地方政府绩效评估指标设计的研究进展与数据挖掘理论的应用

Advances in Local Government Performance Evaluation Indicator Designing Researches and Using of Data Mining Theory

在中国这样的大国,各个地区之间在自然、地理、历史、社会等很多方面千差万别,由于这种异质性的存在,地方政府绩效评估指标的设计是一个噪音很多的系统工程,它会受到各方面因素的影响。实际上,解决绩效指标设计的最有效手段,就在于解决指标设计中所收集、采用的各种绩效信息。目前国内外的各种指标设计方法尽管取得了很大进展,解决了一系列理论与实践问题,但仍然不能有效解决从大量噪音指标中筛选有效指标的难题,其实这可以利用"未来十大技术"之三的数据挖掘技术作为工具,充分发掘这些信息中所隐含的地方政府绩效现实,并把它们融入到绩效指标的设计中去,这是解决地方政府绩效指标设计问题的治本之策。

In a tremendously large country like China,there are substantial differences on natural geographic features,history background,and social realities among the varied regions.Because of the heterogeneity,the local government performance evaluation indicator designing is a systematical engineering which could be affected by various factors.Actually,the most effective means to design the government performance evaluation indicators is the collecting and adopting of different kinds of performance information.During the process of daily operation,local government gathers diverse objective information in the internal and external activities of administration,including the National Statistical Yearbook,Local Statistical Yearbook,Sector Statistics Yearbook,government departments' work plans & programs,work records,material inputs & consumption,recordings of solving practical problems,actual results of the work achieved and social effects,survey data of public opinion,and even the climate and customs of the regions the governments located in.We can use DM theory to mine the indictors in the different kind of information.