Adaptive Threshold Tuning for Real-Time Fraud Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Fraud detection in financial transactions has become increasingly difficult due to sophisticated fraud techniques, such as imposter fraud, where fraudsters convince customers to transfer money, making it challenging for financial institutions to detect fraudulent activities based on browser information or technological analysis.

Innovation Solution

A system that systematically models transaction attributes using a simulation manager, determines ceiling thresholds for transaction characteristics, and provides real-time detection of anomalous transactions through a detection manager, allowing for adaptive threshold tuning to minimize false positives and detect fraudulent transactions effectively.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional fraud detection methods based on browser information or technological analysis are used, then the detection process is simple, but the detection accuracy deteriorates due to sophisticated fraud techniques like imposter fraud

Engineering Contradiction:

Improvefraud detection accuracyVSAvoiddetection system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system changes the parameters being monitored from traditional browser information to transaction-level behavioral parameters. It systematically models multiple attributes of transactions (e.g., transaction amount, frequency, recipient patterns) and dynamically adjusts thresholds based on learned normal behavior patterns, enabling detection of imposter fraud where the customer's own device and browser are being used

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces traditional rule-based mechanical detection systems with an adaptive learning system that automatically models transaction attributes and determines ceiling thresholds through systematic analysis of historical transaction data, eliminating the need for manual rule configuration and improving detection of sophisticated fraud

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If static threshold values are used for transaction characteristics, then the system is simple to operate, but the detection effectiveness deteriorates due to varying normal transaction patterns

Engineering Contradiction:

Improvedetection reliabilityVSAvoidthreshold adaptability

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system implements dynamic thresholds that automatically adjust based on learned transaction patterns. The limitation manager continuously analyzes historical transactions to determine ceiling thresholds for each transaction characteristic, allowing the detection system to adapt to changing normal behavior patterns while maintaining high reliability in detecting true anomalies

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The detection system performs self-adjustment by automatically modeling transaction attributes and determining appropriate thresholds without manual intervention. The systematic modeling of transaction characteristics and automatic threshold determination enables the system to serve itself in adapting to new fraud patterns and legitimate behavior changes

Inventive Principle:

Principle #25Self-service

3Measurement precision

If comprehensive transaction analysis is performed to improve fraud detection, then the detection accuracy improves, but the processing time increases

Engineering Contradiction:

Improvetransaction analysis accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary modeling of transaction attributes on historical transaction data before real-time detection. By pre-establishing baseline patterns and ceiling thresholds for each transaction characteristic, the system prepares detection criteria in advance, enabling fast real-time comparison and decision-making without extensive processing of each new transaction

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent segments the transaction analysis into distinct components: systematic modeling of individual transaction attributes, separate determination of ceiling thresholds for each characteristic, and modular detection rules. This segmentation allows parallel processing and efficient real-time evaluation of multiple transaction dimensions independently

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11216816B1Identification of anomalous transaction attributes in real-time with adaptive threshold tuning

Publication Date: 2022.01.04 WELLS FARGO BANK NA

AI summary

Identification of anomalous transaction attributes in real-time with adaptive threshold tuning is provided. A set of historical transactions conducted during a defined time period are analyzed and categorizing into defined groups. Outlier transactions are identified and removed from the set of historical transactions and a set of non-anomalous transactions are determined. When a new transaction is received, the new transaction is automatically allowed based on a determination that the subsequent transaction conforms to the set of non-anomalous transactions. Alternatively, an alert for further analysis for the new transaction is output based on a determination that the subsequent transaction does not conform to the set of non-anomalous transactions.