Discovery of associated consumer demands: Construction of a co-demanded product network with community detection. (15th September 2021)

• Record Type: Journal Article
• Authors: Wu, Jiacong; Wang, Yu; Shafiee, Sara; Zhang, Dongsong
• Is Part Of: Expert systems with applications. Volume 178(2021)
• Publication Date: 2021-09-15
• Subjects: Associated consumer demands -- Co-demanded products -- Product networks -- Community detection; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.115038
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: Consumers have various product demands that are associated with each other. Online shopping behaviors are considered to construct a co-demanded product network. Several influential factors are highlighted and considered in network construction. DACCN can detect diverse, practical and reliable associations among products. Abstract: Some consumers have various product demands that are associated with each other. Although there are methods for discovering the associations among co-purchased products, they have limitations, including redundancy of the extracted association rules, the potential to miss novel and interesting associations among co-demanded products hidden in shopping behaviors, and neglect of several important influential factors. In order to provide effective product recommendations, it is necessary and beneficial to discover the associations among the products co-demanded by the same consumers in a short period based on the consumers' shopping behaviors. Therefore, this paper proposes a novel model for discovering associated consumer demands based on a co-demanded product network. First, the model identifies each consumer's product demands and calculates their intensity based on various online shopping behaviors. Second, the model constructs a co-demanded product network based on the products demanded by the same consumers within a short period. The model also considers several important factors, previous neglected in the literature, that can improve the detection of associations among co-demanded products, including the time interval between two product demands from the same consumers, the popularity of each demanded product, and the number of product demands from each consumer. Third, the model uses an algorithm for the detection of overlapping communities to identify the tightly connected co-demanded products within the network as communities of associated consumer demands, and ranks them based on their information density. We use a real-world dataset collected from a well-known e-commerce platform to validate the proposed model. The results show that the proposed model can detect more modular, diverse, practical, and reliable communities of associated products than the existing network analysis–based market basket analysis methods.