User repurchase behavior prediction for integrated energy supply stations based on the user profiling method

Under the guidance of the Dual Carbon goal, integrated energy supply stations have gradually become an essential facility for the energy transition. Promoting user repurchase has become a vital marketing strategy for integrated energy supply station enterprises. This paper proposes a prediction method based on the user profiling method to predict user repurchase behavior accurately. First, using an improved RFM model and the K-means algorithm, this paper constructs user profiles by dividing 10,000 users into three clusters: general-value developmental users, high-value new users, and low-value loyal users. Next, this paper uses the random forest, light gradient boosting machine, and extreme gradient boosting to predict the repurchase behavior of non-clustered users and the three clusters and compares their prediction performance. In addition, this paper adopts the stacking method for model fusion to improve the prediction performance further. The results show that the accuracies of the best prediction models for the three clusters are 93.28 %, 93.68 %, and 92.84 %, respectively. Finally, this paper provides each cluster with the corresponding prediction model of user repurchase behavior and marketing strategy. For the application scenario of integrated energy supply stations, this study accurately predicts the repurchase behavior of each cluster with unique consumption characteristics. It helps to provide personalized services for new energy vehicle consumers, optimize their consumption experience, and facilitate sustainable consumption.

Automated summary

This paper proposes a prediction method based on the user profiling method to accurately predict user repurchase behavior. The results show that the prediction models have high accuracy, making them valuable for integrated energy supply stations and the energy transition.