Taking inspiration from real-life friendship formation patterns, we propose a new generative model of evolving social networks. Each person in the network has a distribution over social interaction spheres, which we term "contexts." The model allows for birth and death of links and addition of new people. Model parameters are learned via Gibbs sampling, and results are demonstrated on real social networks. We study the robustness of our model by examining statistical properties of simulated networks, and compare against well-known properties of real social networks.

16 pages

*Robotics Institute, Carnegie Mellon University

This page maintained by reports@cs.cmu.edu