GraphFAS improves fraud detection with automated graph features
GraphFAS: A Distributed System for Automated Graph Feature Generation and Selection in Industrial Transaction Networks
Machine LearningArtificial IntelligenceDistributed, Parallel, and Cluster Computing
Summary
Fraud detection in financial transactions often depends on experts creating complex features that miss important connection patterns between users. The authors introduce GraphFAS, a system that automatically creates easy-to-understand features based on the network of transactions without needing complex learning models. This system selects the most useful features in a robust and scalable way, making it easier to catch fraud while explaining the results. GraphFAS was successfully used in a big payment platform, improving efficiency and matching expert-crafted and advanced learning methods.
What this means in practice
- •For financial risk teams: Reduce manual effort in building fraud detection features while maintaining interpretability and scalability on large transaction graphs.
- •For payment platform engineers: Integrate a distributed automated graph feature system that feeds traditional models and supports transparent fraud detection explanations.
Authors
Yice Luo, Yun Zhu, Xi Chen, Yongchao Liu, Xintan Zeng, Chengying Huan, Kai Zhang, Jinrui Zhang, Juelu Zhang, Jiajun Zheng
Abstract
Industrial fraud detection often relies on costly expert-crafted features that overlook graph-structured relational signals, while GNNs often do not meet the interpretability and deployment requirements of financial risk control. We propose GraphFAS (Graph Feature Automated Selection), a distributed feature selection procedure based on Boruta that bridges this gap through: (1) a non-parametric graph feature generation module that constructs explicit, interpretable structural features via multi-hop subgraph extraction and multi-scale aggregation without learned parameters; and (2) an automated distributed feature selection algorithm extending Boruta with median-based aggregation across partitions to robustly identify informative features at scale with minimal domain expertise. Compared with end-to-end GNN pipelines, GraphFAS decouples feature aggregation from model training, enabling direct integration with tabular models and direct compatibility with TreeSHAPbased explanations. Deployed in Alipay, GraphFAS delivers orderof-magnitude improvements in engineering efficiency while showing strong performance against expert-driven and graph-learning baselines on large-scale graphs.