Issuer Authorization Vulnerability Testing With Synthetic Fraud Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for determining vulnerabilities in issuer authorization networks require manual setup and testing, lack intelligent test cases, and face challenges with high cardinality categorical columns, leading to privacy concerns and inefficiencies in data processing.
Innovation Solution
A computer-implemented method using a memory-efficient conditional generative adversarial network (MeTGAN) to generate synthetic transaction data, enrich it with historical card velocity features, and simulate an issuer authorization model to assess vulnerabilities automatically.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual testing methods are used to evaluate vulnerabilities in issuer authorization networks, then detailed vulnerability reports can be generated, but the process requires significant manual effort and network downtime
Solution Approach 1:
The patent creates a simulated authorization model that replicates the behavior of the actual issuer authorization network. This copy allows vulnerability testing to be performed on the simulation rather than the live system, eliminating network downtime while maintaining detection accuracy. The simulated model mirrors the authorization logic and can be independently tested without affecting production operations.
Solution Approach 2:
The system performs vulnerability assessment in advance by generating synthetic transaction data and testing against the simulated authorization model before deploying changes to the live system. This preliminary testing identifies potential vulnerabilities ahead of time, allowing the actual authorization network to remain operational without interruption.
2Reliability
If synthetic transaction data is generated using existing GAN models, then data privacy concerns are addressed and training data availability is improved, but memory footprint increases significantly
Solution Approach 1:
The patent extracts and utilizes only the essential characteristics and patterns from real fraudulent transaction data to generate synthetic data. By focusing on key features rather than complete data replication, the system maintains privacy protection while reducing the memory resources required for data storage and processing.
Solution Approach 2:
The system transforms real transaction data into synthetic data by changing specific parameters and characteristics while preserving the underlying fraud patterns. This parameter transformation approach maintains the reliability of privacy protection through synthetic data generation while optimizing memory usage by not storing both original and complete synthetic datasets simultaneously.
3Measurement precision
If comprehensive transaction data is processed to create test cases, then vulnerability detection coverage is improved, but data processing complexity and time requirements increase
Solution Approach 1:
The patent segments the vulnerability assessment process into distinct components: synthetic data generation, enrichment with historical features, feasibility classification, and simulated model testing. This segmentation allows each component to be optimized independently and processed in manageable stages, reducing overall complexity while maintaining comprehensive coverage through systematic progression through each segment.
Data Source
AI summary
Embodiments provide artificial intelligence methods and systems for evaluating vulnerability risks of issuer authorization system. Method performed by a server system includes accessing a set of payment transaction data including subset of fraudulent transaction data. Method includes generating via a machine learning model, set of synthetic transaction data based on the subset of fraudulent transaction data. Method includes accessing set of historical card velocity features and collating the set of synthetic transaction data and the set of historical card velocity features to generate set of enriched synthetic transaction data. Method includes extracting via a classifier, subset of feasible fraudulent transaction data from the set of enriched synthetic transaction data. Method includes generating simulated authorization model based on the set of payment transaction data. Method includes classifying via simulated authorization model, each enriched synthetic transaction from the subset of feasible fraudulent transaction data as one of fraudulent transaction and non-fraudulent transaction.