Determining profitable products in the retail market with consideration of cash limitation and exhibition periods

Abstract

In the selection of profitable products, consumer preferences and retailer constraints in products supply must be considered. When data mining algorithms are used to discover the consumer's preferences from transaction database, the results may be biased, if the exhibition period of the products has not be considered. In this study a new method is proposed to adjust the support and confidence coefficients of traditional association rule mining algorithms such as Apriori or FP-growth taking into consideration of common exhibition periods. On the supply side, the retailer may have some limitations in terms of buying and supplying products in the store. In the most recent researches, only the shelf space constraint has been considered. In this study, financing as an important constraint in the retail market and the opportunity cost of money are imported in the selection of profitable products.The researcher's experiment on real world data shows that the number of frequent itemsets increases significantly when products exhibition periods are taken into consideration. In this case, if the retailer considers the opportunity cost of money as 1%, the profitable set composition will be changed by 10%. Also, when the opportunity cost is 1% and the retailer faces cash limitation, the number of products is reduced by 21% in the most profitable set, whereas the new set composition is 29.4% different from the base set.