Research And Application Of The Prediction Of O2O Coupon' Usage Based On Data Mining

In recent years,with the development of e-commerce in China,O2O(Online To Offline)business in consumer market has attracted more and more attention.Using coupons to activate old users or attract new customers into the store is an important marketing method of O2 O.At present,the random delivery strategies commonly used by merchants tend to waste coupons,while personalized delivery based on user characteristics can improve the cancellation rate of coupons.The users generate huge amount of consumption behavior data in their using the Internet platform.Through these data,users' purchase intention and consumption habits can be analyzed to realize accurate delivery of coupons,thus improving the marketing ability of merchants.In this paper,the prediction of user consumption behavior based on big data analysis is selected as the research subject.Through data mining of real data provided by the TianChi big data platform,the usage of users after receiving coupons is predicted,and the prediction effects based on different algorithms are compared and analyzed.The conclusion is verified by an example.The main research contents of this paper are as follows:(1)Research on O2 O coupon usage prediction related technologies and methods.Research data analysis and processing,feature selection and principles of GBDT,XGboost and Lightgbm algorithms.Discuss the evaluation indexes of prediction model--AUC value and F1 value.(2)Based on the analysis and processing of consumer behavior data provided by the TianChi big data platform and the selection of features,the combination of features is carried out by analyzing the basic feature groups mined,and some results of feature selection are presented.Experimental results of GBDT algorithm,XGboost algorithm and Lightgbm algorithm are compared and analyzed.(3)Based on the accurate release demand of coupons of a hotel in Wuhan,this paper forecasts and analyzes the relevant data of the hotel's consumers,and the results verify the rationality and effectiveness of the above conclusions,and accurate release of coupons is carried out according to the predicted results,and finally the hotel's marketing ability is significantly improved in practical application.