Replacement Model for Automated Product Taxonomy Labeling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Manual addition of hundreds of thousands of products to a hierarchical taxonomy in an online concierge system is inefficient, necessitating an automated method to label unlabeled products based on their categories.

Innovation Solution

An online concierge system employs a replacement model that determines the likelihood of labeled products serving as replacements for unlabeled products, selecting the highest likelihood product to label the unlabeled product with its category within the hierarchical taxonomy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual addition of products to hierarchical taxonomy is performed, then labeling accuracy is maintained, but productivity is severely reduced due to the large number of products (hundreds of thousands)

Engineering Contradiction:

Improveproduct labeling speedVSAvoidtime for manual taxonomy population

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system enables unlabeled products to automatically find their own categories by using the replacement model to identify similar labeled products. The product effectively labels itself through the similarity matching process, eliminating the need for manual intervention in taxonomy population.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system copies category assignments from existing labeled products to unlabeled products based on similarity. By finding replacement products that are already labeled, the system replicates their category information to the new product, enabling rapid taxonomy population without manual work.

Inventive Principle:

Principle #26Copying

2Productivity

If automated category assignment is implemented, then productivity is improved, but measurement precision of category classification may deteriorate

Engineering Contradiction:

Improveproduct labeling throughputVSAvoidcategory classification accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The replacement model provides a likelihood score indicating how well a labeled product matches an unlabeled product. This feedback mechanism allows the system to assess classification confidence and potentially adjust the automation threshold, balancing speed and accuracy based on the quality of matches found.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent replaces manual mechanical classification work with an automated machine learning-based replacement model. The model uses learned patterns from training data to automatically determine category assignments, substituting human judgment with algorithmic decision-making that can process hundreds of thousands of products efficiently.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20220292567A1Inferring categories in a product taxonomy using a replacement model

Publication Date: 2022.09.15 MAPLEBEAR INC

AI summary

An online concierge system accesses a hierarchical taxonomy of products each labeled with a category of the hierarchical taxonomy. The online concierge system receives, from an inventory database, an unlabeled product, which not included in the hierarchical taxonomy. The online concierge system inputs the unlabeled product to a replacement model. The replacement model is trained to output, for each of one or more labeled products from the hierarchical taxonomy, a likelihood that a user would select the labeled product as a replacement for an input product. The online concierge system selects a labeled product from the one or more labeled products based on the likelihoods. The online concierge system adds the unlabeled product to a category of the hierarchical taxonomy based on the selected labeled product.