基于CiteSpace的地质大数据与人工智能研究热点及前沿分析

Research hotspots and cutting-edge analysis of geological big data and artificial intelligence based on CiteSpace

为研究地质学领域的大数据和人工智能研究现状、热点和前沿,在中国知网(CNKI)核心期刊和Web of Science(WoS)核心数据库收集了2000—2022年相关中文文献3600篇、英文文献1803篇,利用社区结构分析软件CiteSpace,从合作作者、研究国家、研究机构、关键词聚类、关键词时空分布图谱等进行可视化分析,并统计了2021—2022年间,地质学领域国际顶级期刊(综合影响因子10以上)的文献进行前沿分析。分析结果表明,近10年内该研究领域全球累计发文量激增,以中国为代表的亚洲国家和以美国为代表的欧美国家研究为主,双方累计发文量相差不大,论文中介中心性欧美国家普遍较高。我国研究机构之间的交流合作居多,与国外的研究机构交流合作较少,国外研究机构则与之相反。该领域以应用机器学习类方法、知识图谱构建等,在地质灾害防治、地震解释、石油与天然气勘查、固体矿产资源预测等方向进行的科学研究为研究热点,以深度学习、集成学习、智能平台搭建等为手段的地球演化过程中的重大地质事件研究、全球性气候变化、极地及海洋地质研究、数字地质建模及定量分析、地震预报、地灾易发性精准评估等为研究前沿。

To investigate the current status,hotspots,and frontiers of big data and artificial intelligence research in the field of geology,this study conducts literature screening of relevant research publications between 20002022 using China National Knowledge Infrastructure(CNKI)and Web of Science(WoS)core databases.A total of 3600 Chinese and 1803 English articles are collected,and community structure analysis software CiteSpace is used for visual analysis of cooperation authors,research countries/institutions,keyword clustering,and keyword spatiotemporal distribution maps.Furthermore,a stochastic frontier analysis correction is conducted on publications by international top-tier geoscience journals(comprehensive impact factor≥10)between 20212022.The global cumulative publication volume in this research field had surged in the past decade,led by Asian countries represented by China and European/American countries represented by the United States,with China and the United States showing no significant differences,and the betweenness centrality measures generally higher for European/American countries than for Asian countries.In China,research collaborations were mostly among domestic institutions and relatively rare with foreign research institutions,whilst the opposite was true in foreign countries.The research hotspots in this field were geological disaster prevention and control,earthquake interpretation,petroleum and natural gas exploration,and solid mineral resource prediction using machine learning and knowledge graphs.Research frontiers included significant geological events during Earth’s evolution,global climate change,polar and marine geology,digital geological modeling and quantitative analysis,earthquake prediction,and accurate assessment of geological disaster susceptibility by means of deep learning,integrated learning,and intelligent platform.