Problem
Research questions and friction points this paper is trying to address.
Graph Neural Networks
Multi-type Nodes
Efficiency
Clarify Confused Nodes via Separated Learning
🤖 AI Summary
Heterophilous nodes in real-world graphs violate the homophily assumption of Graph Neural Networks (GNNs), degrading classification performance. To address this, we propose Neighborhood Confusion (NC), a novel metric that quantifies node-wise classification confidence disparity to explicitly distinguish homophilous from heterophilous nodes—marking the first such confidence-based heterophily diagnosis. Building on NC, we design NC-guided GCN (NCGCN), which adaptively partitions nodes into homophilous and heterophilous groups, enables group-specific parameter sharing, and incorporates multi-scale neighborhood aggregation. Our approach maintains architectural simplicity while effectively decoupling heterogeneous semantics and enhancing local representation fidelity. Extensive experiments demonstrate that NCGCN significantly outperforms state-of-the-art methods on both homophilous and heterophilous graph benchmarks, validating the efficacy and generalizability of our node separation mechanism. The code is publicly available.
📝 Abstract
Graph neural networks (GNNs) have achieved remarkable advances in graph-oriented tasks. However, real-world graphs invariably contain a certain proportion of heterophilous nodes, challenging the homophily assumption of traditional GNNs and hindering their performance. Most existing studies continue to design generic models with shared weights between heterophilous and homophilous nodes. Despite the incorporation of high-order messages or multi-channel architectures, these efforts often fall short. A minority of studies attempt to train different node groups separately but suffer from inappropriate separation metrics and low efficiency. In this paper, we first propose a new metric, termed Neighborhood Confusion (NC), to facilitate a more reliable separation of nodes. We observe that node groups with different levels of NC values exhibit certain differences in intra-group accuracy and visualized embeddings. These pave the way for Neighborhood Confusion-guided Graph Convolutional Network (NCGCN), in which nodes are grouped by their NC values and accept intra-group weight sharing and message passing. Extensive experiments on both homophilous and heterophilous benchmarks demonstrate that our framework can effectively separate nodes and yield significant performance improvement compared to the latest methods. The source code will be available in https://github.com/GISec-Team/NCGNN.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Neighborhood Confusion Guided
Graph Convolutional Network
Node Classification
🔎 Similar Papers
No similar papers found.
