基于本体和图数据库的流域干旱灾害风险知识图谱构建及应用
Construction and Application of Knowledge Graph Concerning Drought Risk in River Basin Based on Ontology and Graph Database
张欢 1易善桢 1冷创 1吴琪1
作者信息
- 1. 华中科技大学 数字流域科学与技术湖北省重点实验室,湖北 武汉 430074;华中科技大学 土木与水利工程学院,湖北 武汉 430074
- 折叠
摘要
针对干旱灾害领域内数据多源异构、难以统一管理与分析的问题,从知识图谱理论出发,首先根据领域知识要素,自顶向下构建干旱灾害风险本体;其次基于全球离散格网系统设计了一种知识抽取方法,并以汉江流域丹江口上游为例,采用该方法从流域时空数据中提取出知识三元组后,利用图数据库进行存储,形成流域干旱灾害风险知识图谱;最后以所建图谱作为空间分析框架,结合游程理论、聚类等方法实现了对流域干旱灾害风险的综合评估.
Abstract
In order to address the problem of heterogeneous and difficulty in unified management and analysis of multi-source data in the field of drought disasters,this article starts from the theory of knowledge graph.Firstly,the drought disaster risk ontology was constructed based on the domain knowledge elements.Secondly,a knowledge extrac-tion method was designed based on the global discrete grid system.Taking the upstream area of Danjiangkou in the Han-jiang River Basin as an example,this method was used to extract knowledge triplets from spatiotemporal data in the ba-sin.Then the triplets were stored using a graph database,thus forming the knowledge graph concerning drought risk in river basin.Finally,the constructed graph was used as a spatial analysis framework,combining methods such as run-length theory and clustering were used to achieve comprehensive assessment of drought disaster risk in the basin.
关键词
知识图谱/旱灾风险/知识抽取/聚类分析Key words
knowledge graph/drought risk/knowledge extraction/clustering analysis引用本文复制引用
基金项目
重点交叉研究项目(2016JCTD115)
出版年
2024
NETL
NSTL
万方数据