基于加权迭代聚类方法的云平台小微企业信用评价  

Cloud-Platform Credit Evaluation of Small and Micro Enterprises Based on Weighted Iterative Clustering Method

作　　者：孙瑜 牛艳芳[1] 毕秀玲[1] SUN Yu;NIU Yanfang;BI Xiuling(School of Accounting,Shandong University of Finance and Economics,Jinan 250014,China)

机构地区：[1]山东财经大学会计学院,山东济南250014

出　　处：《山东财经大学学报》2021年第6期86-97,共12页Journal of Shandong University of Finance and Economics

基　　金：国家社会科学基金项目“新形势下大数据审计能力的结构、测度和提升路径”(20BJY033)。

摘　　要：以云平台为依托,基于企业行为理论,设计由财务维度和行为维度构成的"双维度"小微企业信用评价指标体系,以443家批发和零售业小微企业为研究对象,构建熵权法与迭代聚类融合的加权迭代聚类方法,对小微企业进行分类并判定所属信用等级。评价结果发现:443家小微企业财务等级总体较高,但行为等级普遍较低,致使综合信用等级偏低,说明小微企业可能存在忽视记账规范性或通过异常记账行为使之呈现良好财务结果的行为,证实将行为维度融入评价体系能够对财务维度起到有效的辅助作用。同时,研究发现加权迭代聚类方法相比层次分析法更具严谨性、科学性和客观性。Relying on cloud platform and based on enterprise behavior theory,this study designs a"two-dimensional"small and micro enterprise credit evaluation index system composed of financial dimension and behavior dimension,and with 443 wholesale and retail small and micro enterprises as the research object,constructs a weighted iterative clustering method which integrates entropy weight method and iterative clustering so as to classify small and micro enterprises and determine their credit rating.The evaluation results show that the overall financial grade of the 443 small and micro enterprises is high while their behavior grade is generally low,resulting in a low comprehensive credit grade,which indicates that small and micro enterprises may ignore the normalization of bookkeeping or present good financial results through abnormal bookkeeping behavior.This study confirms that integrating behavior dimension into the evaluation system can play an effective auxiliary role in financial dimension.Meanwhile,it is found that the weighted iterative clustering method is more rigorous,more scientific and more objective than the analytic hierarchy process.

关 键 词：云平台 小微企业 信用评价体系 加权迭代聚类方法 

分 类 号：F276.3[经济管理—企业管理]