赵星.学术文献用量级数据Usage的测度特性研究[J].中国图书馆学报,2017,43(3):44~57
Exploring the Features of Usage Data for Academic Literatures
学术文献用量级数据Usage的测度特性研究
Received:March 20, 2017  
DOI:
Key words:Bibliometrics  Usage data  Academic evaluation  
中文关键词:  文献计量学  使用数据  学术评价  引文分析  信息行为
基金项目:本文系国家自然科学基金青年项目“h型信息网络测度的机理与实证研究”(编号:71503083)的成果之一
Author NameAffiliation
ZHAO Xing 华东师范大学经济与管理学部信息管理系、华东师范大学学术评价与促进研究中心 上海200241 
Hits: 2491
Download times: 879
Abstract:
For a long time,citation has always been considered the most important basic data for measuring the impact of academic literature. It leads to the rise of Bibliometrics and Informetrics but also is controversial in academia. One of the critic opinions is that citation data is only one side of the impact,biased and difficult to fully measure complex academic activities. Therefore,an innovational supplementary perspective is needed.〖JP.

    In September 2015,Web of Science(WoS)announced Usage,a new set of data to reflect the attention that a single article received. This data is the first standard given by WoS,which is not based on citation analysis. Unlike citation,Usage reflects abroader interest of academic users and summarizes two types of user data:1)The number of downloads;2)The times of export for bibliographic. It can be seen that Usage mainly reflects the users specific usage of articles. Compared with citation,the academic behavior it characterized is more basic. Thus,it can be viewed as an underlying data for academic literature.This study first discusses the potential value and theoretical characteristics of Usage data in WoS. Then,based on the 166 767 articles in Physics,Computer Science,Economics and Library and Information Science,we study the basic statistical characteristics and the informetric distribution of Usage,and focus on the comparison between Usage and citation. Finally,the representative articles of library and information science are discussed.〖JP.

    The results show that,compared with citation,Usage is more discriminative and sensitive. In terms of distribution,Usage has an approximate positively skew distribution at high frequency part,and an approximate power-law distribution is found. The evaluation result of Usage is independent to a certain degree,and there is no essential opposition to the result of citation. The study indicates that Usage can be used as the promising data for academic evaluation.

    The limitations of this study include that the data is limited to four disciplines and the future studies still need to be more detailed. In addition,because it is the newly emerged data,the current time window of Usage is limited. Thus,it is difficult to expand the span of data. This article demonstrates the basic features and feasibility of Usage in the academic evaluation and is expected to provide a reference for the application and development of this new data. 4 figs. 6 tabs. 23 refs.

中文摘要:
      新型数据的涌现一直是推动信息计量学前行的重要动力,Web of Science(WoS)给出的学术文献用量级数据Usage可望在影响力维度、学科差异、行为动机三个理论方面为引文分析提供补充视角。本研究以2013年物理学、计算机、经济学和图书情报学被SCI和SSCI收录的166 767篇论文为研究样本,探索性地研究了该Usage数据的测度特征。结果显示,相较于引文数据,Usage更具测评区分度与敏感性;其在高频局部呈现出近似正偏态分布,于累积整体涌现出近似幂律分布;Usage的测评结果具有独立性,且与引文结果没有本质对立。虽然WoS平台的Usage仍有笼统性、可伪性和封闭性之局限,但仍不失为配合引文数据、提供丰富影响力测评视域的可能选择之一。图4。表6。参考文献23。
View Full Text   View/Add Comment  Download reader