Anonymous Identifier Fraud Detection via Temporal Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in distinguishing trusted customers from cybercriminals due to sophisticated online fraud schemes, particularly in identifying fraudulent behavior in time to prevent cybercrimes, as they often rely on high thresholds that overlook collusion from groups staying below these thresholds.
Innovation Solution
The system uses unique, anonymous alphanumeric identifiers to associate entity attributes and activities with temporal information, determining linkage and anomaly scores to calculate an aggregate threat score, thereby granting or blocking electronic access based on this score, effectively preventing fraudulent account access or transactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If high thresholds are used to identify fraudulent behavior, then false positives are reduced, but organized groups staying below thresholds can collude undetected
Solution Approach 1:
The system transforms the static threshold parameter into a dynamic scoring mechanism. Instead of using fixed high thresholds that miss subtle collusion patterns, the patent implements aggregate threat scores that continuously evaluate multiple attributes (linkage scores, anomaly scores, temporal information) to detect coordinated fraudulent behavior below traditional threshold levels.
Solution Approach 2:
The patent adds temporal dimension to fraud detection by incorporating temporal information weighting into the threat score calculation. This transforms the detection from a single-point threshold check to a multi-dimensional evaluation considering timing patterns, sequence of events, and temporal relationships between attributes, enabling detection of organized groups that operate below static thresholds.
2Measurement precision
If comprehensive attribute analysis is performed to improve fraud detection, then detection accuracy increases, but computational complexity increases
Solution Approach 1:
The system segments the comprehensive attribute analysis into distinct scoring components: linkage scores for attribute-entity relationships, anomaly scores for behavioral deviations, and temporal weightings for timing patterns. This segmentation allows parallel computation of each component and modular aggregation into the final threat score, reducing overall computational complexity while maintaining comprehensive analysis.
Solution Approach 2:
The patent implements selective attribute analysis by focusing computational resources on the most discriminative attributes for each entity. The system calculates linkage and anomaly scores only for relevant attributes rather than uniformly analyzing all possible attributes, achieving high detection precision with reduced computational overhead through targeted partial analysis.
3Reliability
If temporal information is incorporated to detect organized groups, then detection of collusion improves, but data processing requirements increase
Solution Approach 1:
The system extracts and isolates temporal information from the broader attribute data set, separating timing patterns and temporal relationships into distinct weightable components. This extraction allows the system to focus computational processing specifically on temporal patterns relevant to collusion detection without redundantly processing entire data sets, reducing overall data processing requirements.
Solution Approach 2:
The patent performs preliminary temporal analysis by pre-calculating temporal weightings and relationships between attributes before the main fraud detection evaluation. This preliminary action prepares temporal data in advance, allowing faster integration with linkage and anomaly scores during the actual threat score calculation, thereby reducing real-time data processing requirements.
Data Source
AI summary
Systems and methods are disclosed herein for tracking related and known attributes and/or online activities connected with a digital identity of an entity. In one embodiment, a computing apparatus is configured to associate unique, anonymous alphanumeric identifiers with an entity and to build a unique mapping of entity attributes/activities with associated temporal information to identify suspicious/outlier behaviors so that fraudulent account access or transactions may be prevented.