Growing Better Graphs With Latent-Variable Probabilistic Graph Grammars
This work addresses the challenge of improving graph generation models for applications in network analysis, though it appears incremental as it builds on prior HRG methods.
The paper tackles the problem of generating realistic graphs by extending probabilistic hyperedge replacement grammars (HRGs) with latent variables trained via Expectation-Maximization, resulting in models that consistently outperform existing methods in generalizing to test data as measured by likelihood.
Recent work in graph models has found that probabilistic hyperedge replacement grammars (HRGs) can be extracted from graphs and used to generate new random graphs with graph properties and substructures close to the original. In this paper, we show how to add latent variables to the model, trained using Expectation-Maximization, to generate still better graphs, that is, ones that generalize better to the test data. We evaluate the new method by separating training and test graphs, building the model on the former and measuring the likelihood of the latter, as a more stringent test of how well the model can generalize to new graphs. On this metric, we find that our latent-variable HRGs consistently outperform several existing graph models and provide interesting insights into the building blocks of real world networks.