基于知识网络的标签层级演化研究  

Research on Hierarchical Structure Evolution of Tags Based on Knowledge Network

作　　者：刘雅姝[1] 滕广青[1] 王东艳[1] 毕强[2] 白淑春[3] 

机构地区：[1]东北师范大学信息科学与技术学院,长春130117 [2]吉林大学管理学院,长春130022 [3]吉林大学图书馆,长春130012

出　　处：《图书情报工作》2017年第11期13-20,共8页Library and Information Service

基　　金：国家自然科学基金项目"基于网络结构演化的Folksonomy模式中社群知识组织与知识涌现研究"(项目编号:71473035)研究成果之一

摘　　要：[目的/意义]基于知识网络的核心-边缘结构揭示Folksonomy知识组织模式中标签层级演化的模式与规律,对洞悉开放网络环境下知识组织的层级结构具有重要意义。[方法/过程]以网络思维为指导,基于标签之间的同现关系构建领域知识网络。依据核心-边缘结构分析方法,遵循时间序列对标签层级的演化情况,从描述统计、显著性分析、层级升降演化等方面进行跟踪研究。[结果/结论]标签的核心/边缘层级结构随着领域知识的发展演化变迁,逐渐收敛趋于稳定。标签层级的晋升与淡化主要受知识关联驱动,核心层标签之间的关联关系占有更高的权重。标签层级演化规律的揭示,有助于把握Folksonomy模式中知识组织的结构特征,对于揭示开放网络环境下知识关联及其结构关系具有积极的促进作用。[ Purpose/significance ] Core-periphery structure of knowledge networks can reveal the laws and models of tags hierarchy evolution. The detection has important significance for insight into hierarchical structure of knowledge or- ganization in the open network environment. [ Method/process] Under the guidance of network thinking, the domain knowledge networks are constructed based on the co-occurrence relationships between tags. According to the core-periphery structure analysis method, the evolution of the tags hierarehyon time series is tracked from the description of statistics, sig- nificant analysis, and rising/falling of tags hierarchy. [ Result/conclusion ] The results show that the tags hierarchy of core/peripherystrueture changes with the development of domain knowledge, and gradually approaches stabilize. The ris- ing and falling of tags hierarchy are mainly driven by knowledge cross-correlation and the relationships between core layer tags occupy a higher weight. These rules of hierarchy evolution help to grasp the structural features of the knowledge orga- nizationin the folksonomy model, and help to promote positive role for revealing the knowledge cross-correlation and struc- tural relations in the open network environment.

关 键 词：知识网络 层级结构核心一边缘Folksonomy 

分 类 号：G254[文化科学—图书馆学]