WeatherPon: A Weather and Machine Learning-Based Coupon Recommendation Mechanism in Digital Marketing

Machine learning has become an innovative technology tool to improve coupon recommendation systems in digital marketing. Users’ information, coupon content, and customers’ shopping history are the main factors to predict revenues, resolve promotions, and adjust distribution in past coupon recommendation models. However, weather is a critical factor that can affect human activities in daily life but has barely been awarded in previous studies. Many studies observed that weather change could impact human mood, behavior, and economics. Therefore, we proposed a weather and machine learning-based coupon recommendation mechanism named "WeatherPon." We employ real-time weather data to optimize coupon recommendation systems. Considering the non-linear relationship and multiple uncertainties, we utilize the deep neural networks and use Shapley Value to interpret the black-box model. The WeatherPon is a coupon recommendation mechanism not only improves accuracy but also helps enterprises achieve automated marketing solutions in real-time.