Fraud Detection System Using Historical Transaction Clustering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing payment card fraud detection systems have not kept pace with evolving fraudulent activities and increased computing capabilities, failing to effectively utilize historical transaction data and advanced statistical methods for real-time and batch processing.

Innovation Solution

A computerized system and method that receives data from transacting entities, applies it to models for generating fraud scores, and identifies fraudulent transactions by using historical transaction data, clustering analysis, and external data sources, enabling real-time and batch mode fraud detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If existing fraud detection systems process transactions using traditional methods, then system simplicity is maintained, but detection accuracy fails to keep pace with evolving fraud patterns

Engineering Contradiction:

Improvefraud detection accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The fraud detection system is segmented into multiple independent components: transaction data receiver, historical data repository, statistical model applicator, score generator, and fraud indicator producer. Each component performs a specific function, allowing the system to achieve high detection accuracy through specialized processing while maintaining manageable complexity through modular architecture.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system dynamically adapts to evolving fraud patterns by continuously applying statistical models to both current transaction data and historical transaction data. The fraud detection process is dynamic rather than static, allowing the system to keep pace with changing fraudulent activities while maintaining a structured processing framework.

Inventive Principle:

Principle #15Dynamics

2Reliability

If historical transaction data is incorporated into fraud detection, then detection effectiveness improves, but data processing time increases

Engineering Contradiction:

Improvefraud detection effectivenessVSAvoiddata processing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

Historical transaction data is pre-collected and stored in a dedicated repository before fraud detection is needed. This preliminary preparation of data allows the system to quickly retrieve and analyze historical patterns during actual fraud detection operations, improving effectiveness without adding processing delays during critical transaction evaluation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates a copy of historical transaction data that can be independently analyzed without affecting the processing of current transactions. By working with replicated historical data sets, the system can perform comprehensive pattern matching and statistical analysis while the original transaction processing continues uninterrupted.

Inventive Principle:

Principle #26Copying

3Measurement precision

If advanced statistical models are applied to transaction data, then fraud detection precision improves, but computational requirements increase

Engineering Contradiction:

Improvefraud score precisionVSAvoidcomputational energy consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system applies statistical models selectively to the most relevant features of transaction data rather than processing all possible data elements. By focusing computational resources on key indicators and patterns that most strongly correlate with fraud, the system achieves high detection precision while minimizing unnecessary computational energy consumption.

Inventive Principle:

Principle #16Partial or excessive action

4Speed

If real-time fraud detection is implemented, then transaction processing speed is maintained, but system complexity increases

Engineering Contradiction:

Improvetransaction processing speedVSAvoidreal-time processing complexity

Core Design Contradiction:

SpeedVSDevice complexity

Solution Approach 1:

The real-time fraud detection process is segmented into discrete, quickly-executable steps: receive transaction data, retrieve relevant historical data, apply statistical models, generate fraud scores, and produce fraud indicators. This segmentation allows each step to be optimized for speed while the overall system maintains real-time processing capability despite the added complexity of multiple processing stages.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS7668769B2System and method of detecting fraud

Publication Date: 2010.02.23 CORELOGIC SOLUTIONS LLC

AI summary

Embodiments include systems and methods of detecting fraud. In particular, one embodiment includes a system and method of detecting fraud in transaction data such as payment card transaction data. For example, one embodiment includes a computerized method of detecting that comprises receiving data associated with a financial transaction and at least one transacting entity, wherein the data associated with the transacting entity comprises at least a portion of each of a plurality of historical transactions of the transacting entity, applying the data to at least one first model, generating a score based on the first model, and generating data indicative of fraud based at least partly on the score. Other embodiments include systems and methods of generating models for use in fraud detection systems.