Streaming Transaction Data Analysis for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in efficiently analyzing large volumes of transaction data to detect fraudulent activity, leading to processing and storage resource burdens and latency issues.
Innovation Solution
A computer-implemented method and system that receives predetermined data identifiers, filters streaming transaction data using these identifiers, and asynchronously stores filtered data in a data structure based on predetermined parameters, allowing for real-time analysis and reduced resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large amounts of transaction data are analyzed to detect fraud, then detection accuracy is improved, but processing and storage resources are burdened causing latency
Solution Approach 1:
The patent segments transaction data into streaming batches that are processed in real-time, separating the data flow into manageable units that can be analyzed without overwhelming system resources. This allows continuous fraud detection while maintaining processing speed through batched stream processing.
Solution Approach 2:
The system performs preliminary filtering and processing of transaction data as it streams in, preparing and pre-processing the data before full analysis. This preliminary action reduces the computational burden during final fraud detection, thereby reducing latency while maintaining accuracy.
2Measurement precision
If large amounts of transaction data are analyzed to detect fraud, then detection accuracy is improved, but processing and storage resources are burdened
Solution Approach 1:
The patent extracts and separates relevant transaction data from the larger stream, focusing analysis only on the necessary fields and transactions that pose fraud risk. This extraction reduces the volume of data requiring full processing, thereby improving throughput while maintaining detection accuracy through targeted analysis.
Solution Approach 2:
The system applies partial processing to transaction data, analyzing only the essential parameters needed for fraud detection rather than processing every detail. This partial action reduces computational resource consumption and increases processing throughput while maintaining sufficient accuracy for fraud detection purposes.
3Speed
If transaction data is processed in real-time, then response time is improved, but system complexity increases
Solution Approach 1:
The patent implements a universal processing framework that handles multiple functions within the same real-time stream processing architecture, including filtering, analysis, and storage operations. This multi-functionality reduces the need for separate specialized systems, thereby managing complexity while maintaining real-time response capability.
Data Source
AI summary
Provided is a computer-implemented method for performing analysis of transaction data. The method may include receiving streaming data, wherein the streaming data comprises a plurality of data identifiers and transaction data associated with each transaction of a plurality of transactions, wherein each data identifier is associated with transaction data associated with a transaction so that the transaction data associated with each transaction is separated by a data identifier, filtering the streaming data using at least one predetermined data identifier value to provide filtered transaction data, determining whether the filtered transaction data corresponds to at least one predetermined transaction parameter, and asynchronously storing the filtered transaction data associated with the at least one transaction in at least one data structure. A system and computer program product are also disclosed.