# A mixture model approach for clustering bipartite networks

## TL;DR

This paper introduces a flexible mixture model for clustering bipartite networks, capturing latent groups and individual propensities, and demonstrates its application on terrorist network data to identify key groups and traits.

## Contribution

It presents a novel mixture model approach for bipartite network clustering that accounts for both group structure and individual variability, estimated efficiently via variational inference.

## Key findings

- Identified main latent groups of terrorists.
- Estimated latent trait scores for individuals.
- Demonstrated model's effectiveness on real terrorist network data.

## Abstract

This chapter investigates the latent structure of bipartite networks via a model-based clustering approach which is able to capture both latent groups of sending nodes and latent variability of the propensity of sending nodes to create links with receiving nodes within each group. This modelling approach is very flexible and can be estimated by using fast inferential approaches such as variational inference. We apply this model to the analysis of a terrorist network in order to identify the main latent groups of terrorists and their latent trait scores based on their attendance to some events.