作　　者： ; ; (刘佩珊）;

机构地区： 暨南大学信息科学技术学院计算机科学系

出　　处： 《计算机集成制造系统》 2009年第7期1436-1441,1449,共7页

摘　　要： 为从数据挖掘结果中获得真正有用的规则,探讨了评估规则的度量,提出了度量关系的新概念,并应用于发现与评估有用规则集的算法中。经过度量函数之间的相关分析,确定适合的度量,并据此对规则进行兴趣度评估,指导用户对有用规则的选择。此外,为了解决数据集成系统中存在"垃圾"数据的问题,提出了数据质量挖掘与控制系统框架,将数据挖掘规则应用于数据质量管理中。使用经过评估筛选得到的规则,能够检测数据库中的异常和不完整数据,解决数据集成过程中部分数据质量问题。通过企业信息集成实例,说明了规则评估及应用方法的有效性。 To get really useful rules from data mining results, measurements for evaluation rules were studied, and the new concept of measurement relationship was defined, which was applied in the algorithm for discovering and e valuating useful rules. By correlative analysis of measurement functions, suitable measures for rule interestingness evaluation were determined, so as to guide users for useful rules selections. In addition, to solve the problem of ＂garbage＂ data in integration systems, the framework of data quality mining and control system was presented. Data mining rules were applied in data quality management. Using the evaluated and selected rules, the abnormal, incomplete data were detected, so as to solve some data quality problems in data integration. Example of enterprise information integration illustrated the effectiveness of the rules＇ evaluation and the application method.

关 键 词： 数据挖掘 规则评估 度量函数 数据质量 集成系统

领　　域： [自动化与计算机技术] [自动化与计算机技术]