Spherical latent space models for social network analysis

TL;DR

This paper proposes a spherical latent space model for social networks that better captures community structures and cyclical patterns, improving interpretability and model fit over traditional Euclidean models.

Contribution

It introduces a novel spherical embedding approach for social network analysis, addressing degeneracy and enhancing the representation of complex relationships.

Findings

Improved model fit on benchmark datasets

Enhanced interpretability of social structures

Better representation of cyclical patterns

Abstract

This article introduces a spherical latent space model for social network analysis, embedding actors on a hypersphere rather than in Euclidean space as in standard latent space models. The spherical geometry facilitates the representation of transitive relationships and community structure, naturally captures cyclical patterns, and ensures bounded distances, thereby mitigating degeneracy issues common in traditional approaches. Bayesian inference is performed via Markov chain Monte Carlo methods to estimate both latent positions and other model parameters. The approach is demonstrated using two benchmark social network datasets, yielding improved model fit and interpretability relative to conventional latent space models.