【摘要】 由于数据的不确定性,传统频繁模式挖掘方法难以适用到不确定性数据中.针对不确定性数据的特点,把挖掘确定性数据频繁模式的经典垂直挖掘算法Eclat算法扩展到不确定性数据中,提出了UP-Eclat算法.该算法分别对Tid集和项集搜索树进行扩展:把原来只有一个id域的Tid扩展成两个域,即id域和概率域;用扩展后的Tid集代替原来的Tid集,生成扩展后的项集搜索树.扩展后的Tid集可以表示不确定性数据,然后利用扩展后的项集搜索树进行频繁模式挖掘.通过实验与分析,UP-Eclat算法可行,高效.更多还原

【Abstract】 Because of the uncertainty,the traditional way of mining frequent patterns is not available in uncertain data.As a result,this paper extends the classic vertical mining algorithm Eclat for mining frequent patterns from uncertain data and then proposes UP-Eclat algorithm.This algorithm extends the tidset as well as the itemset search tree.The Tid that contains only one id field is extended to a new Tid that contains both id field and probability field.Then the extended itemset search tree is consisted of the new tidset.The extended tidset can describe uncertain data,and the extended itemset search tree is built to mine the frequent patterns.The UP-Eclat algorithm is proved to be efficient according to the experimentation.更多还原