摘要
科学文献之间通过引用关系构成了特定研究主题的知识网络,其单向无回路的特征揭示了学科主题的知识结构和发展过程。本文以WOS数据库中XML研究论文所构成的引文网络为例,利用引文关系权重与文献节点权重确定核心文献,并在此基础上从阈值和权值"高地"两个角度对核心文献进行聚合。研究发现:文献核心程度的确定过程充分考虑了不同引用实质上的重要程度区别,据此计算得到的引文和文献节点权重能够准确反映文献的质量;阈值聚合能够迅速发现整个学科发展过程中最核心的文献和引文;权值"高地"聚合分析结果更为多样,并能弥补阈值聚合在揭示次重要子结构方面的不足,发现整个知识体系发展过程中丰富的研究维度。图6。表4。参考文献11。
Intereormected by citation, scientific literature construct knowledge network of researches on specific subjects. As an acyclic network, the knowledge network reveals the knowledge structure and development of research subjects. This paper uses the citation network formed by research literature on XML collected from WOS database in order to calculate the degree of core literature by weighing citation ties and literature vertices, and finally to present the cohesion of key literature according to the weights and highland. It is found that the core literature substantially reveals the importance for each citation, and the weights calculated from citation ties and literature vertices can accurately detect the quality of literature; cohesion based on the weights can help to fast discover the most core literature and citation during the developments of specific subjects; adequate re- suits from analyzing weight highland in cohesion make up for drawback of analyzing weights in revealing substructure, showing profound research dimensions in the development of the whole knowledge system. 6 figs. 4 tabs. 11 refs.
出处
《中国图书馆学报》
CSSCI
北大核心
2013年第2期111-120,共10页
Journal of Library Science in China
基金
国家社会科学基金重大项目"基于语义的馆藏资源深度聚合与可视化展示研究"(项目编号:1182D152)的研究成果之一
关键词
引文网络
核心文献
深度聚
合权值“高地”
Citation network. Core literature. Deep integration. Weight highland.
