Analysis of Residence Time in Shopping Using RFID Data -- An Application of the Kernel Density Estimation to RFID

This study shows a method of determining and visualizing the existence probability of customers from shopping-path data in supermarkets using a database collected by a RFID (Radio Frequency Identification) technique, which allows us to analyze the detailed behaviors of customers. First, we present a method to estimate customer existence probability density on the sales floor using a Kernel density estimation. The kernel density estimation obtains continuous distribution of the existence probability density and grasps the detailed movements of customers. This estimation is better than an aggregation of residence time in each sales floor zone. Secondly, we visualize the customer existence probability density as cartographic output. Thirdly, we assess the relations between customer existence probability and sales in each sales floor zone using both shopping-path data and customer purchasing records. Finally, we assess the relation between customer existence probability and sales on each store shelf to verify the utility of this method for sales and marketing.