Fused Risk Score Generation for Real-Time Call Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fraud detection systems in call centers are ineffective in adapting to changing fraud methods, as they rely on outdated risk scoring methods that fail to accurately determine the likelihood of fraud in real-time call data.
Innovation Solution
A system that generates fused risk scores by combining uniquely calculated fraud risk scores from multiple sub-modules, using voice and non-audio data comparisons with fraudster databases, to determine the likelihood of fraud and classify calls as legitimate or fraudulent.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional risk scoring methods are used, then system simplicity is maintained, but fraud detection accuracy deteriorates
Solution Approach 1:
The system divides fraud detection into multiple independent sub-modules, each specializing in specific fraud patterns or data types. Each sub-module generates its own risk score, and these scores are then fused to create a comprehensive assessment. This segmentation allows each module to be optimized for its specific function while maintaining overall system manageability.
Solution Approach 2:
The system combines risk scores from multiple sub-modules through a fusion process that integrates diverse fraud detection signals. By merging the outputs of specialized sub-modules, the system achieves comprehensive fraud detection accuracy that exceeds what any single module could provide alone.
2Measurement precision
If multiple sub-modules are used to generate unique risk scores, then fraud detection accuracy is improved, but processing time increases
Solution Approach 1:
The system performs preliminary fraud assessment using multiple sub-modules in parallel during the call interaction itself, rather than sequentially after the call ends. This allows risk scores to be generated and fused in real-time, enabling immediate fraud detection decisions without significant processing delays.
Solution Approach 2:
The system replaces traditional sequential mechanical processing with a parallel computational architecture where multiple sub-modules operate simultaneously. This substitution of processing methodology dramatically reduces the time required to aggregate and analyze multiple risk factors.
3Productivity
If real-time fraud detection is implemented, then fraud identification capability is improved, but system resource consumption increases
Solution Approach 1:
The system applies partial processing by having different sub-modules activate based on the specific call context and detected fraud indicators. Not all sub-modules process every call at full capacity; instead, the system dynamically adjusts the level of analysis based on risk indicators, reducing unnecessary resource consumption while maintaining high fraud detection capability.
Data Source
AI summary
Systems, methods, and media for generating fused risk scores for determining fraud in call data are provided herein. Some exemplary methods include generating a fused risk score used to determine fraud from call data by generating a fused risk score for a leg of call data, via a fuser module of an analysis system, the fused risk score being generated by fusing together two or more uniquely calculated fraud risk scores, each of the uniquely calculated fraud risk scores being generated by a sub-module of the analysis system; and storing the fused risk score in a storage device that is communicatively couplable with the fuser module.