Skip to Main Content
Online shoppers purchase a wide variety of items ranging from books to electronics. Keeping in view of the large number of items available for online purchase, there is a need for sophisticated techniques to help users explore the available products. Composite items, which associate a central item with a set of packages formed by satellite items, can be used for this purpose. The existing approach for constructing and exploring composite items use a random walk algorithm for this purpose but has some problems. In this paper, we propose a statistically efficient algorithm to construct and explore the composite items in real time. The proposed approach considers that not only the price but also the quality of the product influences the consumer behavior. The experimental study is conducted to show that the proposed approach significantly improves the performance in terms of time, search space and the quality of the recommendation provided using the composite items over the existing method.