Item-Triggered Recommendation for Identifying Potential Customers of Cold Sellers in Supermarkets

Recommendation has achieved successful results in many applications. However, for supermarkets, since the transaction data is extremely skewed in the sense that a large portion of sales is concentrated in a small number of hot seller items, collaborative filtering recommenders usually recommend hot sellers while rarely recommend cold sellers. But recommenders are supposed to provide better campaigns for cold sellers to increase sales. In this paper, we propose an alternative “item-triggered” recommendation, which aims at returning a ranked list of potential customers for a given cold-seller item. This problem can be formulated as a problem of rare class learning. We present a Boosting-SVM algorithm to solve the rare class problem and apply our algorithm to a real-world supermarket database. Experimental results show that our algorithm can improve from a baseline approach by about twenty-five percent in terms of the area under the ROC curve (AUC) metric for cold sellers that as low as 0.7% of customers have ever purchased.