摘要
利用CNKI引文数据库,以图情领域共19本期刊53243篇文献为统计数据源,从单篇论文、作者、期刊三种粒度,分别对文献下载频次与被引频次进行数据正态性检验、相关性分析及曲线估计,并探讨利用下载频次预测被引频次的可行性。实验表明,下载频次与被引频次的相关性在不同粒度下差异较大:单篇论文粒度下相关性不强,作者粒度下呈显著的二次函数正相关,而期刊粒度下呈显著的三次函数正相关。因此,从作者或期刊粒度,利用下载频次预测被引频次是可行的。
In the granularity of the journal papers, the authors and the journals, normality test, correlation analysis andcurve estimation were conducted to test the relevance between the download frequency and the citation frequency on the da-taset which were collected from the CNKI citation database covering 53243 journal papers and 19 journals in total in thefield of Library and Information Sciences, and explored the feasibility of predicting the citationa frequency from the down-load frequency. The experiments show granularity relevance differences between the download frequency and the citationfrequency. To be more specifically, little relevance between the download frequency and the citation frequency was found inthe granularity of the journal papers whereas quadratic function, cubic function were related in the granularity of the authorsand the journals, respectively. As a consequence, it is possible to predict the citation frequency from the download frequen-cy in the granularity of the authors or the journals.
出处
《情报科学》
CSSCI
北大核心
2016年第1期3-8,共6页
Information Science
基金
国家科技支撑计划课题(2012BAH33F03)
国家自然科学基金面上项目(71173164)
关键词
下载频次
被引频次
正态性检验
相关性
download frequency
citation frequency
normality test
correlation analysis
作者简介
陆伟（1974-），男，辽宁鞍山人，教授，博士研究生导师，主要从事信息检索、知识管理等研究．
