Combination Tree for Mining Frequent Patterns Based on Inverted List

In this paper, a combination-tree algorithm is presented for mining frequent patterns based on inverted list. Compared with Apriori algorithm and FP-growth algorithm, our algorithm has better efficiency. Our algorithm insert items one by one with inverted list to build frequent tree, then transfer count between branches in order to make branches independent, our algorithm need only scan data set twice, can share more common items of transactions, can omit the local infrequent items, at the same time, avoid lots of recursive operations. Our performance study and theory analysis show that it is efficient in both dense datasets and sparse datasets