Chi-Square Wavelet Graph Neural Networks for Heterogeneous Graph Anomaly Detection

2025-05-25

AI Summary

Addressing three key challenges in heterogeneous graph anomaly detection—difficult heterogeneous semantic modeling, loss of high-frequency anomaly signals, and severe class imbalance—this paper proposes ChiGAD. First, it introduces a novel chi-square-distribution-based wavelet filter that precisely preserves high-frequency anomaly components in the spectral domain. Second, it designs multi-graph chi-square filtering coupled with interactive meta-graph convolution to jointly model anomaly propagation paths across diverse node and edge types. Third, it proposes a contribution-aware cross-entropy loss that dynamically reweights hard-to-classify anomalous samples. ChiGAD is the first framework to systematically integrate wavelet spectral theory into heterogeneous graph anomaly detection. Extensive experiments on multiple public and industrial datasets demonstrate significant improvements over state-of-the-art methods. Moreover, its homogeneous variant, ChiGNN, achieves top performance on seven standard benchmarks.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Graph Anomaly Detection (GAD) in heterogeneous networks presents unique challenges due to node and edge heterogeneity. Existing Graph Neural Network (GNN) methods primarily focus on homogeneous GAD and thus fail to address three key issues: (C1) Capturing abnormal signal and rich semantics across diverse meta-paths; (C2) Retaining high-frequency content in HIN dimension alignment; and (C3) Learning effectively from difficult anomaly samples with class imbalance. To overcome these, we propose ChiGAD, a spectral GNN framework based on a novel Chi-Square filter, inspired by the wavelet effectiveness in diverse domains. Specifically, ChiGAD consists of: (1) Multi-Graph Chi-Square Filter, which captures anomalous information via applying dedicated Chi-Square filters to each meta-path graph; (2) Interactive Meta-Graph Convolution, which aligns features while preserving high-frequency information and incorporates heterogeneous messages by a unified Chi-Square Filter; and (3) Contribution-Informed Cross-Entropy Loss, which prioritizes difficult anomalies to address class imbalance. Extensive experiments on public and industrial datasets show that ChiGAD outperforms state-of-the-art models on multiple metrics. Additionally, its homogeneous variant, ChiGNN, excels on seven GAD datasets, validating the effectiveness of Chi-Square filters. Our code is available at https://github.com/HsipingLi/ChiGAD.

Problem

• Detecting anomalies in heterogeneous graphs with node and edge diversity
• Capturing abnormal signals across diverse meta-paths in heterogeneous networks
• Addressing class imbalance and high-frequency retention in anomaly detection

Innovation

• Multi-Graph Chi-Square Filter for anomaly detection
• Interactive Meta-Graph Convolution preserves high-frequency data
• Contribution-Informed Loss addresses class imbalance