Since there are usually a large number of distinct single items in a typical
transaction database, and their combinations may form a very huge number
of itemsets, it is challenging to develop scalable methods for mining frequent
itemsets in a large transaction database.