Product Map Analysis from a Crowd of Small- and Medium-Sized E-Commerce Sites: A Bottom-Up Approach

Abstract

The study of product maps in e-commerce has garnered significant attention from academics and practitioners, as they provide insights into the relationship between products, such as complementarity and competition. However, existing studies have focused on the perspectives of large manufacturers and retailers, using data from these central sources. This paper adopts a bottom-up approach based on crowd intelligence, with small- and medium-sized e-commerce (SME) sites serving as independent data providers. This approach allows for the decentralized processing of data and enables the aggregation of diverse perspectives and insights from a large number of independent sources. A graph term frequency-inverse document frequency method is proposed, which can measure the similarities of products and build a product map. The method was employed to find a hierarchical community structure using data from over 90 000 products from 52 SME sites. The results showed that products within the same site tend to be distributed across the same community. Our findings can assist e-commerce sites in making informed decisions about pricing and product offerings, leading to more diversified production.