Fraud Detection for Infrequent Users via Transaction Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fraud detection systems are ineffective in scaling strong authentication solutions to millions of users in a cost-effective manner, particularly for consumer applications, and struggle to protect infrequently active users from online fraud and identity theft.
Innovation Solution
A method and system that categorize transactions as anomalous or not anomalous by comparing recent transactions with historical transactions using a similarity value and predetermined threshold, allowing for incremental analysis and confidence computation based on transaction similarity and factors like age or status.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If enterprise authentication solutions are deployed to secure consumer applications, then security and fraud detection capability are improved, but cost and scalability deteriorate
Solution Approach 1:
The patent employs lightweight, computationally inexpensive authentication mechanisms that can be deployed at scale to millions of users. Instead of using heavy enterprise-grade authentication systems, the invention uses simple risk scoring based on transaction behavior patterns that can be rapidly processed and discarded, enabling cost-effective scaling to consumer-level volumes while maintaining fraud detection capability.
2Measurement precision
If behavioral engines are used to learn consumer behavior patterns, then fraud detection accuracy is improved, but effectiveness for infrequently active users deteriorates
Solution Approach 1:
The patent pre-establishes risk thresholds and authentication requirements before transactions occur. For infrequently active users, the system pre-defines higher risk thresholds that trigger additional authentication steps, rather than relying on learned behavior patterns that require extensive historical data. This preliminary configuration ensures protection is in place even when insufficient behavioral data exists.
Solution Approach 2:
The system dynamically adjusts risk scoring parameters and authentication thresholds based on user activity patterns. For infrequently active users, the patent modifies the parameters by incorporating broader contextual factors and setting different threshold values compared to active users, allowing effective fraud detection despite limited transaction history.
3Reliability
If strong authentication measures are implemented, then security against identity theft is improved, but user convenience and transaction volume deteriorate
Solution Approach 1:
The patent implements dynamic authentication that adapts to each transaction's risk level. Instead of applying uniform strong authentication to all users, the system continuously assesses transaction risk based on behavior patterns, device information, and contextual factors, adjusting authentication requirements in real-time. This dynamic approach maintains security while minimizing friction for low-risk transactions.
Solution Approach 2:
The system applies different authentication strengths to different users and transactions based on their specific risk profiles. Rather than implementing blanket strong authentication, the patent tailors the level of security measures to each individual case, applying enhanced verification only where necessary based on localized risk assessment, thereby preserving user convenience for low-risk scenarios.
Data Source
AI summary
A method of categorizing a recent transaction as anomalous includes a) receiving information about a recent transaction and b) accessing information about one or more historical transactions. The one or more historical transactions have at least one party in common with the recent transaction. The method also includes c) determining a similarity value between the recent transaction and a transaction i of the one or more historical transactions and d) determining if the similarity value is greater than or equal to a predetermined threshold value. The method further includes e) if the similarity is greater than or equal to the predetermined threshold value, categorizing the recent transaction as not anomalous or f) if the similarity is less than the predetermined threshold value, determining if there are additional transactions. If there are additional transactions, incrementing counter i and repeating steps c) through f).