Towards a new approach for mining frequent itemsets on data stream