Graph Homophily Booster: Rethinking the Role of Discrete Features on Heterophilic Graphs

2025-09-15

AI Summary

Graph Neural Networks (GNNs) suffer significant performance degradation on heterophilic graphs, where adjacent nodes exhibit large feature and label discrepancies, undermining the efficacy of message passing. Existing approaches primarily modify model architectures without directly addressing the root cause of heterophily—often underperforming even simple MLPs. This paper proposes GRAPHITE, the first framework to explicitly enhance graph structural homophily at the data level by introducing learnable auxiliary feature nodes. Leveraging a homophily-aware metric, GRAPHITE constructs these auxiliary nodes to promote homogenized message propagation, while jointly optimizing them with a lightweight MLP for node classification. On highly heterophilic benchmarks (e.g., Actor), GRAPHITE substantially outperforms state-of-the-art methods; it also maintains competitive accuracy on homophilic graphs. By shifting focus from model-centric design to data-level homophily enhancement, GRAPHITE breaks the paradigm limitation of relying solely on architectural innovations.

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: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Graph neural networks (GNNs) have emerged as a powerful tool for modeling graph-structured data. However, existing GNNs often struggle with heterophilic graphs, where connected nodes tend to have dissimilar features or labels. While numerous methods have been proposed to address this challenge, they primarily focus on architectural designs without directly targeting the root cause of the heterophily problem. These approaches still perform even worse than the simplest MLPs on challenging heterophilic datasets. For instance, our experiments show that 21 latest GNNs still fall behind the MLP on the Actor dataset. This critical challenge calls for an innovative approach to addressing graph heterophily beyond architectural designs. To bridge this gap, we propose and study a new and unexplored paradigm: directly increasing the graph homophily via a carefully designed graph transformation. In this work, we present a simple yet effective framework called GRAPHITE to address graph heterophily. To the best of our knowledge, this work is the first method that explicitly transforms the graph to directly improve the graph homophily. Stemmed from the exact definition of homophily, our proposed GRAPHITE creates feature nodes to facilitate homophilic message passing between nodes that share similar features. Furthermore, we both theoretically and empirically show that our proposed GRAPHITE significantly increases the homophily of originally heterophilic graphs, with only a slight increase in the graph size. Extensive experiments on challenging datasets demonstrate that our proposed GRAPHITE significantly outperforms state-of-the-art methods on heterophilic graphs while achieving comparable accuracy with state-of-the-art methods on homophilic graphs.

Problem

• Addressing poor performance of GNNs on heterophilic graphs
• Transforming graph structure to directly increase homophily
• Creating feature nodes to enable homophilic message passing

Innovation

• Transforms graph to increase homophily directly
• Creates feature nodes for homophilic message passing
• Boosts heterophilic graph performance with minimal size increase