融合K值算法与三指标的神经科学领域“睡美人”论文识别及影响因素探析

Identification of Sleeping Beauties in Neuroscience Through Combining K Value and Three Indicator Methods and Analysis on Their Influencing Factors

[目的/意义]从不同学科领域识别出"睡美人"论文并在科学界中广泛传播与使用,能够极大程度实现此类科技成果的科学价值,促进科学领域的发展与进步。[方法/过程]综合运用K值算法、三指标法和文献被引延迟指数,从神经科学领域1990-2010年发表的905 418篇论文中识别出"睡美人"论文,并对"睡美人"论文的期刊分布、论文篇幅、作者数量和睡眠特征等影响因素进行计量分析。[结果/结论]实证结果显示:(1)融合K值算法与三指标法能够从神经科学领域90余万篇论文中识别出26篇"睡美人"论文,识别准确率较高;(2)文献被引延迟指数方法识别出的"睡美人"论文数量较多,达到65篇,识别准确率略低,然而该方法的计算效率较高;(3)两类方法识别出的26篇共同"睡美人"论文的睡眠深度范围为0.11~1.63次,睡眠时长相对较短,平均时长为9.88年。此外,除总被引频次外,神经科学领域"睡美人"论文形成的影响因素与期刊影响因子、论文作者数量和篇幅等特征均不显著相关。

[Purpose/Significance] The identification of Sleeping Beauties from different disciplines and their wide dissemination and utilization in the scientific community can greatly come true the scientific value of such scientific achievements and promote the development and progress of scientific fields. [Methodology/process] By using k-value algorithm,three-indicator recognition method and citation delay index,the Sleeping Beauties were identified from 905418 articles published in Neuroscience from 1990 to 2010 were analyzed,as sell as influencing factors of these Sleeping Beauties as journal distribution,paper length,number of authors. [Result/Conclusion] The empirical results showed that:(1)26 Sleeping Beauties were identified via the combination of K-value algorithm and three-index method from more than 900000 articles in the field of Neuroscience,with high recognition accuracy.(2)A large number of Sleeping Beauties were identified via the citation delay index,up to 65,with slightly lower recognition accuracy,but the method had higher computational efficiency.(3)The sleeping depth of the 26 same Sleeping Beauties identified by two types of methods from Neuroscience ranged from0. 11 to 1. 63 times,and the sleeping duration was relatively short,with an average of 9. 88 years. In addition,the influencing factors of the formation of Sleeping Beauties in Neuroscience were not significantly correlated with the impact factor,the number of authors and length of papers,except the total citation frequency.