地质科学2024，Vol.59Issue(2)：588-602,15.DOI:10.12017/dzkx.2024.041
面向地质图的知识图谱构建及智能问答应用
Geological map-oriented knowledge graph construction and intelligent Q&A application
摘要
Abstract
Massive geological maps contain rich basic knowledge of geology and expert experience knowledge.Geological maps mainly express geological knowledge of the Earth's surface(e.g.,stratigraphic units,rock bodies,fractures,etc.)acquired through regional geological surveys and mineral geological surveys.How to quickly extract geological knowledge from vector geological maps and form knowledge services is currently the forefront of research on geoscientific knowledge graph and knowledge services.Traditional geological map knowledge extraction mainly relies on manual methods for comprehensive analysis.This paper focuses on the research of knowledge expression and extraction of vector geological map,proposes a geological map knowledge expression framework,extracts geological entities and relationships contained in geological maps,expresses geological map information in the form of knowledge graph,and carries out intelligent question and answer applications based on geological vector knowledge map.Based on the experimental validation analysis of the vector dataset of Yinkeng Formation in Yudu County,Jiangxi Province.Finally,the experimental validation analysis is carried out with the vector data set of Yinkeng in Yudu County,Jiangxi Province,and the results show that the method of this paper can obtain the information of each geological object in geological maps in a more comprehensive way,improve the effect of semantic expression of geological maps,and at the same time improve the understanding and knowledge of geological learners about geological maps,so that computers can obtain the knowledge content of geological maps on a large scale.关键词
地质图知识表达模型/地质知识图谱/地质矢量图件/智能问答/空间认知Key words
Geological map knowledge expression model/Geological knowledge graph/Geological vector map/Intelligent question and answer/Spatial cognition分类
天文与地球科学引用本文复制引用
段雨希,邱芹军,田苗,马凯,谢忠,陶留锋,刘俊杰..面向地质图的知识图谱构建及智能问答应用[J].地质科学,2024,59(2):588-602,15.基金项目
国家自然科学基金项目(编号:42301492)、国家重点研发计划项目(编号:2022YFB3904200,2022YFF0711601)、湖北省自然科学基金项目(编号:2022CFB640)和地质探测与评估教育部重点实验室主任基金项目(编号:GLAB2023ZR01)资助 （编号:42301492）
