ne of themost popular datamining approaches is to fi nd frequent itemsets froma transaction

dataset and derive association rules. Finding frequent itemsets (itemsetswith frequency larger

than or equal to a user specifi ed minimum support) is not trivial because of its combinatorial

explosion. Once frequent itemsets are obtained, it is straightforward to generate association

rules with confi dence larger than or equal to a user specifi ed minimum confi dence