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

中文关键词: 地球科学；数据共享；语义标签；相似推荐；知识图谱；元数据

英文关键词: Earth science, data sharing, semantic tags, similarity recommendation, knowledge graph, metadata

中文摘要:

      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。

英文摘要:

      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.