文章导读
总览 评价 牛琪锴 1, , 王春凤 2, , 狄增如 1, , 樊瑛 1,* ( 1、 北京师范大学系统科学学院; 2、 北京师范大学理科学报编辑部; ) 摘要: 评价论文影响力是科学计量学领域的重要内容。科学引文网反应了论文之间的相互引用关系。本文基于H指数的思
牛琪锴1,, 王春凤2,, 狄增如1,, 樊瑛1,*
(
1、北京师范大学系统科学学院; 2、北京师范大学理科学报编辑部; )
摘要:
评价论文影响力是科学计量学领域的重要内容。科学引文网反应了论文之间的相互引用关系。本文基于H指数的思想对于评价论文影响力提出"H算法",进而提出改进后的"H+算法"用以解决H算法简并度高的问题。把算法应用在美国物理学会(APS)提供的科学引文网中,实验结果表明,两种算法计算出的评价指标呈现高端幂律分布。这两种算法比PageRank算法的时间复杂度都要低;在评价优秀论文方面的表现上,"H算法"准确率优于PageRank算法。
关键词:
H指数;论文影响力;科学引文网
NiuQikai1,, WangChunfeng2,, DiZengru1,, Fanying1,*
(
1、School of Systems Science,Beijing Normal University; 2、Science journal,Beijing Normal University; )
Abstract:
The article influence evaluation is quite important in scientific metrology.The science citation network reflect the reference relationship of papers issued by scientists. In this paper, we propose the "H algorithm" to evaluate the article influence based on the H-index and propose a improved "H+ algorithm" to solve the high degeneracy of "H algorithm". We use the algorithm in the science citation network offered by APS and the result indicate that the evaluation index of two algorithms show top power-law distribution. The time complexity of the two algorithms is lower than the PageRank algorithm; and the accuracy rate to evaluate the excellent papers of "H algorithm" is higher than PageRank algorithm.
Tag:
点此返回栏目查看更多>>>参考论文