摘要
针对因子分析和主成分分析在期刊评价中存在的问题,提出了一种新的学术期刊评价方法——因子理想解法,其原理是,首先采用因子分析筛选出关键因子,然后确定各关键因子的权重,在此基础上将关键因子标准化后采用加权TOPSIS进行评价。研究结果表明,因子理想解法能消除指标间的相关性导致的重复计算问题,具有较好的区分度,权重设置比较合理,距离函数采用直线距离也更加科学,可以在期刊评价中广泛进行运用。
To overcome the shortcoming of factor linalysis and principle component analysis in academic journals, this paper puts forward a new method, factor ideal solution (FIS). The' fwst step of HS is to conduct key factors selection based on factor analysis, and then' to evaluate academic journals by weighted TOPSIS after key factors standardization. The results show that FIS may solve Repeat calculation caused by indictors' correlation and has good distinguish degree. The weight giving and distance function are also reasonable.
出处
《情报杂志》
CSSCI
北大核心
2012年第10期22-26,共5页
Journal of Intelligence
基金
国家社会科学基金后期资助课题"科技评价中多属性评价方法基本理论研究"(编号:10FTQ003)