摘要
成果地质资料的知识化是地质资料智能化应用的前提,以成果地质资料为研究对象,讨论了成果地质资料知识化多个环节的实践方法。首先,以人工智能知识表示方法为基础讨论了地质资料知识化表示及语义融合,并结合成果地质资料特点,提出把地质资料图谱分为“学科图谱”和“应用图谱”的分类思想以及基于OWL语义扩充的知识融合思路,构建了知识表示、知识组成、融合语义为支撑的地质资料知识图谱的基础逻辑框架。其次,结合知识化技术发展和成果地质资料的多模态特点,把成果地质资料知识化由浅到深分为基于关键词、元数据、全文本及多模态等四个层次。最后,对成果地质资料知识化质量评价方法进行了探讨,提出以三元组网罗度指标和一致性指标两个方面对图谱质量进行评价的初步构想。论及地质资料知识表示、分类、融合、层次划分及质量评价等关键环节,对地质资料管理部门系统化开展地质资料知识化工作具有参考意义。
The knowledge-based application of geological data is a prerequisite for the intelligent application of geological data.Discusses the practical methods of multiple links in the process of knowledge conversion of geological data.Firstly,the classification of knowledge graph and semantic fusion of knowledge graph are discussed,and the classification of“subject graph”and“application graph”and the fusion idea based on OWL semantic expansion are proposed.Secondly,combined with the development of knowledge technology and the multimodal characteristics of geological data,the knowledge conversion of geological data is divided into four levels based on keywords,metadata,full text and multimodality from shallow to deep.Finally,the evaluation method of knowledge graph quality is discussed,and a idea is proposed to evaluate the quality of the graph in terms of coverage index and consistency index.Involves knowledge representation,classification,fusion,hierarchical division and quality evaluation of knowledge conversion,which is of reference significance for the systematic development of geological data conversion work.
作者
廉永海
王斌
胡瑞斌
傅俊鹤
吕腾达
LIAN Yonghai;WANG Bin;HU Ruibin;FU Junhe;LYU Tengda(Beijing Yidaotongren Network Technology Co.,Ltd.,Beijing 100036,China;Command Center of Natural Resource Comprehensive Survey,China Geological Survey,Beijing 100055,China;Development and Research Center,China Geological Survey,Beijing 100037,China;Inner Mongolia Geological Survey and Research Institute,Hohhot 010010,China;Zhejiang Institute of Geosciences,Hangzhou 311203,China)
出处
《中国矿业》
2023年第6期175-182,共8页
China Mining Magazine
基金
中国地质调查局项目资助(编号:DD20221785)
内蒙古自治区地质调查研究院本级综合项目(2022)资助。
关键词
地质资料
自然语言处理
知识图谱
知识化
语义融合
geological data
natural language processing
knowledge graph
knowledge conversion
semantic fusion
