关联数据发布流程与关键问题研究——以科技文献、科学数据发布为例

A Research on Publishing Workflow and Key Issues of Linked Data: Experience with Publishing Scientific Literature and Scientific Data as Linked Data

本文结合中国科学引文数据库和中国生态系统研究网络通量数据的关联数据发布,以关联数据的发布技术框架为研究对象,采取实例阐释的方法,提出了关联数据发布过程中可参考的标准化流程,并详细分析了其中的关键问题。研究表明,关联数据发布流程可以分解成数据建模、实体命名、实体RDF化、实体关联化、实体发布、开放查询六个关键步骤,发布过程中需要考虑到多语种问题、值词表的发布、RDF词表的发布等关键问题。关于利用D2R Server发布数据,本文建议:不要采用空白节点;尽量做好关系型数据库的前期设计;指定非文本属性的数据类型;适当进行实体表的拆分与合并。图5。表3。参考文献31。

Combined with the publication of linked data in Chinese Science Citation Database （CSCD） and the flux data in Chinese Ecosystem Research Network （CERN）, this paper studies the standardized publishing workflow of linked data and key issues in such process by studying on the technical framework of publishing linked data with examples. The standardized pub- lishing workflow of linked data consists of such six critical steps as data modeling, entity naming, turning entity into RDF, make connections between entities, entity publication and open-type inquiry. Some key issues need to be considered in the publishing process, such as dealing with multilingual values, publishing value vocabularies and publishing RDF vocabularies. In addition, this paper gives some suggestions in case IY2R Server is used to publish linked data： not using blank note; trying the best to make pre-design of relational database better; specifying the data type of non-text attributes; splitting and combining table enti- ties properly. 5 figs. 3 tabs. 31 refs.