A latent space model for multilayer network data

A Bayesian statistical model to simultaneously characterize two or more social networks defined over a common set of actors is proposed. The key feature the hierarchical prior distribution that allows user represent entire system jointly, achieving compromise between dependent and independent networks. Among others things, such specification provides an easy way visualize multilayer network data in low-dimensional Euclidean space, generate weighted reflects consensus affinity actors, establish measure correlation networks, assess cognitive judgments subjects form about relationships among perform clustering tasks at different instances. model's capabilities are illustrated using real-world synthetic datasets, taking into account types sizes, relations.