To address disadvantages of the Apriori algorithm, a new method is presented to optimize the Apriori. It can reduce the number of invalid candidate item sets through optimizing the strategy of frequent item sets pruning. In order to reduce the connections of items, avoid repeated comparison of the same items, it optimizes the joining strategy. And it also removes the useless transactions from database step by step in order to reduce the times of scanning transactions. The results of experiment show that the optimized algorithm is more efficient.