Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters

By Yingtong Dou, Zhiwei Liu, Li Sun, Yutong Deng, Hao Peng, Philip S. Yu. International Conference on Information and Knowledge Management, 2020

744 citations recorded by the source corpus.

Graph Neural Networks (GNNs) have been widely applied to fraud detection problems in recent years, revealing the suspiciousness of nodes by aggregating their neighborhood information via different relations. However, few prior works have noticed the camouflage behavior of fraudsters, which could hamper the performance of GNN-based fraud detectors during the aggregation process. In this paper, we introduce two types of camouflages based on recent empirical studies, i.e., the feature camouflage and the relation camouflage. Existing GNNs have not addressed these two camouflages, which results in their poor performance in fraud detection problems. Alternatively, we propose a new model named CAmouflage-REsistant GNN (CARE-GNN), to enhance the GNN aggregation process with three unique modules against camouflages. Concretely, we first devise a label-aware similarity measure to find informative neighboring nodes. Then, we leverage reinforcement learning (RL) to find the optimal amounts of neighbors to be selected. Finally, the selected neighbors across different relations are aggregated together. Comprehensive experiments on two real-world fraud datasets demonstrate the effectiveness of the RL algorithm. The proposed CARE-GNN also outperforms state-of-the-art GNNs and GNN-based fraud detectors. We integrate all GNN-based fraud detectors as an opensource toolbox https://github.com/safe-graph/DGFraud. The CARE-GNN code and datasets are available at https://github.com/YingtongDou/CARE-GNN.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/14156438bafed28a626738630b5181b83ed5d79c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xiyang Hu
2. Songqiao Han
3. Hailiang Huang
4. Mingqi Jiang
5. Zi-Chao Gao
6. Qing He
7. Yue Zhao
8. Nedim Lipka

Citations in the corpus

• Graph Neural Networks with Heterophily (2020, background): “…under implicit homophily assumptions suffer from poor performance in heterophily settings, which can be problematic for applications like fraudster detection (Pandit et al. 2007; Dou et al. 2020) or analysis of protein structures (Fout et al. 2017), where heterophilous connections are common.”
• Pick and Choose: A GNN-based Imbalanced Learning Approach for Fraud Detection (2021, methodology): “CARE [10] enhances the GNN aggregation process against camouflage for opinion fraud detection.”
• ADBench: Anomaly Detection Benchmark (2022, background): “Camouflage detection finds concealed objects in the background, e.g., camouflaged anomalies blurred with normal objects [110], where camouflage-resistant AD methods [40] help detect concealed objects (that look normal but are abnormal).”
• Rethinking Graph Neural Networks for Anomaly Detection (2022, methodology): “…(Wang et al., 2019a; Cui et al., 2020; Liu et al., 2021a), (2) using the resampling strategy to aggregate neighborhood information selectively (Dou et al., 2020; Liu et al., 2020; 2021b), (3) designing auxiliary losses to enhance the network training power (Ding et al., 2019; Zhao et al.,…”