Deep Fraud Detection on Non-attributed Graph

Abstract

Fraud detection problems are usually formulated as a machine learning problem on a graph. Recently, Graph Neural Networks (GNNs) have shown solid performance on fraud detection. The successes of most previous methods heavily rely on rich node features and high-fidelity labels. However, labeled data is scarce in large-scale industrial problems, especially for fraud detection where new patterns emerge from time to time. Meanwhile, node features are also limited due to privacy and other constraints. In this paper, two improvements are proposed: 1) We design a graph transformation method capturing the structural information to facilitate GNNs on non-attributed fraud graphs. 2) We propose a novel graph pre-training strategy to leverage more unlabeled data via contrastive learning. Experiments on a large-scale industrial dataset demonstrate the effectiveness of the proposed framework for fraud detection.

References in corpus (8)

• Inductive Representation Learning on Large Graphs
• Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters
• How Powerful are Graph Neural Networks?
• CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances
• Combining Label Propagation and Simple Models Out-performs Graph Neural Networks
• Unsupervised Outlier Detection using Memory and Contrastive Learning
• On Node Features for Graph Neural Networks
• On Positional and Structural Node Features for Graph Neural Networks on Non-attributed Graphs