资源数据质量诊断的理论与方法初探——以统计型数据为例  

THEORY AND METHODOLOGY OF RESOURCES DATA QUALITY DIAGNOSIS:THE CASE OF STATISTIC DATA

作　　者：杨小唤[1] 王乃斌[1] 刘红辉[1] 

机构地区：[1]中国科学院地理科学与资源研究所,北京100101

出　　处：《资源科学》2002年第1期11-14,共4页Resources Science

基　　金：中国科学院知识创新工程重要方向性项目"国家资源环境数据库建设与数据共享研究"(KZCX2 30 8 4 )资助

摘　　要：随着信息化技术的发展 ,各学科越来越重视各种数据的应用 ,数据的质量问题就显得更加重要 ,数据质量的好坏直接关系到数据开发的效率和数据应用的程度。如何诊断现有数据的质量 ,是数据与信息处理领域亟待解决的问题。本文首次以资源科学研究中最常用的统计型数据为例提出了数据质量诊断的方法 ,其关键是根据各种数据序列的不同特性 ,研究其内在规律 ,确定诊断模型。然后根据诊断模型寻找不符合常规的数据点 ,对该点数据进行核实后做技术处理。该项研究为提高数据的可信度和增强数据应用程度提出了一套较为简便易行的理论方法。With the development of information technology, growing attention has been devoted to data application. Thus the quality of data, directly involving in the efficiency of data exploring and the extent of data application, becomes more important than before. Therefore, how to diagnose the quality of existing data is a critical problem to be resolved in the field of data information processing. By employing the statistic data commonly used in the study of resources science, a diagnosis method for data quality, for the first time, is put forward in this paper. The key issues of such method are to study the inherent laws of the data series according to their characteristics and then to establish a diagnosis model. Through this diagnosis model, record not consistent with the general can be dug out and then technical adjustment can be done. With this respect, the diagnosis model provides a simple and convenient tool to improve the reliability of data as well as to enhance the extent of data application.

关 键 词：数据质量 诊断模型 资源 统计型数据 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]