Representation Learning on Large Non-Bipartite Transaction Networks using GraphSAGE

2025-09-12

AI Summary

Traditional graph embedding methods struggle to simultaneously ensure scalability and generalization to unseen nodes in large-scale, dynamically evolving, non-bipartite heterogeneous banking transaction networks. Method: We propose an inductive GraphSAGE-based node representation learning framework designed for efficient and scalable embedding on anonymized customer-merchant transaction graphs, naturally incorporating heterogeneous attributes such as geographic and demographic features. Contribution/Results: The method exhibits strong inductive capability—directly generalizing to newly arriving nodes—and yields interpretable, semantically meaningful node clusters. Evaluated on real-world banking data, the learned embeddings significantly improve ranking performance for high-risk accounts in money mule detection, boosting NDCG@10 by 12.7%. This demonstrates the framework’s effectiveness and practical utility for downstream anti-fraud applications.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Financial institutions increasingly require scalable tools to analyse complex transactional networks, yet traditional graph embedding methods struggle with dynamic, real-world banking data. This paper demonstrates the practical application of GraphSAGE, an inductive Graph Neural Network framework, to non-bipartite heterogeneous transaction networks within a banking context. Unlike transductive approaches, GraphSAGE scales well to large networks and can generalise to unseen nodes which is critical for institutions working with temporally evolving transactional data. We construct a transaction network using anonymised customer and merchant transactions and train a GraphSAGE model to generate node embeddings. Our exploratory work on the embeddings reveals interpretable clusters aligned with geographic and demographic attributes. Additionally, we illustrate their utility in downstream classification tasks by applying them to a money mule detection model where using these embeddings improves the prioritisation of high-risk accounts. Beyond fraud detection, our work highlights the adaptability of this framework to banking-scale networks, emphasising its inductive capability, scalability, and interpretability. This study provides a blueprint for financial organisations to harness graph machine learning for actionable insights in transactional ecosystems.

Problem

• Scalable graph embedding for dynamic banking transaction networks
• Inductive learning for unseen nodes in evolving financial data
• Interpretable node embeddings for fraud detection and risk analysis

Innovation

• GraphSAGE for non-bipartite transaction networks
• Inductive framework generalizing to unseen nodes
• Embeddings enable clustering and fraud detection