Biometric Fraud Detection Using Profile-Specific Correction Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional voice biometric fraud detection systems face inefficiencies due to the reliance on global score threshold settings, which can lead to either false positives or false misses, particularly with outlier profiles, resulting in increased costs and potential financial and reputational damage.
Innovation Solution
Implementing a method to define detection policies for each biometric profile based on individual detection performance metrics, such as a target false accept rate, and adjusting biometric scores using correction factors to adapt to unique profile behaviors, thereby reducing the impact of outliers and improving detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If global score threshold settings are used for biometric fraud detection, then the system is simple to operate, but it produces false positives and false misses particularly with outlier profiles
Solution Approach 1:
The patent segments the fraud detection system by creating individual detection policies for each biometric profile rather than using a single global threshold. Each profile receives customized scoring based on its specific characteristics, which eliminates false positives and false misses caused by outlier profiles while maintaining systematic operation through automated policy generation.
Solution Approach 2:
The patent applies local quality by tailoring the detection policy to each specific biometric profile. Each profile receives customized scoring weights and thresholds based on its unique characteristics, allowing the system to adapt to local variations in biometric data quality and characteristics rather than applying uniform global standards.
2Measurement precision
If individual detection policies are defined for each biometric profile, then detection accuracy improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent performs preliminary action by pre-defining detection policies for each biometric profile before actual fraud detection occurs. During enrollment or setup, the system analyzes biometric characteristics and establishes customized scoring policies in advance, so that during operational detection, only simple score comparison is needed rather than complex real-time analysis.
Solution Approach 2:
The system performs self-service by automatically generating detection policies based on biometric profile characteristics without requiring manual intervention. The automated policy generation process analyzes each profile's features and configures optimal scoring parameters independently, reducing both computational burden and processing time.
3Reliability
If correction factors are applied to adjust biometric scores, then false alerts are reduced, but the system complexity increases
Solution Approach 1:
The patent applies parameter changes by introducing correction factors that modify biometric scores based on profile-specific characteristics. These correction factors adjust the scoring parameters dynamically for each profile, compensating for variations in biometric data quality, characteristics, and reliability without requiring complete redesign of the detection system.
Data Source
AI summary
A method, computer program product, and computing system for defining a correction factor for a biometric profile of a plurality of biometric profiles based upon, at least in part, a detection performance metric associated with the biometric profile. The biometric profile may be adjusted based upon, at least in part, the detection policy for the biometric profile.