Dynamic Summarization for Real-Time Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fraud detection systems face challenges in accurately identifying fraudulent financial transactions, particularly in real-time, and struggle to address activity deemed suspicious across multiple entities or peer groups.
Innovation Solution
The system implements a dynamic summarization process that generates fraud-indicative information at the entity or peer group level, using predictive models to analyze financial transaction data in real-time or near real-time, and applies financial transaction business rules to determine matching scores and associations between authorized and fraudulent transaction records.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If real-time fraud detection is implemented using predictive models, then fraud detection speed is improved, but system complexity increases
Solution Approach 1:
The system segments fraud detection into multiple independent predictive models that operate in parallel, each analyzing different aspects of transaction data. This allows real-time processing while distributing system complexity across modular components rather than requiring a single complex monolithic system.
Solution Approach 2:
The patent introduces intermediary components including data normalization layers and scoring mechanisms that mediate between raw transaction data and final fraud decisions. These intermediaries simplify the overall system architecture by standardizing data flows and providing clear decision boundaries.
2Measurement precision
If multiple entity dynamic summarization is performed, then fraud detection accuracy is improved, but processing time increases
Solution Approach 1:
The system performs preliminary summarization of entity transaction patterns at regular intervals, creating pre-computed profiles that capture fraud-indicative information. When a new transaction arrives, the system quickly compares it against these pre-existing summaries rather than analyzing all historical data, thus maintaining high accuracy while reducing processing time.
Solution Approach 2:
The patent implements dynamic summarization that focuses computational resources on the most relevant entities and time periods. Rather than uniformly processing all entity data, the system selectively summarizes transactions based on risk indicators and temporal patterns, achieving sufficient accuracy with reduced processing overhead.
3Reliability
If operation reason information is generated for fraud scores, then fraud analysis quality is improved, but computational overhead increases
Solution Approach 1:
The system generates operation reason information selectively based on local conditions. Rather than producing detailed explanations for all fraud scores, the system provides operation reasons only when the fraud score exceeds certain thresholds or when specific fraud patterns are detected, thus improving analysis quality where needed while minimizing unnecessary computational overhead.
Solution Approach 2:
The patent dynamically adjusts the level of operation reason detail based on fraud risk parameters. For low-risk transactions, minimal processing is performed. For high-risk transactions, the system generates comprehensive operation reasons including fraud type probabilities and risk factor breakdowns, optimizing the balance between analysis quality and computational resource usage.
Data Source
AI summary
Systems and methods are provided for operation upon data processing devices are provided for operating with a fraud detection system. As an example, a system and method can be configured for receiving, throughout a current day in real-time or near real-time, financial transaction data representative of financial transactions initiated by different entities. At multiple times throughout the day, a summarization of the financial transaction data (which has been received within a time period within the current day) is generated. The generated summarization is used to determine whether fraud has occurred with respect to a financial transaction contained in the received authorization data or with respect to a subsequently occurring financial transaction.