基于文献数据规律的机构知识库关键技术研究——以北京邮电大学机构知识库为例

Research on the Key Technology of Literature Data Regularity-based Institutional Repository:A Case Study of Institutional Repository in Beijing University of Posts and Telecommunications

文章针对目前国内高校图书馆构建机构知识库普遍存在关联学者、机构、学科或主题不精准等问题,以北京邮电大学图书馆通过研究文献数据预处理的关键技术来探索文献数据中的规律为例,设计了文献实体关系模型并进行高校实体的扩展,在图书馆信息化管理的顶层设计下自主研发了机构知识库。通过认领作者机构的新模式,提出了文献与实体之间可持续的间接精准关联机制,为学术生态圈的形成和深层次学科服务的开展提供了数据支持。

Aiming at the problem of inaccurate data correlation with scholars, institutions, subjects or topics which is widespread in institutional repositories of domestic university libraries, Library of BUPT develops key technologies of preprocessing to explore the regularities in bibliographic data, and designs paper - entity relationship model which is expanded by colleges＇ entities, and develops institutional repository based on top - level design of information management. By a new mode of identifying author affiliations, a sustainable correlation mechanism for indirect and accurate correlation between paper and entity is proposed, and accurate data correlation provides the support for academic ecosystem and profound subject service.