基于Mendeley与WoS的选择性计量指标与传统科学计量指标相关性研究  

Study on Correlation Between Altmetrics Indicators and Traditional Scientometric Indicators Based on Mendeley and WoS

作　　者：刘晓娟[1] 周建华[1] 尤斌[1] 

机构地区：[1]北京师范大学政府管理学院,北京100875

出　　处：《图书情报工作》2015年第3期112-118,共7页Library and Information Service

摘　　要：[目的 /意义]进一步探讨Web 2.0环境下选择性计量指标的有效性。[方法 /过程]以"data mining"为检索词,获得Mendeley与Web of Science两平台的交叉文献集合,分别对交叉文集的被引频数与阅读数、被引频数与标签数进行相关性检验后,从每组选取指标值差异最大与最小的各100篇文献进行具体分析。[结果 /结论]传统计量指标被引频数与Mendeley中的阅读数和标签数均存在弱相关性,证实了以阅读数和标签数为代表的选择性计量指标可以在一定程度上评估文献的影响力,且文献类型、出版年份和作者h指数会对用户阅读、引用等文献利用行为产生影响。未来文献影响力评价的发展方向应为传统文献计量方法与选择性计量方法的结合。[Purpose/significance ] This paper will explore the effectiveness of ahmetrics indicators in the Web 2.0 environment. [ Method/process] It achieves a set of overlapping documents in the data mining domain from Mendeley and Web of Science, verifies the correlation between citation number and reader counts as well as citation number and tags frequency, and selects 100 articles with great difference and articles with little difference to analyze specifically. [ Re- suit/conclusion ] It finds that the citation number in traditional indicator has weak correlation with reader counts and tags frequency in Mendeley. Thus confirms that reader counts and tags frequency as the representatives of altmetrics indicators can evaluate the literature influence in some extent, and document types, publication date and H index of authors can af- fect the usage rate and influence of literature. In the future, the literature influence evaluation should combine the traditional scientometric and newly ahmetrics

关 键 词：选择性计量指标 相关性分析 文献影响力 Mendeley 科学计量指标 

分 类 号：G353.1[文化科学—情报学]