User Defined Security Parameters for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current neural network-based fraud-prevention systems are ineffective during the learning phase, may not accurately reflect actual user behavior, and are slow to adapt to changing patterns, leading to vulnerabilities in electronic transaction fraud prevention.
Innovation Solution
A network-based security system that allows users to set personalized security parameters, such as geographic location, monetary value ranges, and transaction modes, which are stored in a user security parameter database and used to screen transactions, with the option to warn or block transactions that fall outside these parameters, and incorporates a neural network for additional analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a neural network is used to learn legitimate behavior patterns, then fraud detection capability is improved, but the system is ineffective during the learning phase and slow to adapt to changing patterns
Solution Approach 1:
The system performs preliminary actions by collecting transaction data and establishing baseline behavior patterns before actively using the neural network for fraud detection. This preliminary data collection and pattern establishment phase prepares the system in advance, reducing the ineffective learning period when the system is deployed.
Solution Approach 2:
The neural network is designed to be dynamic and continuously adapt to changing transaction patterns. The system updates its learned patterns over time as new data becomes available, allowing it to quickly adapt to evolving legitimate and fraudulent behavior patterns rather than remaining static.
2Measurement precision
If a neural network learns behavior patterns, then detection accuracy is improved, but the learned pattern may not fully reflect the customer's or business' actual pattern of behavior
Solution Approach 1:
The system incorporates feedback mechanisms where transaction outcomes and user confirmations are fed back into the neural network. This feedback loop allows the system to continuously refine its learned patterns based on actual customer behavior, improving the accuracy of behavior pattern representation over time by correcting deviations from true user intent.
3Reliability
If security parameters are set to block suspicious transactions, then fraud prevention is improved, but legitimate transactions may be blocked
Solution Approach 1:
The system applies partial blocking actions rather than complete blocking for suspicious transactions. Instead of immediately blocking all potentially fraudulent transactions, the system can flag them for review, apply partial restrictions, or require additional verification, thereby preventing fraud while allowing most legitimate transactions to proceed without interruption.
Data Source
AI summary
A fraud-prevention system having user security parameters based on user instruction. A user specifies the security parameter for a transaction. The security parameter may include, but is not limited to a geographic location, a monetary value range, a transaction mode, an account access parameter, a class of goods, or a class of services. The fraud-prevention system acquires this security parameter information from the user and stores it in a user security parameter database. The fraud-prevention system uses these user security parameters to screen subsequent transactions.