Contextual Graph Markov Model: A Deep and Generative Approach to Graph Processing

TL;DR

The paper presents the Contextual Graph Markov Model, a deep generative neural network approach for scalable graph data processing and classification, combining probabilistic modeling with efficient context diffusion.

Contribution

It introduces a novel deep architecture that integrates probabilistic models with neural networks for scalable, context-aware graph processing.

Findings

Effective in structure classification benchmarks

Scalable and efficient context diffusion across graph vertices and edges

Combines generative modeling with discriminative classifiers

Abstract

We introduce the Contextual Graph Markov Model, an approach combining ideas from generative models and neural networks for the processing of graph data. It founds on a constructive methodology to build a deep architecture comprising layers of probabilistic models that learn to encode the structured information in an incremental fashion. Context is diffused in an efficient and scalable way across the graph vertexes and edges. The resulting graph encoding is used in combination with discriminative models to address structure classification benchmarks.