A Sequence Pattern Matching Approach to Shopping Path Clustering

We can have a new perspective about customer by analyzing customer shopping path data because customer should go to a specific place to buy some product in store. The analysis of path data is time-consuming work. In this way, the commonly used method is clustering algorithm in order to understand the tendency of data. But a general clustering algorithm is not suitable to identify the shopping path of customers because of various spatial constraints such as aisle layout or other physical obstructions in the store. In this paper, we propose a shopping path clustering algorithm with sequence pattern matching method, LCSS (longest common subsequence) method, which groups similar moving path for shop in order to understand characteristics of shopping. Experimental results using real data obtained from a grocery in Seoul confirm the good performance of the proposed method in finding the hot spot, dead spot and major path patterns of customer movements.