图书情报工作 ›› 2018, Vol. 62 ›› Issue (17): 84-95.DOI: 10.13266/j.issn.0252-3116.2018.17.011

• 情报研究 • 上一篇    下一篇

基于熵权TOPSIS及因子分析的学术期刊综合评价研究

奉国和1, 周榕鑫2, 武佳佳1   

  1. 1. 华南师范大学经济与管理学院信息管理系 广州 510006;
    2. 华南理工大学工商管理学院 广州 510641
  • 收稿日期:2018-01-02 修回日期:2018-05-02 出版日期:2018-09-05 发布日期:2018-09-05
  • 作者简介:奉国和(ORCID:0000-0002-0774-1544),教授,博士,E-mail:ghfeng@163.com;周榕鑫(ORCID:0000-0003-1812-2586),硕士研究生;武佳佳(0000-0002-7342-8388),硕士研究生。
  • 基金资助:
    本文系国家社会科学基金项目"基于文本挖掘的科技文献知识发现研究"(项目编号:16BTQ071)和2016年华南师范大学研究生创新项目"基于深度学习的科技文献挖掘研究"(项目编号:2016wkxm62)研究成果之一。

Research on the Comprehensive Evaluation of Academic Journals Based on Entropy Weight TOPSIS and Factor Analysis

Feng Guohe1, Zhou Rongxin2, Wu Jiajia1   

  1. 1. The Department of Information Management, School of Economics & Management, South China Normal University, Guangzhou 510006;
    2. School of Business Administration, South China University of Technology, Guangzhou 510641
  • Received:2018-01-02 Revised:2018-05-02 Online:2018-09-05 Published:2018-09-05

摘要: [目的/意义] 对图书馆学情报学期刊的综合质量进行评价,发现样本期刊质量在评价年间的变化,为期刊评价研究提供新的思路。[方法/过程] 以2011-2017年版《中国科技期刊引证报告》为数据源,选取34种图书馆学情报学类期刊为样本期刊,经过筛选得到7个指标,使用熵权法结合因子分析计算各指标权重,通过TOPSIS法计算各年期刊综合得分并构建综合评价矩阵。最终对综合评价矩阵进行聚类分析得到2010-2016年图书馆学情报学类期刊分档情况。[结果/结论] 结果表明:①进行期刊评价时应对评价指标进行严格筛选;②各指标所属的公因子趋于稳定;③为指标设置不同权重以体现其意义;④载文质量高于载文数量;⑤图书馆学情报学期刊存在"马太效应";⑥图书馆学情报学期刊需要进一步发展。

关键词: 期刊评价, 熵权法, TOPSIS法, 面板数据, 因子分析

Abstract: [Purpose/significance] This paper evaluates the comprehensive quality of Library Science and Information Science's journals to find the quality changes in sample journals and purpose new thoughts for the journal evaluation research. [Method/process] This paper used the data from Chinese S&T Journal Citation Reports(2011-2017), choosed 7 terms after a series of filtering work, then set different weight through entropy method and factor analysis, evaluated the comprehensive scores and ranks for 34 kinds of Library Science and Information Science' journals during 2010-2016 by using TOPSIS model and built the comprehensive evaluation matrix. Finally, this paper used the clustering method to set the grades for the journals. [Result/conclusion] The results are as follows:① We should filter the attributes strictly when evaluate the qualification of Academic Journal; ②The common factors of each index tend to be stable; ③ The weight set for different indicators show these indicators' significance; ④ For the published articles, the quality is more important than the quantity; ⑤ "Matthew effect" exists in the evaluation of library and information science journals;⑥ Library and information science journals need further development.

Key words: journal evaluation, entropy weight method, TOPSIS model, panel data, factor analysis

中图分类号: