Abstract: It is discussed how to pretreat enormous data in online data mining.It is presented a data pretreating algorithm based on cube.The algorithm can collect data effectively and reduce data enormously.The performance study shows that the pretreatment is resultful for data aggregation.