Network Message Set Identification for Real-Time Fraud Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Banking networks face challenges in efficiently identifying sets of messages related to fraudulent activity due to their complexity and high transaction volumes, leading to significant delays in detecting illicit fund dispersion and making it difficult to stop fraudsters and retrieve stolen funds.

Innovation Solution

An apparatus, method, and computer program product that receive information about sets of messages exchanged in a network, select a portion of these messages based on properties of their source messages, generate individual values for the nodes, and identify sets of messages of interest using these values, thereby increasing processing efficiency and reducing computational effort.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If complete network scanning is performed to identify fraudulent transactions, then detection accuracy is improved, but processing time and computational resources increase significantly

Engineering Contradiction:
Improvefraud detection accuracyVSAvoiddetection delay
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the large-scale network scanning task into smaller, manageable units by dividing the banking network into multiple sub-networks or zones. Each scanner is assigned to scan a specific segment, generating partial sets of messages that are then aggregated. This segmentation reduces the computational burden on individual scanners and enables parallel processing across multiple devices, thereby maintaining detection accuracy while reducing overall processing time and resource consumption.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements preliminary actions by pre-processing transaction data before full analysis. Scanners first perform initial filtering and validation on incoming messages to identify potentially fraudulent patterns before conducting comprehensive analysis. This preliminary action reduces the volume of data requiring full processing, enabling faster detection without compromising accuracy by focusing computational resources on high-risk transactions.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If comprehensive message analysis is performed across all nodes, then fraud detection reliability is improved, but device complexity and computational effort increase

Engineering Contradiction:
Improvefraud detection reliabilityVSAvoidprocessing system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent divides the complex analysis task into segments performed by different components: scanners at individual nodes perform local message analysis, a central coordinator aggregates partial results, and a fraud detection engine performs final validation. This segmentation of analytical functions maintains comprehensive coverage and reliability while distributing computational complexity across multiple simpler components rather than requiring a single complex system.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediary components that facilitate comprehensive analysis without requiring direct complex interactions between all system elements. A central coordinator acts as an intermediary that collects partial message sets from multiple scanners, performs aggregation and deduplication, and forwards consolidated data to the fraud detection engine. This intermediary layer simplifies the overall system architecture while maintaining comprehensive analysis capability.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If all sets of messages are processed, then complete coverage of fraudulent activity is achieved, but processing efficiency decreases

Engineering Contradiction:
Improvecoverage completenessVSAvoidmessage processing efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent extracts and processes only the essential elements of message sets required for fraud detection. Rather than processing complete message sets in their entirety, the system extracts key features such as transaction amounts, node identifiers, timestamps, and routing information. This extraction approach maintains complete coverage of fraudulent activity patterns while significantly improving processing efficiency by focusing computational resources on critical detection parameters.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent implements partial action by processing a representative subset of message sets in detail while using sampling and aggregation techniques to infer characteristics of the complete set. Scanners process messages in batches, analyzing a partial set thoroughly and using statistical methods to generalize findings across the complete message population. This approach achieves sufficient coverage for reliable fraud detection while maintaining high processing efficiency.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4268167B1Apparatus, method and computer program product for identifying a set of messages of interest in a network
Publication Date: 2026.02.18 VOCALINK INT LTD
  • EP4268167B1 patent drawingFigure 1
  • EP4268167B1 patent drawingFigure 2
  • EP4268167B1 patent drawingFigure 3

AI summary

An apparatus, method and computer program product are provided, the apparatus for identifying a set of messages of interest in a network, the apparatus comprising circuitry configured to: receive information of a plurality of sets of messages which have been exchanged by nodes in a network; select a portion of the plurality of sets of messages which have been received using a property of a source message of each set of messages of the plurality of sets of messages; generate individual first values for the nodes of each set of messages of the selected portion of the plurality of sets of messages using a property of the nodes associated with each set of messages; and identify at least one set of messages of the selected portion of the plurality of sets of messages as a set of messages of interest using the individual first values which have been generated for the selected portion of the sets of messages.