Abstract: Clementine is one of the effective data mining tool. It can quickly build model applied to the business activities of the auxiliary decision-making. Clementine has been widely used at present. Through theory and examples data analyses, it was pointed out that the Clementine software had a lot of defects in the predict result of detecting outliers, feature selection, sampling node and the effect is not ideal