摘要
In this research paper, we introduce a novel model for product segmentation based on data collected from coin-less vending machines called SFM (Steadiness, Frequency and Monetary) in order to track customer behavior through each segments of products and provide vending machine owner the more reasonably way to change the items on the vending machine. We adopted two well-known algorithms from data mining techniques which are association rules mining and K-means clustering. Apriori together with k-means were used in our proposed methodology. After performing the method, we tried our theory on the machines. The results from the real-word experiment showed that there were only 25% of new products that were in low value segment after applied our method while there were 67% before the usage.