Abstract:Complex network data problems are increasingly common in many fields of application. Our motivation is drawn from strategic marketing studies monitoring customer choices of specific products, along with co-subscription networks encoding multiple purchasing behavior. Data are available for several agencies within the same insurance company, and our goal is to efficiently exploit co-subscription networks to inform targeted advertising of cross-sell strategies to currently mono-product customers. We address this goal by developing a Bayesian hierarchical model, which clusters agencies according to common mono-product customer choices and co-subscription networks. Within each cluster, we efficiently model customer behavior via a cluster-dependent mixture of latent eigenmodels. This formulation provides key information on mono-product customer choices and multiple purchasing behavior within each cluster, informing targeted cross-sell strategies. We develop simple algorithms for tractable inference, and assess performance in simulations and an application to business intelligence.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to efficiently analyze and utilize data on customers' selection of specific products and multi - purchase behaviors in the co - subscription network by using the Bayesian modeling method in complex business intelligence problems, so as to guide the cross - selling strategies for single - product customers. Specifically, the authors focus on formulating effective cross - selling strategies by analyzing the purchase choices of single - product customers and the co - subscription network in multiple branches of insurance companies, thereby increasing the multi - product purchase rate of these customers, improving customer retention, and reducing customer churn. The core objective of the paper is to develop a Bayesian hierarchical model that can cluster branches according to the common single - product customer choices and co - subscription networks. Within each cluster, customer behaviors are effectively modeled through a cluster - dependent latent feature model - mixture model. This model provides key information about single - product customer choices and multi - purchase behaviors within each cluster, thus providing a basis for targeted cross - selling strategies. In addition, the author has also developed simple algorithms to achieve feasible inferences and evaluated the performance of the model through simulations and practical applications. The key to the paper lies in how to process complex network data through statistical methods to support business decisions, especially cross - selling strategies for single - product customers. This not only helps to improve marketing efficiency but also enhances customer satisfaction and reduces customer churn caused by frequent aimless cross - selling attempts.

Bayesian modeling of networks in complex business intelligence problems

A Mixed Bayesian Network for Two-Dimensional Decision Modeling of Departure Time and Mode Choice

Bayesian model-based clustering for populations of network data

Bayesian inference of financial networks

Bayesian probabilistic models for corporate context, with an application to internal audit activities

A Bayesian Model of node interaction in networks

Structural Regularity Exploration in Multidimensional Networks Via Bayesian Inference

A Bayesian Network-based Customer Satisfaction Model: a Tool for Management Decisions in Railway Transport

A modeling framework for detecting and leveraging node-level information in Bayesian network inference

Advances in Bayesian network modelling: Integration of modelling technologies

Bayesian Modeling of Marketing Attribution

Network Structure Exploration Via Bayesian Nonparametric Models

Bayesian dynamic financial networks with time-varying predictors

Non-parametric Bayesian modeling of complex networks

Bayesian Inference of Networks Across Multiple Sample Groups and Data Types

Scalable Inference of Customer Similarities from Interactions Data using Dirichlet Processes

Network and interaction models for data with hierarchical granularity via fragmentation and coagulation

Bayesian network structural learning from complex survey data: a resampling based approach

Customer online shopping behaviours analysis using bayesian networks

Complex Networks Identification Using Bayesian Model With Independent Laplace Prior

A robust Bayesian latent position approach for community detection in networks with continuous attributes