Random Field Models for Graph Layout

Abstract

We propose a new framework for the layout of graphs, which both unifies and generalizes many approaches known from the literature. Doing so, we briefly survey a number of models, particularly including force directed placement approaches. Our model is based on the stochastic concept of random fields, which have applications in various other areas such as physics, biology, economics, or medical imaging. By allowing, yet not enforcing, the decoupling of model and computation, it integrates algorithmic and declarative approaches to graph layout.