图书情报工作 ›› 2019, Vol. 63 ›› Issue (2): 91-98.DOI: 10.13266/j.issn.0252-3116.2019.02.011

• 情报研究 • 上一篇    下一篇

基于创新扩散理论的学术论文影响力广度研究

梁国强1,2, 侯海燕1, 高桐3, 孔祥杰2,3, 胡志刚1   

  1. 1. 大连理工大学科学学与科学技术管理研究所暨WISE实验室 大连 116023;
    2. 印第安纳大学伯明顿分校信息、计算与工程学院 伯明顿 47408;
    3. 大连理工大学软件学院 大连 116024
  • 收稿日期:2018-07-01 修回日期:2018-08-12 出版日期:2019-01-20 发布日期:2019-01-20
  • 通讯作者: 胡志刚(ORCID:0000-0003-1835-4264),副教授,硕士生导师,通讯作者,E-mail:huzhigang@dlut.edu.cn
  • 作者简介:梁国强(ORCID:0000-0002-9669-4048),博士研究生;侯海燕(ORCID:0000-0002-2790-9973),教授,博士,博士生导师;高桐(ORCID:0000-0002-5464-9433),本科生;孔祥杰(ORCID:0000-0003-2698-3319),副教授,硕士生导师。
  • 基金资助:
    本文系国家社会科学基金项目"高科技前沿监测中的知识图谱方法与应用研究"(项目编号:14BTQ030)研究成果之一。

Impact Breadth of Scientific Papers Based on Innovation Diffusion Theory

Liang Guoqiang1,2, Hou Haiyan1, Gao Tong3, Kong Xiangjie2,3, Hu Zhigang1   

  1. 1. WISE Lab of Dalian University of Technology, Dalian 116023;
    2. School of Informatic and computing, Indiana University Bloomington, Bloomington 47408;
    3. School of Software, Dalian University of Technology, Dalian 116024
  • Received:2018-07-01 Revised:2018-08-12 Online:2019-01-20 Published:2019-01-20

摘要: [目的/意义]采用被引次数衡量学术论文影响力存在诸多弊端。本文认为学术论文影响力包括其传播的深度、速度和广度3个方面,熵可用于衡量学术论文影响力传播的广度。[方法/过程]选择1901-2017年生物医学领域的诺贝尔奖论文作为最具影响力的论文纳入实验组,并根据1:1配对原则设立对照组,比较两组论文发表后5年内其施引文献所属学科数量、熵以及熵与被引次数的相关性。[结果/结论]实验组中65%以上的论文施引文献学科数量高于对照组;实验组熵的均值介于0.552-0.772,对照组介于0.251-0.481,有显著性差异(P<0.05);两组论文的被引次数与熵的相关性较弱,均小于0.3。结果表明:①70%以上的高影响力论文在发表早期能够影响较多的学科;②采用熵对论文影响力广度进行识别具有可行性。

关键词: 创新扩散, 熵, 诺奖论文, 生物医学

Abstract: [Purpose/significance] Using citations as a measurement of the impact of an article have long been criticized. This study argues that the impact of scientific paper includes its depth, speed, and breadth in diffusion. We can measure the impact breadth of an article by using entropy.[Method/process] This study regards Nobel Prizes winning articles in Physiology and Medicine field as the most influential scientific research, collecting Nobel Prizes winning articles into the test group and matching them at a ratio of 1:1 into the control group. The citing articles' disciplinary diversity, within five years after these papers' publishing, was explored. In addition, this study employs entropy to measure the diversity.[Result/conclusion] 65 percent of articles in the test group have relatively higher disciplinary diversity compared to the control group. The values of entropy in the test group are between 0.552 and 0.772, and between 0.251 and 0.481 in the control group (p<0.05). There is a weak correlation, which below 0.3, between citation counts and values of entropy. The paper argues that above 70 percent of highly influential articles have a high disciplinary diversity in their early stage, and it is possible by using entropy as an indicator to measure the breadth of an article's impact.

Key words: innovation diffusion, entropy, Nobel Prizes winnings articles, Physiology and Medicine

中图分类号: