An improved scheme for determining top-revenue itemsets for placement in retail businesses
- PDF / 2,292,899 Bytes
- 17 Pages / 595.276 x 790.866 pts Page_size
- 7 Downloads / 178 Views
REGULAR PAPER
An improved scheme for determining top-revenue itemsets for placement in retail businesses Parul Chaudhary1
· Anirban Mondal2 · Polepalli Krishna Reddy3
Received: 8 September 2019 / Accepted: 24 April 2020 © Springer Nature Switzerland AG 2020
Abstract Utility mining has been emerging as an important area in data mining. While existing works on utility mining for retail businesses have primarily focused on the problem of finding high-utility itemsets from transactional databases, they implicitly assume that each item occupies only one slot. Here, the slot size of a given item is the number of (integer) slots occupied by that item on the retail store shelves. However, in many real-world scenarios, the number of slots consumed by different items typically varies. Hence, this paper considers that a given item may physically occupy any fixed (integer) number of slots. Thus, we address the problem of efficiently determining the top-utility itemsets when a given number of slots is specified as input. The key contributions of our work are three fold. First, we present an efficient framework to determine the top-utility itemsets for different user-specified number of slots that need to be filled. Second, we propose a novel flexible and efficient index, designated as Slot Type Utility (STU) index, for facilitating quick retrieval of the top-utility itemsets for a given number of slots. Third, we conducted an extensive performance evaluation using both real and synthetic datasets to demonstrate the overall effectiveness of the STU index in quickly retrieving the top-utility itemsets by considering a placement scheme in terms of execution time and utility (net revenue) as compared to recent existing schemes. Keywords High-utility itemset mining · Top-k mining · Retailing · Supermarkets · Product placement · Indexing
1 Introduction Over the past decade, we have been witnessing the prevalence of several popular medium-to-mega-sized retail stores with relatively huge retail floor space. Walmart Supercenters are examples of medium-sized retail stores, and Macy’s Department Store at Herald Square (New York City, USA), Dubai Mall (Dubai), Shinsegae Centum city Department Store (Busan, South Korea) are examples of mega-sized retail stores. Typically, such mega-sized stores have more than a million square feet of retail floor space [1]. In this regard,
B
Parul Chaudhary [email protected] Anirban Mondal [email protected] Polepalli Krishna Reddy [email protected]
1
Shiv Nadar University, Greater Noida, India
2
Ashoka University, Sonipat, India
3
International Institute of Information Technology, Hyderabad, India
research efforts are going on to address issues such as scalable supply chain management [2], inventory management [3], stock-out management [4,5] and placement of items (products) [6] for strategically improving the revenue of retail stores. In this paper, we address the problem of item (and itemset) placement for revenue improvement of the retailer. Incidentally, the placement of items i
Data Loading...
