空间关联规则的双向挖掘

Mining Spatial Association Rules in Two-direction

空间数据库中关联规则挖掘不仅需要考虑关系元组属性之间的关系———纵向关系,更需要挖掘元组之间的关系———横向关系,如相邻、相交、重叠等。本文通过分析空间数据库的存储模式,借鉴事务数据库关联规则的挖掘方法,对空间关联规则进行完整定义,并对规则的兴趣度度量进行探讨。根据挖掘的方向将空间数据挖掘归纳为纵向挖掘、横向挖掘、双向挖掘。在双向挖掘中,提出一种新算法,该算法根据挖掘任务进行约束,缩小挖掘空间,然后通过空间计算将空间关系转化为非空间关系,经过多次循环,获取非空间项集,进而挖掘出空间关联规则。据此提出空间数据双向挖掘工作流程,并通过实例进行了验证。

Spatial data mining is different from data mining in transaction DB. In most case, the relationships between the tuples in transaction DB do not be taken into account. But in spatial database, there are relationships not only between the attributes, but also between the tuples, and most of the associations exists between the tuples-objects, such as adjacent, intersection, overlap and other topological relationships. So the tasks of spatial data association rules mining include not only mining the relationships between attributes of spatial objects, which we call vertical direction DM, but also mining the relationships between the tuples, which we call horizontal direction DM. This paper analyses the storage models of spatial data, uses for reference the technologies of data mining in transaction DB, defines spatial association rules, including vertical direction association rule, horizontal direction association rule and two-direction association rule, discusses the ts of interestingness of spatial association rules, and propose the work flows of spatial association rules data mining. During two-direction spatial association rules mining, we propose an algorithm to get non-spatial itemsets. By spatial analysis, we can transfer the spatial relations into non-spatial associations and get non-spatial itemsets. Based on the non-spatial itemsets, we can make use of Apriori algorithm or other algorithms to get the frequent itemsets and then, spatial association rules come into being. To confirm that, we mine in the land using spatial DB to get spatial association rules to validate the algorithm. The test results show that the algorithm is efficient and can mine the interesting spatial rules.