Distributed Fraud Scoring System for Payment Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing payment processing networks face interruptions in determining fraudulent electronic transactions, especially during maintenance, as fraud scoring systems become inoperable, leading to inefficiencies and additional processing needs to correct fraudulent transactions.
Innovation Solution
A distributed fraud risk scoring system using two scoring computing devices, one for odd and one for even bank identification number transactions, with real-time data replication and fallback processes to ensure continuous operation during maintenance, allowing for robust and failure-tolerant fraud scoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single centralized fraud scoring system is used, then the system structure is simple and easy to manage, but the system becomes inoperable during maintenance causing interruptions in fraud determination
Solution Approach 1:
The fraud scoring system is segmented into multiple independent scoring devices (first scoring device and second scoring device). Each device can independently process fraud scoring for transactions, allowing the system to maintain operability during maintenance by switching between devices. This segmentation resolves the contradiction by distributing functionality across multiple units rather than relying on a single centralized system.
Solution Approach 2:
The system implements beforehand cushioning through predefined failover mechanisms and redundant scoring devices. When maintenance is performed on one scoring device, the system is already prepared to switch to the other device, cushioning against service interruptions. This preparatory redundancy ensures continuous fraud scoring capability while maintaining relatively simple operational procedures.
2Reliability
If multiple scoring devices are deployed for redundancy, then system availability during maintenance is improved, but the system complexity and data synchronization requirements increase
Solution Approach 1:
The patent applies local quality by assigning specific characteristics to different scoring devices based on their bank identification number ranges. The first scoring device handles transactions with odd-numbered BINs while the second scoring device handles even-numbered BINs. This local differentiation allows each device to operate independently with specialized data, reducing the complexity of full system-wide synchronization while maintaining redundancy and availability.
Solution Approach 2:
The transaction processing is segmented by bank identification number parity, with odd BIN transactions routed to one scoring device and even BIN transactions to another. This segmentation reduces inter-device data dependency and synchronization complexity compared to a fully redundant system where all devices must maintain identical data states for all transactions.
3Loss of information
If fraud scoring is performed centrally without redundancy, then data consistency is maintained, but additional transactions are needed to correct fraudulent transactions during system interruptions
Solution Approach 1:
The redundant scoring device architecture enables continuous fraud scoring operations without interruption during maintenance. While one scoring device is being maintained, the other continues to process transactions, eliminating the need to stop fraud scoring or require additional corrective transactions. This continuity maintains both data consistency and processing efficiency simultaneously.
Data Source
AI summary
A co-processing fraud risk scoring system for scoring electronic payment transactions for potential fraud is described. Additionally, a method and a computer-readable storage medium for scoring electronic payment transactions for potential fraud are described.