In-store shopping activity modeling based on dynamic bayesian networks

RFID technology has been recently adopted in retail environments to track consumer in-store movements, bringing about new exciting opportunities for spatial data mining-enabled marketing. In this paper, we propose a Dynamic Bayesian Networks (DBN)-based model of customer in-store shopping trips and activities. This model infers a customer's product purchase interest given the observations of customer in-store movement data collected through a RFID-based wireless network. We also report a preliminary evaluation of our approach using a real-world dataset. Our proposed approach can be potentially used to help create an intelligent shopping environment, in which store operators can target their marketing efforts at providing effective location-aware real-time product recommendation for individual customers.