Association rule mining algorithm based on matching array

The association rule mining algorithm Apriori need to repeatedly scan the transaction database and a lot of I/O loads, moreover it may generate huge candidate sets, the complexity of time and space is relatively high. Aiming at the limitation of the algorithm, an algorithm is proposed for association rule mining based on matching array. The algorithm only needs to scan the database once, screens out 1-frequent item sets and encode them. The latter process of the algorithm is candidate encoding operation and the effectiveness verification of frequent item sets, without scanning the database, and does not process the invalid frequent items for matching array. The experimental results under the same condition show that the improved algorithm can reduce the times of scanning database and the number of candidate by deleting candidate, thereby improving efficiency of the algorithm.