基于关联图的频繁闭模式挖掘

将关联图的数据挖掘思想应用到频繁闭模式的挖掘中,使用位向量的技术简化项集支持度的计算,构造关联图表示项集间的频繁关系。在此基础上,提出一种频繁闭模式挖掘算法,针对频繁闭模式的特点,结合剪枝策略、子集检测策略、搜索策略等技术手段,优化算法性能。实验结果表明,该算法在时间性能上优于经典的频繁闭模式算法CLOSET。

Frequent Close Pattern Mining Based on Association Graph

The data mining thoughts based on association graph was applied to frequent close pattern mining.The calculation of item set support was simplified by using bit vector technique.The frequent relationship among items was expressed by constructing an association graph.Further more,a mining algorithm of frequent close pattern was proposed.According to the characteristics of the frequent closed pattern,the algorithm performance was improved by using pruning method,subset test strategy and search strategy of,etc.The experimental results showed that the algorithm was superior to CLOSET which was the classic algorithm for frequent close pattern mining.