A Dynamic Approach for Discovering Maximal Frequent Itemsets

We present a novel method, which reads the database at regular intervals as in dynamic itemsets counting technique and creates a tree called dynamic itemset Tree containing items which may be frequent, potentially frequent and infrequent. This algorithm requires less time to discover all maximal frequent itemsets since it involves a method for reducing the size of the database. This method prunes the transactions and items of the transactions which are not of our interest after every scan of the database. Also, this method is independent of the order of the items.