情报科学 ›› 2022, Vol. 40 ›› Issue (3): 174-182.

• 博士论坛 • 上一篇    下一篇

图情档一级学科点的优势研究方向识别研究

  

  • 出版日期:2022-03-01 发布日期:2022-03-08

  • Online:2022-03-01 Published:2022-03-08

摘要: 【目的/意义】在双一流学科建设背景下,本文基于学科研究主题分类模型、z指数评价模型,以国内图书情
报与档案管理领域 25 所一级学科点为例,建立了高校学科点的优势研究方向识别方法。【方法/过程】利用
Word2vec 模型结合 k-means 进行关键词聚类,将图情档学科划分为 12 个研究主题;其次,基于匹配的主题-学科
点-被引频次三维关系,建立25所学科点在12个研究主题下的300个评价数据集;最后,计算学科点在不同研究主
题中的zsi指数,并识别学科点的优势研究方向。【结果/结论】实验发现:z指数评价模型较好地实现了各机构发表论
文数量、质量、被引分布之间的平衡。其次,基于细粒度研究主题的学科点评价方法具有分类评价、多元化评价的
相对优势。【创新/局限】本文提出的评价模型可为不同科研群体提供一定的信息参考与决策支持,同时也为国内
高校学科评价提供了一个相对公平、个性化、多元化的评价视角;但该评价模型易受主题分类质量的影响。

Abstract: 【Purpose/significance】Under the background of the construction of double first-class disciplines,based on the subject clas‐
sification model and z-index evaluation model,this paper takes 25 first-level disciplinary institution in the field of library information
and archives management in China as examples to establish a method for identifying the superiority research direction of institution.
【Method/process】Keyword clustering was carried out by Word2vec model combined with K-means,and the subject of library informa‐tion and archives management was divided into 12 research topics. Secondly, based on the matching three-dimensional relationship
subject-institution-cited frequency,the 300 evaluation data sets of 25 disciplinary institution under 12 research topics were estab‐
lished.Finally,the zsi-index was used to identify the superiority research direction of institution【. Result/conclusion】The experimental
results show that the z-index achieves a good balance between the quantity,quality and citation distribution of papers published by var‐ious institutions.Secondly,the subject evaluation method based on fine-grained research topic has the advantages of classified evalua‐
tion and diversified evaluation.【Innovation/limitation】The evaluation model proposed in this paper can provide certain information
reference and decision support for different research groups,and at the same time.it also provides a relatively fair,personalized and di‐
versified evaluation perspective for the disciplinary evaluation of domestic universities.However,the evaluation model is susceptible to
the quality of subject classification.