面向共享的地学数据语义标签提取与推荐方法  

Semantic Tag Extraction and Recommendation Method for Geoscience Data Sharing

作　　者：王敬悦 王卷乐[2,3] 韩保民 张敏 李凯[5] WANG Jingyue;WANG Juanle;HAN Baomin;ZHANG Min;LI Kai(Shandong University of Technology,School of Civil and Architectural Engineering,Zibo 255049;State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101;Collaborative Innovation Center for Development and Utilization of Geographic Information Resources in Jiangsu Province,Nanjing 210023;University of Chinese Academy of Sciences,College of resource and Environment,Beijing 100049;College of Geoscience and Surveying Engineering,China University of Mining&Technology(Beijing),Beijing 100083)

机构地区：[1]山东理工大学建筑工程学院,山东淄博255049 [2]中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室,北京100101 [3]江苏省地理信息资源开发与利用协同创新中心,江苏南京210023 [4]中国科学院大学,北京100049 [5]中国矿业大学(北京)地球科学与测绘工程学院,北京100083

出　　处：《中国科技资源导刊》2023年第2期83-93,共11页China Science & Technology Resources Review

基　　金：国家重点研发计划项目“地球表层系统科学数据挖掘与知识发现关键技术与应用”(2022YFF0711600);中国科学院战略性先导科技专项(A类)“地球大数据资源分类与多维标签管理系统研发”(XDA19090200,XDA19040501)。

摘　　要：地球科学数据具有丰富的语义信息,这为探索地学奥妙带来广阔空间的同时,也为数据共享带来了挑战。缺少语义的认知和关联使得用户难以从复杂、海量、多源、异构的地球科学数据中发现符合自身需求的数据。以中国科学院《地球大数据科学工程》的共享数据集为研究对象,对数据文本进行分词和标签提取,实现数据形式的统一,以达到多源异构数据的规范化管理。研究表明,通过对数据集实现标签提取、标签推荐以及知识图谱的构建,可促进海量地球科学数据的管理和精准服务。研究结果可为更多的科学数据共享平台提供借鉴。Geoscience data has rich semantic information,which not only brings vast space for geoscience exploration,but also challenges for data sharing.The lack of semantic cognition and association makes it difficult for users to find data that meets their own needs from complex,massive,multi-source and heterogeneous scientific geo-data.This paper took the shared dataset of the Big Earth Data Science Project of the Chinese Academy of Sciences as the research object.Through word segmentation and label extraction of data text,the data form was unified to achieve standardized management of multi-source heterogeneous data.The research practice showed that the management and precise service of massive earth science data could be promoted through the realization of label extraction,label recommendation and the construction of knowledge graph on the data set.This study result is expected to provide reference for more scientific data sharing platforms.

关 键 词：地球科学 数据共享 语义标签 相似推荐 知识图谱 元数据 

分 类 号：P208[天文地球—地图制图学与地理信息工程]