Anomaly Profiling with Weighted Clustering for Fraud Patterns

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems for automatically flagging anomalies in transaction processing networks require manual efforts to profile and recommend strategies for anomalies, such as fraudulent transactions, without efficiently determining the type or nature of the anomaly.

Innovation Solution

A method and system that uses unsupervised clustering algorithms to segment and label anomaly transactions based on feature profiles, generating weights for features to automatically profile anomalies, including fraud detection, using distribution-based scoring to identify stronger indicators for specific anomalies.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual efforts are used to profile anomalies, then accuracy in understanding anomaly types can be maintained, but processing efficiency and speed are reduced

Engineering Contradiction:

Improveanomaly profiling efficiencyVSAvoidanomaly type identification accuracy

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The system enables anomaly transactions to profile themselves automatically through unsupervised clustering algorithms. The algorithm autonomously segments anomalies into distinct types based on their feature patterns without requiring manual intervention, while the system self-generates feature profiles and weights to identify anomaly characteristics.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical process of anomaly profiling with an automated computational system. The unsupervised clustering algorithm and distribution-based scoring mechanism substitute human analysts, automatically segmenting anomalies, generating feature profiles, and identifying anomaly types through mathematical computations rather than manual examination.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If all anomaly transactions are analyzed in detail, then comprehensive profiling is achieved, but processing time and computational resources increase

Engineering Contradiction:

Improvefeature profile accuracyVSAvoidprofiling processing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system segments the large set of anomaly transactions into smaller, manageable clusters based on their feature similarities. By dividing the data into distinct segments or types, the system can process each segment with appropriate feature profiles, achieving comprehensive analysis without analyzing every single transaction in detail simultaneously.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system generates feature profiles with weighted features, focusing on the most significant characteristics rather than analyzing all features equally. The distribution-based scoring identifies and emphasizes key features that dominate anomaly patterns, allowing accurate profiling with reduced computational effort on less critical features.

Inventive Principle:

Principle #16Partial or excessive action

3Productivity

If automated clustering algorithms are used to segment anomalies, then processing efficiency is improved, but system complexity increases

Engineering Contradiction:

Improveautomatic anomaly segmentation speedVSAvoidclustering algorithm complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The unsupervised clustering algorithm serves multiple functions simultaneously: it segments anomalies into types, generates feature profiles for each segment, identifies dominant features through distribution-based scoring, and enables future classification of new anomalies. This multi-functionality reduces the need for separate systems for each task, managing complexity while maintaining high productivity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

4Measurement precision

If feature weights are generated based on distribution scoring, then identification of key anomaly indicators is improved, but computational overhead increases

Engineering Contradiction:

Improvekey feature identification accuracyVSAvoidcomputational energy consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system changes the parameter of feature importance by generating dynamic weights based on distribution scoring. Instead of treating all features equally or using fixed importance values, the system computes weights that reflect the actual distribution patterns in the data, identifying which features most strongly distinguish different anomaly types based on their statistical properties.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250390875A1Method, System, and Computer Program Product for Auto-Profiling Anomalies

Publication Date: 2025.12.25 VISA INTERNATIONAL SERVICE ASSOCIATION

AI summary

Methods, systems, and computer program products for auto-profiling anomalies that: receive anomaly transactions, select a subset of anomaly transactions, the subset of anomaly transactions being associated with a plurality of features, generate, based on the plurality of features and a distribution of the plurality of features, a plurality of weights associated with the plurality of features; segment, using an unsupervised clustering algorithm, based on the plurality of features and the plurality of weights, the subset of anomaly transactions into a plurality of segments of anomaly transactions; and label a subset of segments of the plurality of segments with a feature profile including a feature from each segment of the subset of segments associated with a highest weight of the plurality of weights of the plurality of features of the anomaly transactions in that segment.