Synthetic Payment Graph Generation for Real-Time Fraud Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The nascent nature of real-time payment transactions lacks publicly available datasets for machine learning techniques, hindering risk management and decision-making capabilities.
Innovation Solution
A system and method for generating synthetic graphs that simulate real-time payment transactions by creating a base payment graph with nodes representing entities and edges representing transaction probabilities, inserting adversarial activity patterns, and performing machine learning techniques on dynamic payment graphs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If real-time payment transactions are implemented, then transaction speed and immediacy are improved, but the lack of historical datasets worsens the ability to develop machine learning techniques for risk management
Solution Approach 1:
The system generates synthetic transaction data and adversarial patterns in advance, before actual machine learning deployment. This preliminary creation of datasets enables risk management models to be trained and validated without waiting for sufficient real historical data to accumulate, directly addressing the data scarcity problem while maintaining fast transaction processing
Solution Approach 2:
The system creates synthetic copies of real payment graphs and transaction patterns through graph generation algorithms. These synthetic copies replicate the structural and behavioral characteristics of real payment networks, providing sufficient training data for machine learning models without requiring actual historical transaction records
2Reliability
If synthetic payment graphs are generated to provide training data, then machine learning capability is improved, but the complexity of the system increases
Solution Approach 1:
The graph generation system serves multiple functions: it creates synthetic payment graphs, generates adversarial patterns, validates machine learning models, and provides training data. This multi-functionality reduces the need for separate systems for each task, managing overall system complexity while improving risk management capabilities
Solution Approach 2:
The synthetic payment graph system acts as an intermediary between real payment networks and machine learning models. It translates complex real-world payment patterns into structured synthetic data that can be efficiently processed by ML algorithms, simplifying the interface between data generation and model training
Data Source
AI summary
Provided is a computer-implemented method for generating synthetic graphs that simulate real-time payment transactions that includes generating a base payment graph includes a plurality of nodes and a plurality of edges connecting the plurality of nodes, wherein each node represents an entity and each edge represents a probability that a real-time-payment transaction may be conducted involving two entities that are connected by the edge, wherein the real-payment transaction is artificially created, generating a plurality of dynamic payment graphs based on the base payment graph, inserting patterns representing adversarial activity into the plurality of dynamic payment graphs, and performing an action associated with a machine learning technique using the plurality of dynamic payment graphs. Systems and computer program products are also provided.