DuoGNN: Topology-aware Graph Neural Network with Homophily and Heterophily Interaction-Decoupling

TL;DR

DuoGNN is a novel graph neural network architecture that effectively decouples homophilic and heterophilic interactions using topology-aware techniques, improving message passing and scalability across diverse graph structures.

Contribution

The paper introduces a scalable, topology-aware GNN architecture with novel edge-filtering and graph condensation methods to handle both homophilic and heterophilic interactions.

Findings

Consistent improvements on medical and non-medical datasets.

Effective decoupling of homophilic and heterophilic edges.

Enhanced scalability and generalization across graph topologies.

Abstract

Graph Neural Networks (GNNs) have proven effective in various medical imaging applications, such as automated disease diagnosis. However, due to the local neighborhood aggregation paradigm in message passing which characterizes these models, they inherently suffer from two fundamental limitations: first, indistinguishable node embeddings due to heterophilic node aggregation (known as over-smoothing), and second, impaired message passing due to aggregation through graph bottlenecks (known as over-squashing). These challenges hinder the model expressiveness and prevent us from using deeper models to capture long-range node dependencies within the graph. Popular solutions in the literature are either too expensive to process large graphs due to high time complexity or do not generalize across all graph topologies. To address these limitations, we propose DuoGNN, a scalable and generalizable architecture which leverages topology to decouple homophilic and heterophilic edges and capture both short-range and long-range interactions. Our three core contributions introduce (i) a topological edge-filtering algorithm which extracts homophilic interactions and enables the model to generalize well for any graph topology, (ii) a heterophilic graph condensation technique which extracts heterophilic interactions and ensures scalability, and (iii) a dual homophilic and heterophilic aggregation pipeline which prevents over-smoothing and over-squashing during the message passing. We benchmark our model on medical and non-medical node classification datasets and compare it with its variants, showing consistent improvements across all tasks. Our DuoGNN code is available at https://github.com/basiralab/DuoGNN.