InfraredGP: Efficient Graph Partitioning via Spectral Graph Neural Networks with Negative Corrections

2025-08-27

AI Summary

To address the limited representation capability of low-frequency signals in graph community detection, this paper proposes InfraredGP—a training-free spectral graph neural network. Its core innovation is a novel negative correction mechanism that extends the frequency response range of the graph Laplacian operator beyond the conventional [0, 2] interval, thereby uncovering the strong discriminative power of ultra-low-frequency (< 0) signals for community structure. InfraredGP generates separable node embeddings via low-pass filtering, random input initialization, and a single forward pass, then integrates BIRCH for efficient clustering. Evaluated on both static and streaming graph partitioning tasks, InfraredGP achieves 16–23× speedup over baseline methods while maintaining comparable partition quality. The implementation is publicly available.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Content-based information diffusion
• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks

Abstract

Graph partitioning (GP), a.k.a. community detection, is a classic problem that divides nodes of a graph into densely-connected blocks. From a perspective of graph signal processing, we find that graph Laplacian with a negative correction can derive graph frequencies beyond the conventional range $[0, 2]$. To explore whether the low-frequency information beyond this range can encode more informative properties about community structures, we propose InfraredGP. It ( omannumeral1) adopts a spectral GNN as its backbone combined with low-pass filters and a negative correction mechanism, ( omannumeral2) only feeds random inputs to this backbone, ( omannumeral3) derives graph embeddings via one feed-forward propagation (FFP) without any training, and ( omannumeral4) obtains feasible GP results by feeding the derived embeddings to BIRCH. Surprisingly, our experiments demonstrate that based solely on the negative correction mechanism that amplifies low-frequency information beyond $[0, 2]$, InfraredGP can derive distinguishable embeddings for some standard clustering modules (e.g., BIRCH) and obtain high-quality results for GP without any training. Following the IEEE HPEC Graph Challenge benchmark, we evaluate InfraredGP for both static and streaming GP, where InfraredGP can achieve much better efficiency (e.g., 16x-23x faster) and competitive quality over various baselines. We have made our code public at https://github.com/KuroginQin/InfraredGP

Problem

• Efficient graph partitioning via spectral GNN with negative corrections
• Exploring low-frequency information beyond conventional range for community detection
• Achieving high-quality results without training using one feed-forward propagation

Innovation

• Spectral GNN with negative correction mechanism
• Random inputs with one feed-forward propagation
• Untrained embeddings processed by BIRCH clustering