MRFI-The maintenance of representative frequent itemsets

Mining frequent itemsets is an important research task for knowledge discovery, which is to discover the groups of items appearing always together excess of a user specified threshold from a transaction database. However, there may be many frequent itemsets existing in a transaction database, such that it is difficult to make a decision for a decision maker. Recently, mining frequent closed itemsets becomes a major research issue. The reason is that all frequent itemsets can be derived from frequent closed itemsets. In addition, the transactions in a database will increase constantly. It is a challenge that how to update the previous frequent closed itemsets from the increased transactions. In this paper, we propose an efficient algorithm MRFI for incrementally mining frequent closed itemsets without scanning original database. MRFI algorithm generates frequent closed itemsets by performing some operations on the previous closed itemsets and the added transactions without doing any searching operation. Finally, the experimental results show that MRFI algorithm performs much better than the previous approaches.