Tiered Aggregation for Network Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As the number of locations for network interaction fraud detection increases, the processing location faces a bottleneck in handling the aggregated information, impeding the scalability of detection systems.
Innovation Solution
Implementing a tiered aggregation system where network interaction information is processed and aggregated across multiple locations, including local, organizational, segment, country, and centralized aggregators, with rules-based determinations for sharing metadata and using models to assess confidence levels for fraud detection, allowing for efficient data transfer and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If information from multiple locations is aggregated at a single processing location, then detection accuracy is improved, but processing bottleneck increases
Solution Approach 1:
The patent segments the centralized processing function into distributed processing units located at multiple geographic locations. Each processing unit independently processes information from local sensors and makes local detection decisions, eliminating the single processing bottleneck while maintaining aggregated detection capability through distributed collaboration.
Solution Approach 2:
The patent introduces a hierarchical dimension to the processing architecture, organizing processing units across multiple levels (local, regional, national). This multi-level hierarchy allows information to be processed at appropriate granularities, distributing the processing load across vertical levels while maintaining comprehensive detection coverage.
2Adaptability or versatility
If the number of monitoring locations increases, then fraud detection coverage is improved, but system complexity increases
Solution Approach 1:
The patent implements universal processing units that can operate independently or in coordination with other units. Each processing unit is designed to perform the same core detection functions locally, allowing the system to scale by simply adding more identical units rather than designing increasingly complex centralized processors.
Solution Approach 2:
The patent employs a nested hierarchical structure where local processing units are contained within regional aggregates, which are in turn contained within national systems. This nesting allows complex national-level detection to be built from simpler regional and local components, reducing overall system complexity through modular organization.
Data Source
AI summary
A method for analyzing network interaction data is disclosed. Initially, network interaction data is received from a network over time. A predetermined model comprising predetermined values associated with network interaction parameters is also received. The received network interaction data is processed to determine the network interaction parameters and information regarding the network interaction data. A score for the network interaction data is calculated based on the predetermined model and the determined network interaction parameters. The score is compared to a threshold. The information regarding the network interaction data is then forwarded based on the comparison of the score to the threshold.