A new method based on LTB algorithm to mine frequent itemsets

In most data mining algorithms, the core operation is to mine frequent itemsets. Because of data-intensive operation and large output, most operation time is spent in scanning the database. In this paper, we propose a novel algorithm-LTB algorithm to mine frequent itemsets. Loose bounds are used to remove the candidate itemsets whose support cannot satisfy the preset threshold. Tight bounds determine the frequency of some candidate itemsets without scanning the database. For the remainder itemsets after above two steps, we only can scan the database with traditional a priori algorithm. Experiments show that the amount of the candidate frequent itemsets and the operation time can be decreased dramatically.