关联规则在空间数据挖掘中的研究  

Research On Association Rules of Spatial Data Mining

作　　者：曾玲[1] 熊才权[1] 胡恬[1] 

机构地区：[1]湖北工业大学信息工程学院,武汉430068

出　　处：《计算机与数字工程》2005年第6期71-73,共3页Computer & Digital Engineering

摘　　要：在智能化、集成化的空间数据应用领域中,空间数据挖掘是一门很重要的技术,而关联规则分析是空间数据挖掘的主要方法之一。文章基于数据挖掘中的关联规则分析方法,提出不同于一般数据挖掘的算法,设定兴趣度量,并通过将项的概念泛化为空间谓词,事务的概念泛化为邻域,关联规则的概念泛化为同位规则,发现多种形式的有效规则,并用逻辑语言或类SQL语言方式描述规则,以使空间数据挖掘趋于规范化和工程化。最后进行了实评。Spatial data mining plays a very important role in intellective and integrated spatial data application fields. Using association rules to analyze is one of major methods of spatial data mining technology. Based on association rules, this paper presents the algorithm which is different from the general data mining ,it sets the interest degree and discusses how to generalize spatial predicate, adjacent domain from items ,transaction and generalize isotropic rules from association rules to discover different kinds of rules. According to the above methods, the paper describes the rules by using logic language or SQLsimilar language to make spatial data mining tend to standardization. Finally, an experiment is conducted on a real spatial data mining, which shows that the method is practical and effective.

关 键 词：关联规则 空间数据库 数据挖掘 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]