Categorized electronic coupon use prediction method and device and electronic equipment

Abstract

The invention relates to a categorized electronic coupon use prediction method and device and electronic equipment. The method comprises the steps that a model training sample is determined according to historical information of getting categorized electronic coupons of a user; the features of the user and the features of the categorized electronic coupons in the model training sample are acquired to act as sample information of the model training sample; model training is performed on the model training sample based on the preset classification model according to the sample information; and the use probability of the given categorized electronic coupons of the given user is predicted based on the trained preset classification model. The use probability of the specific user for the specific categorized electronic coupons can be predicted and different categories of coupon issuing marketing can be performed for different users so that precision marketing can be performed, the repurchase effect can be enhanced and the repurchase marketing cost can be reduced.

Description

technical field [0001] The present application relates to the technical field of data modeling, in particular, to a method, device and electronic equipment for predicting the use of electronic coupons by category. Background technique [0002] In e-commerce websites, electronic coupons are used more frequently. With the vigorous development of e-commerce, the use of electronic coupons is also increasing. However, in the current precision marketing, the use of electronic coupons is relatively low. How to increase the usage rate of e-coupons more accurately is very important. It can not only bring a certain amount of new users to the website, but also effectively increase the GMV (Gross Merchandise Volume, total merchandise transaction) of the website, and wake up sleeping users . [0003] The existing technical solution is based on the classification model of GBDT (Gradient Boosting Decision Tree, based on the classification and regression algorithm implemented by the decis...

AI Technical Summary

Problems solved by technology

[0004] The above technical solution can only predict all categories of electronic coupons, but the current pre-AUC (Area Under roc Curve, a standard for measuring the quality of a classification model) based on the GBDT single model for all categories of electronic coupons is relatively low. Low, and the use scenarios of all-category e-coupons are limited, and more e-coupons belong to limited categories of e-coupons, resulting in a lower utilization rate of all-category e-coupons, which cannot meet the increasingly refined category operation and marketing

Images