中文摘要：

针对传统空间关联规则挖掘对数据硬化分导致的＂尖锐边界＂问题,提出了一种顾及模糊属性的空间关联规则挖掘方法。该方法引入模糊集理论,将模糊空间属性通过隶属函数转化为隶属度表示的模糊数值,从而将其划分为模糊集合。然后使用改进的模糊关联规则挖掘算法扫描数据库,根据相应的支持度得到频繁项集,最终提取出关联规则。实验结果表明,该方法能够对带有模糊属性的空间数据进行关联规则挖掘,且在一定程度上提高了挖掘结果的兴趣度。

英文摘要：

Aiming at the sharp boundary problem arising from hard division in traditional association rules mining process, a novel association rules mining method is proposed. The new method integrates fuzzy set concepts and adopts membership function to convert fuzzy spatial attributes into fuzzy value denoted by membership degree, and then the improved fuzzy association rule mining algorithm is designed to scan the database and generate frequent items according to the given support degree. Finally the association rules are extracted from the frequent items. The experimental results show this method is excellent in mining association rule from spatial data including fuzzy attributes, and has better interestingness measure than the existing algorithms.