ATLAS: Adaptive Topology-based Learning at Scale for Homophilic and Heterophilic Graphs

2025-12-16

AI Summary

To address the accuracy degradation of graph neural networks (GNNs) on heterophilous graphs and the scalability bottleneck caused by iterative neighborhood aggregation on large-scale graphs, this paper proposes a message-passing-free, multi-scale community topological encoding framework. The method directly encodes adaptive multi-resolution community structures—e.g., derived from Louvain or Leiden algorithms—into concatenated topological features, and employs a fully connected MLP to learn node representations, enabling unified modeling of both homophilous and heterophilous graphs. It supports sampling-free, end-to-end training, achieving high accuracy, strong scalability, and inherent interpretability. Experiments demonstrate up to 20% accuracy gain over GCN on heterophilous graphs and an 11% improvement over vanilla MLP on homophilous graphs. The framework efficiently trains on full graphs with millions of nodes and exhibits systematic performance variation with community resolution, revealing principled structure–performance relationships.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

We present ATLAS (Adaptive Topology-based Learning at Scale for Homophilic and Heterophilic Graphs), a novel graph learning algorithm that addresses two important challenges in graph neural networks (GNNs). First, the accuracy of GNNs degrades when the graph is heterophilic. Second, iterative feature aggregation limits the scalability of GNNs to large graphs. We address these challenges by extracting topological information about graph communities at multiple levels of refinement, concatenating community assignments to the feature vector, and applying multilayer perceptrons (MLPs) to the resulting representation. This provides topological context about nodes and their neighborhoods without invoking aggregation. Because MLPs are typically more scalable than GNNs, our approach applies to large graphs without the need for sampling. Across a wide set of graphs, ATLAS achieves comparable accuracy to baseline methods, with gains as high as 20 percentage points over GCN for heterophilic graphs with negative structural bias and 11 percentage points over MLP for homophilic graphs. Furthermore, we show how multi-resolution community features systematically modulate performance in both homophilic and heterophilic settings, opening a principled path toward explainable graph learning.

Problem

• Addresses GNN accuracy decline in heterophilic graphs
• Solves scalability issues of GNNs on large graphs
• Provides explainable graph learning through multi-resolution features

Innovation

• Extracts multi-level topological community information
• Concatenates community assignments to feature vectors
• Applies MLPs for scalable graph learning without aggregation