Mining of frequent closed itemsets has been shown to be more efficient than mining frequent itemsets for generating non-redundant association rules. The task is challenging in data stream environment because of the unbounded nature and no-second-look characteristics.