论文部分内容阅读
作者采用8项单项指标,运用三种评价方法(熵权法、主成分分析法和简单线性加权法),对我国大陆35种图书情报专业期刊进行综合评价。结果表明,主成分分析法要优于简单线性加权法和熵权评价法。应用SPSS软件对35种图书情报期刊采用Ward Method进行分层聚类(Hierarchical Cluster),聚类结果表明:基本与主成分分析法的排序相吻合;排序总体上比较符合实际情况。
The author uses eight individual indicators, using three kinds of evaluation methods (entropy method, principal component analysis and simple linear weight method) to comprehensively evaluate 35 kinds of library and periodicals in mainland China. The results show that principal component analysis is superior to simple linear weighting and entropy weighting. SPSS software was used to classify 35 kinds of library and information science journals by Ward Method. The clustering results showed that the basic and principal component analysis method were in good agreement with each other, and the ranking was generally in line with the actual situation.