网络中心性视角下的领域知识聚类研究

Domain Knowledge Clustering from the Perspective of Network Centrality

[目的/意义]探索领域知识的中心性与聚类情况有助于关键知识识别与知识群簇发现,对于掌握领域知识发展进程中的知识演化模式具有重要意义。[方法/过程]以网络科学的思想为基础,基于知识关联关系构建领域知识网络。采用中心性与聚类性的相关指标,对领域知识网络从时间序列上进行跟踪与分析。从网络整体与节点个体的层面,分别对领域知识生长发展过程中的中心性与聚类性展开交叉相关分析。[结果/结论]研究结果表明:领域知识网络在整体层面上中心性与聚类性具有非显著的负相关关系;高中心度的领域知识具有低聚类系数;领域知识的聚类性在不同的中心性视角下存在细节差异。[局限]研究中采用的数据难以涵盖所有类型的知识网络,在未来的研究工作中有待进一步拓展探索。

[Purpose/significance] The exploration of the centrality and clustering of domain knowledge contributes to the identification of key knowledge and the discovery of knowledge cluster,which is of great significance to master the pattern of knowledge evolution in the process of domain knowledge development. [Method/process] Based on the thought of network science,domain knowledge networks are constructed according to knowledge correlations. By using the indicators of centrality and clustering,the domain knowledge networks are tracked and analyzed from time series. From the whole network and the individual nodes,cross correlation analysis is carried out on the centrality and clustering in the process of domain knowledge development. [Result/conclusion] The results show that there is a non-significant negative correlation between centrality and clustering of overall domain knowledge network. The domain knowledge with high centrality shows low clustering coefficient,and the clustering of domain knowledge presents differences in details under different centrality perspectives. [Limitations] The data used in the research are difficult to cover all types of knowledge networks. The extensional research needs to be further explored in the future.