基于主成分和集对分析法的期刊评价方法研究

发布时间：2018-04-05 02:34

本文选题：期刊评价　切入点：综合评价指标　出处：《情报杂志》2017年03期

【摘要】：[目的/意义]随着期刊排名关注度的增加,期刊评价方法的研究具有十分重要的意义。为了研究期刊评价方法,分析期刊评价指标间的分离程度及其结构关系,最终达到提高期刊评价质量的目的。[方法/过程]选取影响力指数、基金论文比、总被引频次、引用期刊数等16个指标对期刊进行评价。通过主成分分析筛选指标,构建评价指标体系,再运用集对分析法分析指标间的分离程度。[结果/结论]实证结果显示较高等级期刊的指标同一度较高,差异度较低;较低等级期刊的指标同一度较低,差异度较高;而在对立度方面则没有明显的区分。评价结果和官方给出的期刊排名相差无几,说明集对分析法可对期刊评价指标间的分离度进行分析,在期刊评价方面具有可靠性和有效性。
[Abstract]:With the increasing attention of journal ranking, the research of journal evaluation method is of great significance.In order to study the methods of periodical evaluation and analyze the degree of separation and structural relationship between the evaluation indexes of periodicals, the purpose of improving the quality of periodical evaluation is finally achieved.[methods] Sixteen indexes, such as influence index, fund paper ratio, total citation frequency and number of journals cited, were selected to evaluate the journals.The evaluation index system was constructed by principal component analysis (PCA), and the degree of separation between indexes was analyzed by set pair analysis (SPA).[results / conclusion] the empirical results showed that the index of higher grade journals had higher degree of identity and lower degree of difference; the index of lower grade journals had lower degree of identity and higher degree of difference, but there was no obvious difference in opposites.The evaluation results are no different from the official ranking of periodicals, which shows that the set pair analysis method can analyze the degree of separation between the evaluation indexes of periodicals, and it is reliable and effective in the evaluation of periodicals.
【作者单位】：南昌大学经济管理学院;
【基金】：国家自然科学基金项目“基于增强邻域搜索策略的联合型生产调度问题算法研究”(编号:61473141)资助国家社会科学基金项目“非参数估计的影响因素优化匹配方法与应用研究”(编号:14BTJ016)研究成果之一
【分类号】：G353.1;G237.5

【相似文献】