摘要： 空间关联规则与传统关联规则的主要区别在于空间关联规则挖掘需要考虑空间实体的距离关系、方位关系和拓扑关系，而空间概念层次的形成往往是基于多个空间数据层或面向特定主题的。该文在研究空间关联规则算法的基础上，提出一种基于概念树的多层次空间关联规则挖掘算法，设计和实现一种基于J2EE的空间关联规则原型挖掘系统，并以某市土地利用为例说明了系统的实施过程。该系统挖掘出来的8大类土地利用类型的空间关联规则具有一定的现实意义。

关键词: J2EE技术, 空间关联规则, 空间概念层次关系, 原型系统实现

Abstract: Most differences between spatial association rules and traditional association rules are the fact that distance relation, direction relation and topology relation of the spatial entity are needed to be taken into account. The spatial concept layer relations are based on cell pattern or theme. In the paper, the spatial association rules algorithm is discussed, a multi-level spatial association rules algorithm based on concept tree is proposed, a spatial data mining prototype system based on J2EE is designed and implemented, and an example based on the land use of one city is used to further explain the implementing process. The spatial association rules of the eight types of land use, which are acquired by the system, prove the realistic value.

Key words: J2EE, spatial association rules, relation of spatial concept layer, prototype system implementation

中图分类号: 

• TP391.9