Crowdsourced Merchant Data Verification System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inconsistent and inaccurate merchant data in transaction records due to dissimilar entries across Point of Sale (POS) terminals, leading to duplicate or obsolete data entries, which hinders accurate tracking and processing of transactions.
Innovation Solution
Utilizing localized knowledge from individuals in the vicinity of merchants through crowdsourcing techniques to verify and correct merchant data, including names, locations, and category codes, by identifying and confirming accurate information from close matches within transaction data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional automated matching algorithms are used to identify duplicate merchant entries, then processing speed is improved, but accuracy deteriorates due to dissimilar and errant data entries
Solution Approach 1:
The patent introduces crowdsourced users as an intermediary layer between the automated matching algorithm and the final duplicate identification. Users manually verify close matches that automated algorithms identify, providing a mediating human judgment step that resolves the contradiction between automated speed and accuracy.
Solution Approach 2:
The system implements a feedback loop where user verification results are fed back into the duplicate identification process. This feedback mechanism allows the system to learn from human judgments and improve future automated matching, gradually enhancing accuracy while maintaining processing efficiency.
2Measurement precision
If crowdsourcing techniques are used to verify merchant data, then data accuracy is improved, but system complexity increases
Solution Approach 1:
The patent segments the data verification process into distinct stages: automated close-match identification, user verification of specific candidates, and final duplicate confirmation. This segmentation allows crowdsourcing to be applied only where needed (close matches) rather than throughout the entire process, managing complexity while maintaining accuracy.
Solution Approach 2:
The system leverages users' existing localized knowledge and familiarity with their vicinity merchants to perform verification. Users naturally apply their own expertise without requiring extensive training or complex interfaces, reducing system complexity while achieving high accuracy through self-service verification.
3Reliability
If manual verification of merchant entries is performed, then data reliability is improved, but processing time increases
Solution Approach 1:
The patent applies partial manual verification only to close matches identified by automated algorithms, rather than manually verifying all merchant entries. This partial action approach maintains high reliability for duplicates while avoiding the excessive time cost of complete manual verification of all data.
Solution Approach 2:
The system performs preliminary automated filtering to identify close matches before submitting candidates for user verification. This preliminary action reduces the volume of data requiring manual review, thereby maintaining reliability through user verification while minimizing the time loss associated with manual processes.
Data Source
AI summary
Systems and methods are provided for verifying and/or confirming merchant data for multiple different transactions. One method generally includes accessing merchant data for a merchant where the merchant data includes a master merchant identifier for the merchant, and querying, by a computing device, multiple users as to which of a list of merchant data entries, for multiple different transactions and including different merchant identifiers, corresponds to the master merchant identifier for the merchant. The users are associated with a location within a predefined distance of the confirmed merchant and/or have performed one or more purchase transactions at the merchant. The method also generally includes receiving, at the computing device, a response from at least one of the users, in which the at least one of the users selects a data entry from the list, and determining whether to proceed in querying additional users.