An Improved Apriori Algorithm Based on Goal Orientation and Double Pruning

Aiming at solving the problem of high time complexity of association rule mining among item-sets in Apriori algorithm and obtaining the set of goal-association relations of single-to-single point type, this paper improves the Apriori algorithm based on Goal Orientation and Double Pruning (GODP-Apriori). Oriented by the type of goal, the GODP-Apriori uses double pruning to remove the redundant item set during the calculation of the frequency of item set support, which greatly improves the efficiency of obtaining the frequent item set; in addition to this, by sacrificing space for time, it also records the set of transaction items corresponding to the item when calculating the support degree, which avoids the repetitive scanning in the subsequent calculations and then effectively increases the computation efficiency of traditional Apriori algorithm. The effectiveness of GODP-Apriori algorithm in improving the calculation time efficiency of traditional Apriori algorithm was verified by using a real household electricity consumption data at the end of this research.