Optimization of Apriori Algorithm Based on Mining Association Rules

Aimed at improving the bottlenecks of large quantity of candidate itemsets generated by multiple scanning the database in Apriori algorithm, this paper proposes an optimized algorithm of Apriori. This algorithm optimizes the joint strategy with the feature of self joint when frequent itemsets are generated. The optimized joint strategy is used in boolean matrix, which is a representation of database. The experimental results indicate that the optimized algorithm has better performance than the Apriori, especially in the case of large-scale database.