Block-Structure Based Time-Series Models For Graph Sequences

TL;DR

This paper introduces two novel models for graph sequences that incorporate community and link persistence, along with efficient inference algorithms leveraging existing community detection methods, validated through experiments.

Contribution

It proposes two new models for graph sequences capturing community and link persistence, with efficient inference algorithms based on single-graph community detection.

Findings

Models effectively capture community and link persistence.

Algorithms demonstrate statistical and computational efficiency.

Experimental results validate model suitability on synthetic and real data.

Abstract

Although the computational and statistical trade-off for modeling single graphs, for instance, using block models is relatively well understood, extending such results to sequences of graphs has proven to be difficult. In this work, we take a step in this direction by proposing two models for graph sequences that capture: (a) link persistence between nodes across time, and (b) community persistence of each node across time. In the first model, we assume that the latent community of each node does not change over time, and in the second model we relax this assumption suitably. For both of these proposed models, we provide statistically and computationally efficient inference algorithms, whose unique feature is that they leverage community detection methods that work on single graphs. We also provide experimental results validating the suitability of our models and methods on synthetic and real instances.