Product Matching System Using Surrogate Attributes Without Unique Identifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inefficient and unreliable matching of product offers from different sellers in online marketplaces due to the absence of unique identifiers like UPC or GTIN, leading to duplicate search results, incorrect product grouping, and potential sales of incorrect products or prices.
Innovation Solution
A system and method that utilize sets of rules for title similarity, price outlier detection, image similarity, and attribute extraction to match or mismatch product offers, leveraging machine learning and statistical techniques to identify matches and mismatches, and automatically coordinate displaying or removing product offers based on these determinations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If product offers are matched without unique identifiers, then the system can operate with sellers who do not provide UPC or GTIN, but the matching accuracy and reliability deteriorate
Solution Approach 1:
The patent introduces multiple intermediary comparison attributes (product title, product image, price, product classification, product description) to mediate the matching process when unique identifiers are absent. These intermediaries serve as surrogate keys to identify and match identical products across different sellers, resolving the contradiction by providing alternative reliable bases for matching.
Solution Approach 2:
The system changes the parameters used for product identification from relying solely on unique identifiers (UPC/GTIN) to using a multi-parameter approach including title similarity, image similarity, price comparison, classification matching, and description analysis. This parameter transformation enables reliable matching without unique identifiers.
2Reliability
If manual product matching is performed, then matching accuracy can be maintained, but system productivity and operation speed deteriorate
Solution Approach 1:
The patent replaces manual mechanical product matching with an automated computer-implemented system that uses algorithms to compare product attributes (title similarity, image similarity, price, classification, description). This substitution maintains matching accuracy while dramatically improving system productivity and operation speed.
Solution Approach 2:
The system enables self-service product matching where the computer system automatically performs the matching process without human intervention. The system retrieves product offers, compares attributes, determines matches, and coordinates display automatically, freeing operators from manual matching tasks while maintaining reliability.
3Measurement precision
If multiple product attributes are compared for matching, then matching precision improves, but device complexity increases
Solution Approach 1:
The patent segments the product matching process into distinct modular components: title comparison module, image comparison module, price comparison module, classification matching module, and description analysis module. Each module handles a specific attribute independently, improving measurement precision while managing complexity through modular design that can be implemented and maintained separately.
Data Source
AI summary
Systems and methods including one or more processing modules and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform acts of retrieving a first product offer from a first seller for a first product to offer for sale on a website of an online retailer, determining if the first product offer for the first product matches one or more existing candidate product offers in a catalog of the online retailer using one or more sets of rules, and, if the first product offer matches the one or more existing candidate product offers, coordinating displaying at least a portion of the first product offer on the website of the online retailer.