Topological analysis of citation networks to discover the future core articles
Topological analysis of citation networks to discover the future core articles
复制标题
DOI：
10.1002/asi.20529
复制
发表时间：
2007-04-01
影响因子：
--
通讯作者：
Matsushima, Katsumori
中科院分区：
文献类型：
--
作者：
Shibata, Naoki;Kajikawa, Yuya;Matsushima, Katsumori
在这篇文章中，我们使用引文网络的拓扑分析来研究决定未来学术文章被引用能力的因素。其基本思想是，将被多次引用的文章在过去的拓扑结构中处于“相似”位置。为了验证这一假设，我们研究了未来被引用时间与三种中心性指标之间的相关性：聚类中心性、亲密中心性和中间中心性。我们还分析了被引次数的老化和自相关的影响。案例研究进行了以下两个最近的代表性创新：氮化镓和复杂网络。结果表明，被引次数是解释近未来被引次数的主要因素，中间度中心性与远未来被引次数相关。拓扑位置对被引能力的影响受研究中心从已有领域向新兴领域转移的迁移现象的影响。
In this article, we investigated the factors determining the capability of academic articles to be cited in the future using a topological analysis of citation networks. The basic idea is that articles that will have many citations were in a "similar" position topologically in the past. To validate this hypothesis, we investigated the correlation between future times cited and three measures of centrality: clustering centrality, closeness centrality, and betweenness centrality. We also analyzed the effect of aging as well as of self-correlation of times cited. Case studies were performed in the two following recent representative innovations: Gallium Nitride and Complex Networks. The results suggest that times cited is the main factor in explaining the near future times cited, and betweenness centrality is correlated with the distant future times cited. The effect of topological position on the capability to be cited is influenced by the migrating phenomenon in which the activated center of research shifts from an existing domain to a new emerging domain.