Classification of bibliometric indicators is a fundamental issue in information science.Traditionally,the classification is based on subjective classification.This article presents an empirical study on the mathematic...Classification of bibliometric indicators is a fundamental issue in information science.Traditionally,the classification is based on subjective classification.This article presents an empirical study on the mathematics journals listed in JCR 2019 by using objective classification methods including cluster analysis,factor analysis,and principal component analysis to classify bibliometric indicators.Different classification results are compared and further interpreted,major finding are:the classification results of objective classification methods share similarities;objective classification helps better comprehend bibliometric indicators;objective classification should be used in combination with subjective classification;cluster analysis performs better in classifying bibliometric indicators than factor analysis and principal component analysis;not all the results of objective classification are meaningful;cluster of indicators has sufficient influence on subsequent evaluation and regression analysis.This study provides a new paradigm for journal classification and indicator analysis.展开更多
文摘Classification of bibliometric indicators is a fundamental issue in information science.Traditionally,the classification is based on subjective classification.This article presents an empirical study on the mathematics journals listed in JCR 2019 by using objective classification methods including cluster analysis,factor analysis,and principal component analysis to classify bibliometric indicators.Different classification results are compared and further interpreted,major finding are:the classification results of objective classification methods share similarities;objective classification helps better comprehend bibliometric indicators;objective classification should be used in combination with subjective classification;cluster analysis performs better in classifying bibliometric indicators than factor analysis and principal component analysis;not all the results of objective classification are meaningful;cluster of indicators has sufficient influence on subsequent evaluation and regression analysis.This study provides a new paradigm for journal classification and indicator analysis.
