关键词: 不确定数据的最大频繁项集, 不确定数据最大频繁项挖掘（UMMFI）算法, 有序的压缩不确定树（SCUF-tree）, 逐层逐个地处理节点（NBN）策略

Abstract: In order to resolve the problem of data and route redundant in UF-tree, this paper designs the Sequential Compressed Uncertain Frequent Pattern Tree（SCUF-tree）, and stores different support values of element in the node, so as to compress the storage space and conveniently transplant the algorithm which can mine maximal frequent itemsets in certain data. Then it combines the design idea of the MMFI algorithm and proposes UMMFI algorithm which can mine maximal frequent itemsets in the uncertain database. The UMMFI algorithm adopts a NBN（Node By Node） strategy to mine maximal frequent itemsets. Experiment results manifest that the UMMFI algorithm is effective and adaptable in the uncertain database.

Key words: maximal frequent itemsets in uncertain databases, Mining Maximal Frequent Items from Uncertain data（UMMFI） algorithm, Sequential Compressed Uncertain Frequent pattern tree（SCUF-tree）, Node By Node（NBN） strategy