Inferring signed social networks from contact patterns

2026-01-15

AI Summary

This study addresses the challenge of distinguishing between absent interactions in contact data that arise from missed opportunities versus active avoidance—i.e., negative social ties. To this end, the authors propose a Bayesian framework that models interaction groups and employs Markov chain Monte Carlo (MCMC) inference to differentiate incidental non-interactions from deliberate avoidance, thereby reconstructing signed social networks with both positive and negative edges. Notably, this approach is the first to explicitly model avoidance behavior directly from contact data and validates model plausibility through posterior predictive checks. Experiments demonstrate that the method substantially outperforms existing baselines on synthetic data, particularly in detecting negative ties, and successfully recovers the structure of friendship surveys when applied to real-world contact data collected in a French high school, confirming its effectiveness and practical utility.

Technology Category

• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Social Networks and Social Media: Computational social science
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• User Modeling, Personalization and Recommendation: Social recommender systems and personalization for the social Web

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.

Problem

• signed social networks
• contact patterns
• negative ties
• interaction absence
• network inference

Innovation

• signed social networks
• Bayesian inference
• contact patterns
• negative ties
• MCMC