Merchant Transaction Rule Detection Using Payment Histograms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional financial systems lack a mechanism to identify merchant-enforced card-based transaction rules, such as minimum or maximum purchase amounts, in advance, leading to unexpected surcharges or additional purchases for customers.
Innovation Solution
A determination system that generates histograms of payments based on transaction data, uses machine learning models to identify transaction rules, and sends notifications to user devices when they are proximate to merchants with such rules.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional financial systems process transaction data without analyzing merchant rules, then processing speed is maintained, but customers cannot obtain advance notice of transaction rules
Solution Approach 1:
The system performs preliminary analysis of transaction data to identify merchant-enforced rules before customers make purchases. By proactively detecting patterns in transaction amounts and generating rule profiles in advance, the system provides customers with notice of minimum or maximum purchase requirements before they arrive at merchants, eliminating the need for reactive responses to unexpected surcharges.
Solution Approach 2:
The patent segments the large volume of transaction data into merchant-specific datasets, analyzing each merchant's transaction patterns separately. This segmentation allows the system to identify individual merchant rules while processing data in manageable chunks, maintaining processing efficiency while extracting actionable information about each merchant's enforced transaction requirements.
2Measurement precision
If the system analyzes all transaction data to identify merchant rules, then rule detection accuracy improves, but processing time increases
Solution Approach 1:
The system extracts only the relevant features from transaction data needed for rule detection, such as transaction amounts, frequencies, and patterns. By focusing analysis on these specific extracted features rather than processing entire transaction records, the system achieves accurate rule detection while significantly reducing the computational time and resources required.
Solution Approach 2:
The patent applies partial analysis by examining a representative subset of transaction data for each merchant rather than analyzing every single transaction. This partial action approach provides sufficient accuracy for rule detection while maintaining processing efficiency, as the system can identify patterns from a sufficient sample without the overhead of complete data processing.
3Loss of information
If the system sends notifications to all users about all merchants, then information completeness is improved, but user experience deteriorates due to information overload
Solution Approach 1:
The system implements feedback mechanisms where notifications are triggered based on user context, such as proximity to merchants or intent to make purchases. This feedback-driven approach ensures users receive information about relevant rules only when it matters to them, maintaining information completeness for all merchants while delivering personalized, context-appropriate notifications that avoid overwhelming users with unnecessary information.
Solution Approach 2:
The patent applies local quality by customizing notification delivery based on individual user contexts, merchant locations, and transaction patterns. Rather than uniform notification delivery to all users, the system tailors information delivery to local user needs and situations, ensuring each user receives comprehensive information about rules relevant to their specific circumstances without information overload.
Data Source
AI summary
Systems as described herein determine merchant enforced transaction rules. A determination server may receive transaction data associated with a plurality of merchants. The determination server may generate a histogram of payments associated with a merchant category and filter out transaction data having purchase amounts above or below a predetermined threshold. The determination server may determine a first average purchase amount associated with merchants in the merchant category and a second average purchase amount associated with each merchant in the merchant category. The determination server may determine user spending patterns and that a first merchant in the merchant category enforces one or more card-based transaction rules using machine learning models. After determining that a user device is proximately located to the first merchant, a notification indicating the one or more card-based transaction rules associated with the first merchant may be sent to the user device.