohazard events were carried out,while in the data layer construction,the corresponding knowledge extraction methods were designed for different data,and the unified management of structured and unstructured data was carried out by using the graph database.Finally,taking the epidemic situation of COVID-19 in Zhengzhou as an example,the performance of the constructed knowledge graph was analyzed experimentally.The experimental results show that the knowledge graph constructed in this article can clearly express the objects and relationships in biohazard events,predict the likelihood of event occurrence,and estimate the degree of harm caused by the event,which has a good supporting role in improving the effectiveness of biohazard event risk assessment.

关 键 词：地理知识图谱 生物危害事件 Covid-19 致灾因子 

分 类 号：P208[天文地球—地图制图学与地理信息工程]