Overlap versus partition: Marketing classification and customer profiling in complex networks of products

Abstract

In recent years we witnessed the explosion in the availability of data regarding human and customer behavior in the market. This data richness era has fostered the development of useful applications in understanding how markets and the minds of the customers work. In this paper we focus on the analysis of complex networks based on customer behavior. Complex network analysis has provided a new and wide toolbox for the classic data mining task of clustering. With community discovery, i.e. the detection of functional modules in complex networks, we are now able to group together customers and products using a variety of different criteria. The aim of this paper is to explore this new analytic degree of freedom. We are interested in providing a case study uncovering the meaning of different community discovery algorithms on a network of products connected together because co-purchased by the same customers. We focus our interest in the different interpretation of a partition approach, where each product belongs to a single community, against an overlapping approach, where each product can belong to multiple communities. We found that the former is useful to improve the marketing classification of products, while the latter is able to create a collection of different customer profiles

The regionalization of intergovernmental organization networks: A non-linear process

“…Ahn et al (2010) developed HLC in response to node-based clustering techniques that force nodes into one community. HLC has been applied to a wide variety of highly redundant networks, including bipartite citation graphs (Havemann et al, 2012), product marketing networks (Pennacchioli et al, 2014) and large-scale social networks (Ahn et al, 2010).…”

Building Mini-Categories in Product Networks

“…Pennacchioli et al [10] compare two community discovery approaches: a partitioning approach, where each product belongs to a single community, and an overlapping approach, where each product may belong to multiple communities. The authors apply the approaches to a data set of an Italian retailer and find that the former is useful to improve product classification while the latter can create a collection of different customer profiles.…”

“…A product network [3,6,10] is a graph G reflecting the product co-occurrences in a customer's "basket" [6,8,11]. The graph nodes represent individual material items purchased by customers.…”

“…Detecting communities in complex networks is known as "community discovery" [9]. In recent years, it has been one of the most prolific sub-branches of complex network analysis, with dozens of algorithms proposed and the agreement within the scientific community that there is no unique solution to this problem given the many different possible definitions of "community" for different applications [10]. Raeder and Chawla [8] focus on community discovery in product networks and show how communities of products can be used to gain insight into customer behavior.…”

Combined method to detect communities in graphs of interacting objects

“…In the general case, this task is NP-complete, so the paper proposes to use a method with a number of terminal options since it is required to deal with graphs from various fields [5,6]. These options are designed to possibly refine the resulting community structure and are based on the initial graph topology and corresponding field.…”