科学引文网络中形成多主题覆盖的枢纽节点群发现  

Discovery of a Set of Hub Nodes with Multi-topic Coverage in Scientific Citation Network

作　　者：李文静 刘向[1] 谭琳洁 严婷婷 Li Wenjing;Liu Xiang;Tan Linjie;Yan Tingting(School of Information Management,Central China Normal University,Wuhan 430079,China)

机构地区：[1]华中师范大学信息管理学院,湖北武汉430079

出　　处：《现代情报》2021年第1期144-151,共8页Journal of Modern Information

摘　　要：[目的/意义]多主题覆盖的枢纽节点群能够快速“导航”至领域更多主题的高被引综述文献和权威节点,从而可以使新领域工作者短时间全面了解领域已有研究、现状和未来发展趋势。[方法/过程]本文提出一种节点群发现算法——HubsRank算法,该算法基于引文网络中节点影响力(信息、知识)的传递,通过多轮迭代,得到引文网络中多主题覆盖的枢纽节点群。[结果/结论]最后,与HITs算法进行实证对比分析,本算法能更快、全面地提取枢纽节点群和不同主题的高被引权威节点,且该算法可以有效避免主题集聚效应。[Purpose/Significance]The hub nodes with the multi-topic coverage can quickly lead to the highly cited literatures and authoritative nodes of multiple topics in the field,so that the new field workers can fully understand the existing research,current situation and future development trend in a short time.[Method/Process]This paper proposed a node group discovery algorithm,HubsRank algorithm,which is based on the transfer of node influence(information and knowledge)in citation network.Through multiple iterations,the hub nodes with multi-topic coverage in citation network was obtained.[Result/Conclusion]Finally,compared with HITs algorithm for empirical analysis,this algorithm can extract a set of hub nodes and highly cited authority nodes of different topics more quickly and comprehensively,and this algorithm can effectively avoid topic clustering effect.

关 键 词：HubsRank算法 枢纽节点群 多主题覆盖 科学引文网络 

分 类 号：G250.252[文化科学—图书馆学]