FIT: A Fast Algorithm for Discovering Frequent Itemsets in Large Databases