Payment Service Fraud Detection via Transaction Dissimilarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current payment systems face challenges in efficiently generating and managing receipts for merchants and customers, particularly in scenarios where multiple customers share payment instruments, leading to difficulties in identifying the correct customer for receipt delivery and requiring significant resources from point-of-sale (POS) devices.
Innovation Solution
A payment service that stores receipt templates and generates receipts, allowing merchants to select and customize them, identifies the correct customer using transaction data and customer profiles, and sends receipts electronically, thereby reducing the burden on POS devices and ensuring accurate delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple customers share payment instruments, then payment flexibility and customer convenience are improved, but the ability to accurately identify the correct customer for receipt delivery deteriorates
Solution Approach 1:
The system segments customer identification by creating unique customer profiles associated with each payment instrument, dividing the identification process into distinct segments (device identifiers, transaction patterns, customer information) rather than relying on a single identification method. This allows accurate customer differentiation even when multiple customers share payment instruments.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring transaction data, device identifiers, and customer information to refine and update customer profiles. This feedback loop enables the system to learn from past transactions and improve customer identification accuracy over time, ensuring receipts are delivered to the correct customer.
2Ease of operation
If receipts are generated and managed by POS devices, then receipt delivery control is improved, but the resource burden on POS devices deteriorates
Solution Approach 1:
The system extracts the receipt generation function from POS devices and relocates it to a centralized server. POS devices now only need to transmit transaction data and receive generated receipts, significantly reducing their computational burden and resource consumption while maintaining control over the receipt delivery process.
Solution Approach 2:
A centralized server acts as an intermediary between POS devices and customers for receipt generation and delivery. This intermediary handles the complex processing of generating multiple receipt versions and determining delivery destinations, freeing POS devices from these resource-intensive tasks while maintaining end-to-end control of the receipt delivery process.
3Reliability
If physical receipts are used, then customer receipt delivery is ensured, but environmental waste and distribution costs increase
Solution Approach 1:
The system creates digital copies of receipts that can be delivered electronically to customers via email or mobile devices. These digital copies serve as reliable alternatives to physical receipts, eliminating the need for paper production, distribution, and disposal while maintaining receipt delivery assurance through electronic delivery confirmation.
4Productivity
If receipt templates are pre-configured, then receipt generation speed is improved, but the ability to customize receipts for different scenarios deteriorates
Solution Approach 1:
The system implements dynamic receipt template selection where the appropriate template is automatically chosen based on transaction characteristics, customer preferences, and delivery method. Templates can be dynamically adjusted and customized for different scenarios without sacrificing generation speed, as the system selects from pre-configured options and applies specific parameters based on the transaction context.
Data Source
AI summary
This disclosure describes, in part, a payment service predicting that use of a payment instrument is fraudulent. In examples, a user profile and associated account may be stored in a data store associated with the payment service. The user profile may include an identifier of a payment instrument and historical transaction data for transactions of the user. A payment authorization request may be received by the payment service that includes the payment instrument identifier. If a dissimilarity metric (determined by comparing characteristics of the transaction data for the pending transaction to the historical transaction data) exceeds a threshold, it may be predicted that the payment authorization request is not associated with the user, and a fraud alert may be sent to a user device.


