图书情报工作 ›› 2021, Vol. 65 ›› Issue (15): 111-119.DOI: 10.13266/j.issn.0252-3116.2021.15.013

• 情报研究 • 上一篇    下一篇

基于被引-逆文档权重的专家专长识别与分析——以图情领域为例

唐晓波1,2, 周禾深1, 李诗轩3, 牟昊4   

  1. 1 武汉大学信息管理学院, 武汉 430072;
    2 武汉大学信息系统研究中心, 武汉 430072;
    3 武汉理工大学安全科学与应急管理学院, 武汉 430070;
    4 国网四川省电力公司, 成都 610000
  • 收稿日期:2021-01-31 修回日期:2021-05-12 出版日期:2021-08-05 发布日期:2021-08-07
  • 通讯作者: 周禾深(ORCID:0000-0003-1133-2812),博士研究生,通讯作者,E-mail:zhouheshen@whu.edu.cn
  • 作者简介:唐晓波(ORCID:0000-0001-5885-45090),教授,博士生导师;李诗轩(ORCID:0000-0002-1879-4895),博士,讲师;牟昊(ORCID:0000-0002-1950-9953),高级工程师,博士研究生。
  • 基金资助:
    本文系国家自然科学基金项目"基于大数据的科教评价信息云平台构建和智能服务研究"(项目编号:19ZDA349)研究成果之一。

Identifying and Analyzing Expertise Tags of Scholars Based on the Cited-Inverse Document Frequency in the Library and Information Science Field

Tang Xiaobo1,2, Zhou Heshen1, Li Shixuan3, Mou Hao4   

  1. 1 School of Information Management, Wuhan University, Wuhan 430072;
    2 Center for Studies of Information System, Wuhan University, Wuhan 430072;
    3 School of Safety Science and Emergency Management, Wuhan University of Technology, Wuhan 430072;
    4 State Grid Sichuan Electric Power Company, Chengdu 610000
  • Received:2021-01-31 Revised:2021-05-12 Online:2021-08-05 Published:2021-08-07

摘要: [目的/意义] 识别专家专长有助于发现具有相同或相近研究方向的研究者,对开展细粒度的专家评价与分析具有重要意义。[方法/过程] 基于学术论文关键词构建专长种子词典,采用语义相似度计算对词典进行扩展与对齐;融合专长术语被引频次、作者贡献率与专长术语逆文档频率,提出专家专长术语的被引-逆文档权重计算方法;结合专长权重得分及排名,识别专家的代表性研究专长,并进行专家评价与分析。[结果/结论] 经实验验证,本研究提出的专家专长识别方法能够客观地反映专家专长的影响力,同时在细粒度专家评估、专家推荐以及学科热点分析等相关领域具有一定的实践参考价值。

关键词: 信息计量, 语义挖掘, 专长识别, 专家评价

Abstract: [Purpose/significance] Identifying expertise tags helps to find scholars with the same or similar research capabilities, which is of great significance to support fine-grained scholar evaluation and analysis. [Method/process] In this research, we collected the keywords of academic papers to build an expertise seed dictionary, and used semantic similarity to expand and align the dictionary. Additionally, we combined the citations frequency, author contribution rate and inverse document frequency of expertise terms, and proposed cited-inverse document frequency based weight calculation method for expertise tag. Considering the weights of expertise tags, we could find the representative expertise tags of scholars, and carry out expert evaluation and analysis. [Result/conclusion] Experiment proves that the proposed scholar expertise identification method can objectively reflect the influence of scholar expertise, and provide a practical reference for fine-grained scholar evaluation, expert recommendation, and field hotspot analysis and other related fields.

Key words: informetrics, semantic mining, expertise tag identification, expert evaluation

中图分类号: