How to recognize the tendency of the coupons among the users who receive the coupons and then send the coupon reminder to improve the coupon redemption rate and reduce the marketing cost has become an important issue in the coupon decision-making process. Based on the log data and transaction data in enterprise database, this study combined with the demographics, past purchasing behavior, past coupon usage behavior and the visiting behavior during the coupon validity period to construct the e-coupon redemption behavior prediction model. The model is constructed to help e-commerce enterprises identify the target users who have the coupon proneness after the coupons are issued, so as to send coupon reminders in time and enhance the effectiveness of coupon marketing.