Inferring signed social networks from contact patterns

TL;DR

This paper introduces a Bayesian method to infer signed social networks from contact data, distinguishing between absence of interaction due to opportunity or active avoidance, and validates its effectiveness on synthetic and real data.

Contribution

It presents a novel Bayesian framework with MCMC inference for accurately inferring signed networks from contact patterns, addressing a key challenge in social network analysis.

Findings

The method outperforms baselines in detecting negative edges.

Application to high school data reveals structures consistent with surveys.

Model passes posterior predictive checks indicating good fit.

Abstract

Social networks are typically inferred from indirect observations, such as proximity data; yet, most methods cannot distinguish between absent relationships and actual negative ties, as both can result in few or no interactions. We address the challenge of inferring signed networks from contact patterns while accounting for whether lack of interactions reflect a lack of opportunity as opposed to active avoidance. We develop a Bayesian framework with MCMC inference that models interaction groups to separate chance from choice when no interactions are observed. Validation on synthetic data demonstrates superior performance compared to natural baselines, particularly in detecting negative edges. We apply our method to French high school contact data to reveal a structure consistent with friendship surveys and demonstrate the model's adequacy through posterior predictive checks.