Mining frequent patterns using CP: a comparative study