一种面向关系型数据的可视质量分析方法  

Visual Quality Analysis Method for Relational Data

作　　者：滕东兴[1] 曾志荣[1,2] 杨海燕[1] 王宏安[1] 戴国忠[1] 

机构地区：[1]中国科学院软件研究所人机交互技术与智能信息处理实验室,北京100190 [2]中国科学院研究生院,北京100049

出　　处：《软件学报》2013年第4期810-824,共15页Journal of Software

基　　金：国家自然科学基金(61173057;61100162);国家重点基础研究发展计划(973)(2011CB302205;2013CB329305);国家高技术研究发展计划(863)(2012AA02A608;2012AA02A613);创新基金重大项目专项计划(ISCAS-2010-01)

摘　　要：由于信息系统所提供数据的质量不高(如数据残缺、数据不一致、数据重复等)导致管理者决策过程中经常面临"数据丰富,信息匮乏"的困惑是目前企业普遍存在的现象.为了切实提高信息系统所提供数据的可用性,研究了影响关系数据库数据质量的主要因素,提出了面向多数据源的统一元数据模型和数据库数据质量评估模型,构建了用于数据质量评估的交互式可视形态集.建立了一个面向关系数据库的数据质量可视分析系统,并结合具体企业应用实例进行验证.结果表明,该系统能够有效分析数据质量,提高企业分析决策的可靠性和准确性.Because of the low quality of data provided by information systems such as data missing, data conflicts and data duplicate, it is widespread in enterprise that decision-makers are often faced with ＂rich data but poor information＂. To improve the data availability of information systems, the main factors affecting data quality of relational database are studied in this paper, also a unified metadata model based on multi data sources and a data quality assessment model are proposed, and a set of interactive visual analogues for data quality assessment is built. Finally, a visual analysis system for data quality in relational database is developed, which is verified with several enterprise practical cases. It is indicated that the built system can analyze data quality effectively, and then improve the reliability and accuracy of enterprise decision-making.

关 键 词：信息可视化 可视分析 人机交互 数据质量 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]