图书情报工作 ›› 2018, Vol. 62 ›› Issue (16): 102-110.DOI: 10.13266/j.issn.0252-3116.2018.16.012

• 情报研究 • 上一篇    下一篇

基于“公平性测试”的Altmetrics学术质量评价方法研究

刘红煦1, 王铮2,3   

  1. 1. 同济大学图书馆 上海 200092;
    2. 西北大学公共管理学院 西安 710127;
    3. 南京大学信息管理学院 南京 210023
  • 收稿日期:2018-02-01 修回日期:2018-05-24 出版日期:2018-08-20 发布日期:2018-08-20
  • 作者简介:刘红煦(ORCID:0000-0002-4809-1437),助理馆员,硕士,E-mail:hxliu@lib.tongji.edu.cn;王铮(ORCID:0000-0001-5727-5935),讲师,博士后。
  • 基金资助:
    本文系同济大学图书馆2017年馆内项目"替代计量学与学术评价体系变革"研究成果之一。

Study on Altmetrics Academic Quality Evaluation Method Based on Fairness Test

Liu Hongxu1, Wang Zheng2,3   

  1. 1. Tongji University Library, Shanghai 200092;
    2. School of Public Administration, Northwest University, Xi'an 710127;
    3. School of Information Management, Nanjing University, Nanjing 210023
  • Received:2018-02-01 Revised:2018-05-24 Online:2018-08-20 Published:2018-08-20

摘要: [目的/意义]为了探讨在学科论文评价上,Altmetrics指标间的量度结果是否具有一致性,试图通过已有测量数据构建基于Altmetrics的文献综合评价模型,为评价数据的使用和重用创建新的度量标准。[方法/过程]获取PLoS ALMs上土木工程学科2009年至2016年发表的全部文章的Altmetrics指标,采用相关分析、主成分分析方法将15个原始指标划分到4个主成分,得到适用于具体学科的Altmetrics评价指标体系,并对Altmetrics的单一指标与综合评价指标进行相关度分析,采用"公平性测试"方法探索时间对论文评价相关度的差异。[结果/结论]研究发现采用"公平性测试"方法消除时间影响后,引用指标相关性上升,而与社交媒体传播相关的指标有所下降,但Twitter指标不降反增,且始终与F值具有较高的相关性。

关键词: 替代计量学, 主成分分析, 公平性测试, 学科评价

Abstract: [Purpose/significance] In order to explore whether the measurement results of Altmetrics indicators are consistent in the evaluation of subject papers, this paper attempts to build a comprehensive literature evaluation model based on Altmetrics through the existing measurement data, and to create new metrics for the evaluation and reuse of data.[Method/process] In this study, the Altmetrics indicators of all articles published in the Civil Engineering disciplines from 2009 to 2016 on the PLoS ALMs are obtained. The 15 original indicators are divided into 4 principal components using correlation analysis and principal component analysis methods, which are applied to specific disciplines. It analyzes the correlation between the single index and the comprehensive evaluation index of Altmetrics, and uses the fairness test method to explore the differences in the relevance of time to the evaluation of the paper.[Result/conclusion] The study finds that after using the fairness test method to eliminate time impact, the relevance of the citation indicators has increased, while the indicators related to social media communication have declined. But the Twitter index has not decreased, and it has always been highly correlated with the F value.

Key words: Altmetrics, PCA, fairness test, discipline evaluation

中图分类号: