Extracting Complements and Substitutes from Sales Data: A Network Perspective
arXiv:2103.02042 · doi:10.1140/epjds/s13688-021-00297-4
Abstract
The complementarity and substitutability between products are essential concepts in retail and marketing. Qualitatively, two products are said to be substitutable if a customer can replace one product by the other, while they are complementary if they tend to be bought together. In this article, we take a network perspective to help automatically identify complements and substitutes from sales transaction data. Starting from a bipartite product-purchase network representation, with both transaction nodes and product nodes, we develop appropriate null models to infer significant relations, either complements or substitutes, between products, and design measures based on random walks to quantify their importance. The resulting unipartite networks between products are then analysed with community detection methods, in order to find groups of similar products for the different types of relationships. The results are validated by combining observations from a real-world basket dataset with the existing product hierarchy, as well as a large-scale flavour compound and recipe dataset.
arXiv comment41 pages, 17 figures
References in corpus (15)
- Community detection in graphs
- From Louvain to Leiden: guaranteeing well-connected communities
- Finding community structure in networks using the eigenvectors of matrices
- Random graphs with arbitrary degree distributions and their applications
- Who is the best connected scientist? A study of scientific coauthorship networks
- Community detection algorithms: a comparative analysis
- Stochastic blockmodels and community structure in networks
- Community detection in networks: A user guide
- How to project a bipartite network?
- Vertex similarity in networks
- The map equation
- Random Walks, Markov Processes and the Multiscale Modular Organization of Complex Networks
- Laplacian Dynamics and Multiscale Modular Structure in Networks
- Estimating the resolution limit of the map equation in community detection
- Multiscale dynamical embeddings of complex networks