摘要
[目的 /意义]学术文献影响力评价指标不断推陈出新,但尚缺乏在研究主题层面对文献影响力的评价,为发现不同研究主题内具有高影响力和引用价值的文献,本研究给出一种基于研究主题的文献影响力评价方法。[方法 /过程]以Web of Science数据库中2011年-2015年间情报学领域500篇高被引文献为样本,利用LDA模型对样本文献进行主题建模,将主题对文献的支持度与文献被引频次相结合,计算特定主题文献的被引频次(specific topic cited frequency,简称STCF),并根据每篇文献在相应主题内的STCF值对文献进行影响力排序。[结果/结论]结果表明,STCF值能反映文献的主题内容、细粒度体现文献的学术地位、呈现文献研究主题的多元性,能够有效弥补被引频次、Altmetrics指标的不足。
[Purpose/significance] The indexes of evaluating literature influence continue to innovate, but a lot of indexes are lack of assessment of the literature influence on the topic level. Aimed to solve this problem, this paper presents a method based on the research topic.[Method/process] This paper chooses 500 highly cited literature as samples which attribute to the field of information science in the Web of Science database. LDA model is used to model the literature subject. The support of the topic of the literature is combined with the cited frequency to compute the specific topic cited frequency（STCF）. We rank the literature according to the STCF value of each document in the corresponding subject.[Result/conclusion] The empirical research shows that STCF value reflects the subject content, the subject diversity, and the academic status of literature. This method effectively compensates for the shortage of the cited frequency and Altmetrics index to evaluate the literature impact.
作者
李秀霞
宋凯
Li Xiuxia;Song Kai(School of Communication,Qutu Normal University,Rizhao 27682)
出处
《图书情报工作》
CSSCI
北大核心
2018年第20期88-94,共7页
Library and Information Service
基金
国家社会科学基金项目"文献内容分析与引文分析融合的知识挖掘与发现研究"(项目编号:16BTQ074)研究成果之一
