Testing Transaction Score for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fraud detection systems are ineffective in identifying low-frequency, low-value testing transactions, which are often conducted by fraudulent entities to validate stolen payment account information, leading to higher likelihoods of fraud as these transactions are inconspicuous and differ from typical high-frequency or high-value suspicious transactions.
Innovation Solution
A method and system that utilize a testing transaction score generated by a testing model to classify transactions as testing or non-testing transactions, allowing for the detection and prevention of fraud by recognizing patterns of testing transactions followed by high-value fraudulent activities, and enabling the use of this score to decline subsequent transactions and update the model based on fraud reports.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fraud detection systems focus on high-frequency or high-value transactions, then they can detect typical suspicious transactions, but they fail to detect low-frequency, low-value testing transactions which are used to validate stolen payment account information
Solution Approach 1:
The system changes the detection parameters from focusing on high-frequency or high-value transactions to identifying low-frequency, low-value testing transactions. The testing transaction score model uses specific parameters such as transaction amount, frequency, and temporal patterns to distinguish testing transactions from normal transactions, thereby improving detection accuracy for this specific fraud type while maintaining adaptability to other fraud patterns.
Solution Approach 2:
The system performs preliminary detection and scoring of transactions using the testing transaction score model before they can lead to higher-value fraud. By identifying and flagging testing transactions early in the transaction sequence, the system can take preventive actions such as blocking subsequent transactions or alerting authorities, thereby preventing the progression to more serious fraud.
2Reliability
If the system uses a testing transaction score to identify testing transactions, then it can prevent subsequent high-value fraudulent transactions, but it requires processing and analyzing transaction data in real-time which increases system complexity
Solution Approach 1:
The fraud detection system is segmented into distinct functional components: the testing transaction score model that analyzes transaction patterns, the scoring mechanism that generates risk scores, and the decision-making component that determines whether to block transactions. This segmentation allows each component to be optimized independently and simplifies the overall system architecture while maintaining high reliability in fraud prevention.
Solution Approach 2:
The testing transaction score acts as an intermediary between raw transaction data and fraud prevention decisions. Instead of directly analyzing complex transaction patterns and making decisions, the system uses the score as a mediator that summarizes the risk level, thereby simplifying the decision-making process while maintaining accurate fraud detection and prevention capabilities.
3Reliability
If the system declines transactions based on testing transaction scores, then it can prevent fraud, but it may also block legitimate low-value transactions resulting in false positives
Solution Approach 1:
The system applies partial action by not automatically blocking all transactions with elevated testing transaction scores. Instead, it uses the scores to prioritize monitoring and investigation, blocking only those transactions that exceed specific threshold values or exhibit multiple suspicious characteristics. This approach maintains high fraud detection accuracy while reducing false positives by being selective about which scored transactions are actually blocked.
Solution Approach 2:
The system incorporates feedback mechanisms where the outcomes of transaction decisions (both blocked and approved transactions) are fed back into the testing transaction score model. This feedback allows the model to learn from actual fraud cases and false positives, continuously refining its scoring accuracy and adjusting thresholds to minimize false positives while maintaining reliable fraud detection.
Data Source
AI summary
Embodiments of the invention relate in part to determining a testing model and providing a testing transaction score for transactions. The testing transaction score may indicate a likelihood that the transaction is a testing transaction. One embodiment of the invention discloses a method comprising receiving a first authorization message for a first transaction using an account, determining a testing transaction score for the first transaction using a testing model, and determining that the first transaction is a testing transaction based on the testing transaction score, wherein the testing transaction score for the first transaction is used for declining a second transaction using the same account conducted after the first transaction.