科研管理 ›› 2016, Vol. 37 ›› Issue (3): 133-142.

• 论文 • 上一篇    下一篇

基于PageRank的评审专家信誉度度量方法

李旭彦1,2,宋英华1,杨晓秋2   

  1. 1 武汉理工大学中国应急管理研究中心, 湖北 武汉430070;
    2科学技术部基础研究管理中心,北京100862
  • 出版日期:2016-03-20 发布日期:2016-03-09
  • 通讯作者: 杨晓秋
  • 基金资助:

    国家社会科学基金重大项目: 基于情报流知识库的我国食品安全技术支撑体系优化策略研究(15ZDB168,2016.01-2019,10),国家自然科学基金: 分布决策环境下基于层次任务网络的应急资源规划方法研究(71501151,2016.01-2018,12), 教育部人文社会科学研究基金:基于情报体系的网络舆情应急管理信息平台研究(15YJCZH039,2016.01-2018,12),湖北省自然科学基金: 人员疏散的时空多尺度模拟技术研究(2015CFB593,2015.01-2016.12)。

Reviewer reputation measuring approach based on PageRank Algorithm

Li Xuyan 1,2, Song Yinghua 1, Yang Xiaoqiu 2   

  1. 1. China Research Center for Emergency Management, Wuhan University of Technology, Wuhan 430070, Hubei, China;
    2. Basic Research Service Management Center, Ministry of Science and Technology of the People’s Republic of China, Beijing 100862,China
  • Online:2016-03-20 Published:2016-03-09

摘要: 同行评议是当前对科研项目水平进行科学评价的主要方式之一,然而评议过程中专家评审能力的差别将会对科研项目评审结果产生影响。为此,本文提出了一种基于PageRank算法的评审专家信誉度度量方法,该方法首先利用高斯分布函数计算评审专家的评审能力,然后利用PageRank迭代算法对评审专家的信誉度进行求解,最后通过引入时间因子对评审专家的信誉度进行度量。基于同行评议真实数据集上的实验结果验证了本文提出方法的有效性,该方法将为科研项目评审及专家遴选提供有益参考。

关键词: 科研项目评审, 同行评议, 评审能力, PageRank, 专家信誉度

Abstract: Currently, peer review is one of the most effective approaches to evaluate the scientific project. However, the fairness of peer review results is seriously affected by various levels of reviewers in the process of peer review. Faced with the problem,this paper proposes a reviewer reputation measuring approach based on PageRank algorithm. Firstly, the proposed approach calculates the review level of reviewers based on Gaussian distribution, and then computes the current reputation of each reviewer based on PageRank algorithm. Finally, the proposed approach adopts the time factor to measure the reviewer reputation of all reviewers. Experimental results based on real reviewer data shows the effectiveness of the proposed approach. The approach will provide useful references for selecting peer reviewers.  

Key words: scientific project review, peer review, review level, PageRank, reviewer reputation