Authors: Ali Akbar Niknafs, Mohammad Ebrahim Shiri
Addresses: Faculty of Mathematics and Computer Sciences, University of Shahid Bahonar, 22 Bahman Blvd., Kerman, Iran. ' Department of Computer Sciences, Faculty of Mathematics & Computer Sciences, Amirkabir University of Technology, Hafez Ave., Tehran, Iran
Abstract: This paper presents a new algorithm for recommender systems applied to smart carts. As the customers pass through the store|s aisles, they place their desired products in their cart baskets. In many instances, the customers pick an item from the shelf and place it in the basket but after a while they find a similar item with different specifications. The difference may be in price, quality, weight or other factors. In our proposed plan, based on the customer|s decision from choosing the first item and replacing it with another item from the same group, there will be an attempt to identify the customer|s taste and accordingly recommend a third, fourth, etc., item that might also meet his/her needs. The complete algorithm is introduced as a systematic procedure and the implementation results are shown. The proposed recommender system is designed based on the features of smart carts. We have simulated a part of the smart cart in an application, named NIKSHIRI-Shop, using C# and SQL.
Keywords: smart carts; recommender systems; shopping buddy; e-commerce; electronic commerce; online shopping.
International Journal of Business Information Systems, 2008 Vol.3 No.3, pp.284 - 299
Published online: 22 Feb 2008 *Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article