🤖 AI Summary
This work addresses the dynamic modeling of financial transaction networks. We systematically introduce temporal motifs—the first such application to multi-source real-world financial networks (Mercari, JPMorgan Chase synthetic payment data, and Venmo)—to uncover the critical role of coupling between temporal sequentiality and graph structure in fraud detection and socio-financial relationship modeling. We propose a directed, time-respecting subgraph counting framework coupled with motif-driven feature engineering, integrated with supervised learning and link-prediction-inspired heuristics. Experiments demonstrate: (1) substantial gains over static graph-based fraud detection methods; (2) significantly improved AUC for friend prediction; (3) high accuracy in Venmo merchant identification; and (4) discovery of novel anomalous patterns—e.g., “rare temporal cycles”—validating the cross-platform generalizability and effectiveness of temporal motifs for financial behavior analysis.
📝 Abstract
Understanding the dynamics of financial transactions among people is critically important for various applications such as fraud detection. One important aspect of financial transaction networks is temporality. The order and repetition of transactions can offer new insights when considered within the graph structure. Temporal motifs, defined as a set of nodes that interact with each other in a short time period, are a promising tool in this context. In this work, we study three unique temporal financial networks: transactions in Mercari, an online marketplace, payments in a synthetic network generated by J.P. Morgan Chase, and payments and friendships among Venmo users. We consider the fraud detection problem on the Mercari and J.P. Morgan Chase networks, for which the ground truth is available. We show that temporal motifs offer superior performance than a previous method that considers simple graph features. For the Venmo network, we investigate the interplay between financial and social relations on three tasks: friendship prediction, vendor identification, and analysis of temporal cycles. For friendship prediction, temporal motifs yield better results than general heuristics, such as Jaccard and Adamic-Adar measures. We are also able to identify vendors with high accuracy and observe interesting patterns in rare motifs, like temporal cycles. We believe that the analysis, datasets, and lessons from this work will be beneficial for future research on financial transaction networks.
