Extracting Promising Sequential Patterns from RFID Data Using the LCM Sequence

Recently, supermarkets have been using RFID tags attached to shopping carts to track customers' in-store movements and to collect data on their paths. Path data obtained from customers' movements recorded in a spatial configuration contain valuable information for marketing. Customers purchase behavior and their in-store movements can be analyzed not only by using path data but also by combining it with POS data. However, the volume of path data is very large, since the position of a cart is updated every second. Therefore, an efficient algorithm must be used to handle these data. In this paper, we apply LCMseq to shopping path data to extract promising sequential patterns with the purpose of comparing prime customers' in-store movements with those of general customers. LCMseq is an efficient algorithm for enumerating all frequent sequence patterns. Finally, we construct a decision tree model using the extracted patterns to determine prime customers' in-store movements.