A-Close+: An Algorithm for Mining Frequent Closed Itemsets

摘要

Association Rule Mining (ARM) is the most essential technique for data mining that mines hidden associations between data in large databases. The most important function of ARM is to find frequent itemsets. Frequent closed itemsets (FCI) is an important condense representation method for frequent itemsets, and because of its importance in recent years, there have been many algorithms implemented for it. One of the most fundamental algorithms for frequent closed itemset is A-close. In this paper, we optimize this algorithm using both optimized techniques "reducing pruning time" and "reducing database size", called "A-close+"..Results show that the performance cost of our algorithm is considerably less than A-close.