通用关联规则挖掘框架的设计与实现

Design and Implementation of General Association Rule Mining Framework

对适用于多种数据类型的关联规则挖掘框架进行了研究。从概率论出发讨论了支持度计算问题,提出利用有限测度计算项集支持度的方法,分析了Apriori性质的本质,提出通用关联规则挖掘算法的设计思路。在此基础上,设计并实现了通用关联规则挖掘框架,使用该框架进行了事务、空间和时空数据的挖掘实验,验证了其可行性、通用性及正确性。

According to probability theory,we used a limited measure to calculate the support of an item set.Based on this conclusion,a general association rule mining framework was designed and implemented.The implementation of this framework was tested with transactional,spatial and spatio-temporal data.The coherent results inferred the feasibility,universality and validity of this framework.