Mining frequent itemset from uncertain data

We study the problem of mining frequent itemset from probabilistic data. Firstly, to solve the semantic corruption brought by expected frequent itemset conception, we define the probabilistic frequent itemset which is consistent with possible world model and holds the apriori property. Secondly, we develop a dynamic programming like polynomial algorithm for testing candidate frequent itemsets. Finally, a P-Apriori algorithm for mining top-A probabilistic frequent itemsets is presented, which can incrementally report probabilistic frequent itemsets one-by-one in descending order of their confidences. Comprehensive experiments have been conducted on both real and synthetic datasets to verify the effectiveness and efficiency of the algorithm. The results show that P-Apriori algorithm performs stably on various parameter configurations.