An Effective Algorithm for Simultaneously Mining Frequent Patterns and Association Rules

Within the area of association rules mining, previous algorithms, e.g., FP-Growth and Apriori, have been generally accepted with high appraisals respectively. Most of these algorithms decompose the problem of mining association rules into two subproblems: find frequent pattern and generate the desired rules. Therefore, such a decomposition strategy cannot but bring delay problem when the size of database is considerable and makes user unbearable in a system where the required feedback time is rigor. To solve the problem, we catch a deep insight of FP-Growth algorithm and propose an effective algorithm by utilizing the FP-tree, called AR-Growth (Association Rule Growth), which can simultaneously discover frequent itemsets and association rules (AR) in a large database. It is analyzed in theory that our algorithm is correct and association rules generated by the algorithm are complete. The experiments show that the output of the AR sequence generated by AR-Growth is closely linear with the elapsed runtime, instead of the sudden eruption in FP-Growth.