Research Summary: netseg: a Python Package for Measuring Structural Polarization and Segregation in Social Networks

A new Python package, 'netseg', has been developed to quantify structural polarization and segregation within social networks. This package implements a comprehensive set of established network indices, many of which have been generalized for multi-group analysis and both directed and undirected networks. It aims to address the common issue of unreleased or untested code for these analytical tools, providing a documented and tested solution.

Why it matters

The introduction of a standardized, tested, and documented tool for analyzing network polarization and segregation enhances the reliability and comparability of research in this domain. This can provide more robust insights into group dynamics, information flow, and potential societal divisions, informing strategic decisions related to communication, community building, and risk mitigation.

Key insights

• The 'netseg' Python package quantifies structural polarization and segregation in social networks.
• It implements widely cited network indices, generalized for more than two groups and for directed/undirected networks.
• The package addresses a gap where code for these indices is often not released or tested.
• It ports functionalities from an existing R package and introduces new measures like Random Walk Controversy, Boundary Connectivity, Dipole Moment, and Moran's I.
• The package operates on 'igraph' objects, performing underlying graph operations.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2609.16088