以共引网络识别研究领域的引文评价方法有效性分析 Analysis on the Effectiveness of Citation Evaluation Methods in the Research Field Identification by Co-Citation Network 作者： 宋丽萍 1 王建芳 2 付婕 1 苑珊珊 1 作者单位： 1. 天津师范大学管理学院 天津 300387 2. 中国科学院科技战略咨询研究院 北京 100190 提交时间： 2023-04-01 摘要: [目的/意义]为探讨论文层面学术规范化引用影响力指标，对基于共引网络识别研究领域的引文评价方法Relative Citation Ratio （RCR）的有效性进行分析。[方法/过程]选取iCite、F1000、InCites数据库，以739 604篇文献为样本，采用R语言对RCR与基于先验分类的Category Normalized Citation Impact （CNCI）、期刊影响因子（JIF）等进行相关分析，以检验RCR在学术影响力评价中的有效性。[结果/结论]RCR与CNCI高度相关、与JIF弱相关，因此表明RCR与CNCI具有相似性与差异性，并揭示了其在践行《吃瓜qq群内黄》、摆脱"以刊评文"方面的可操作性；交叉科学中RCR、JIF相关度与学科交叉数目的反变关系以及RCR与JIF相关度低于单一学科的结果，揭示了RCR在交叉科学论文学术影响力评价中具有一定的适用性与有效性。因此，RCR借助共引网络识别研究领域的算法为论文影响力评价，特别是为交叉科学研究成果的评价提供了新的解题思路。 Abstract: [Purpose/significance] To investigate field-normalized reference influence indicators on the level of papers, the effectiveness of a citation evaluation method (Relative Citation Ratio, RCR) based on the research field identification by co-citation network is analyzed.[Method/process] In order to test the effectiveness of RCR in the evaluation of academic impact, the paper chose iCite, F1000, InCites, then took 739 604 articles as a sample, and analyzed the correlation among RCR, Category Normalized Citation Impact (CNCI) and journal impact factor (JIF) by the R language.[Result/conclusion] The paper shows that RCR strongly correlates with CNCI, and poorly correlates with journal impact factor, which indicate the similarities and differences between RCR and CNCI and confirm that it is a good practice for implementing San Francisco Declaration on Research Assessment and assessing research on its own merits rather than on the basis of the journal in which the research is published. It also turns out the inverse relationship of the correlation and the number of interdisciplinary subjects between RCR and journal impact factor in interdisciplinary science, and lower correlation between RCR and journal impact factor than in a single discipline, which appears the applicability and effectiveness of RCR in evaluating the influence of academic papers in interdisciplinary science. The paper concludes that RCR based on research field identification by co-citation network provides new solution for the evaluation of paper influence, especially for interdisciplinary research outputs. CNCI RCR F1000 期刊影响因子 期刊： 图书情报工作 分类： 图书馆学、情报学 >> 情报学 引用： ChinaXiv:202304.00409 (或此版本 ChinaXiv:202304.00409V1 ) doi:10.12074/202304.00409 CSTR:32003.36.ChinaXiv.202304.00409.V1 推荐引用方式： 宋丽萍,王建芳,付婕,苑珊珊.(2023).以共引网络识别研究领域的引文评价方法有效性分析.图书情报工作.[ChinaXiv:202304.00409] 版本历史 [V1] 2023-04-01 16:02:59 ChinaXiv:202304.00409V1 下载全文