政府绩效评估指标设计研究中“失去的研究”的补充研究——指标的可加性探索  

A Study to "the Lost Study" in Government Performance Evaluation Index Desiging——A Probe into Evaluation Index Aggregate Degree and its Additional Attribute

作　　者：尚虎平[1] 陈星宇[2] 

机构地区：[1]厦门大学公共事务学院,福建厦门361005 [2]四川大学公共管理学院,四川成都610064

出　　处：《华东经济管理》2009年第12期111-115,共5页East China Economic Management

基　　金：国家自然科学基金(70873100);江苏省高校自然科学基金研究项目(09KJB630002);教育部2009年哲学社会科学课题青年项目"数据挖掘视域下我国县(市)政府绩效评估指标筛选研究"

摘　　要：针对政府绩效评估指标可加性,如"辖区内三个代表思想掌握状况"与"辖区内土豆3000斤/亩"如何相加的问题,目前学界鲜有研究。对于这种高属性的政府绩效评估指标,在设计中必须按照高属性维对象的聚类特性,首先将高属性维降为低属性维。本研究在标准化方式之外,采用数据挖掘的关联规则与集合理论相结合,借助于二态变量的方式,如在"社会和谐"指标下有"辖区内18岁以上有意愿人口就业率(1/0)"、"辖区内每百平方米乞丐数量(1/0)",用指标聚中度和指标聚中向量实现了海量数据中同级指标的可加性。Because we unanimously ignored the study of the additional attribute of government performance evaluation index as the example of how ＂the reality of grasping the theory of ＇ 3-presentative Thoughts＇ in the area under the government jurisdiction＂ can be added to the index of ＂potato production being 1500kg/mu＂. Facing the high dimensional attribute government indexes, we have to follow the na- ture to cluster the high dimensional subjects to lower the dimension of the subjects. In this study, the binary state variables are used to lower the dimensions and realize the additional attribute. On the basis of the binary state variables, we introduce the new methods in set theory and Data Mining theory to solve the problem of indexes＇ addtitional attribute.

关 键 词：政府绩效评估指标 集合 二态变量 聚中度 聚中向量 

分 类 号：F120[经济管理—世界经济]