Automated Item Attribute Tagging via User Query Engagement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Manual tagging of items with attributes in large ecommerce catalogs is expensive and time-consuming, making it infeasible to assign accurate attribute tags across all items effectively.

Innovation Solution

A data generation system that automatically identifies and tags items with attributes by analyzing user queries and engagement data, using a computing device to generate training datasets and machine learning models to classify items based on engagement thresholds, thereby reducing the need for manual labeling.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual tagging is used to assign attributes to items, then attribute assignment accuracy is improved, but time consumption and cost increase significantly

Engineering Contradiction:

Improveattribute assignment accuracyVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables items to self-tag by analyzing their own query lists and engagement data. Each item's attribute tags are automatically determined based on the attributes that appear most frequently in user queries related to that item, eliminating the need for manual tagging while maintaining accuracy through data-driven automation.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

User queries serve as an intermediary between items and attribute tags. Instead of directly tagging items, the system analyzes the natural language queries users submit when searching for or interacting with items, extracting attribute information from these queries to automatically assign tags. This intermediary approach captures real-world attribute usage context.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If manual tagging is used to assign attributes to items, then attribute assignment accuracy is improved, but cost increases significantly

Engineering Contradiction:

Improveattribute assignment accuracyVSAvoidcost

Core Design Contradiction:

Measurement precisionVSLoss of energy

Solution Approach 1:

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system copies attribute information from user queries to item tags. By analyzing the attributes mentioned in natural language queries and copying them as tags to relevant items, the system creates accurate attribute assignments without human intervention, leveraging existing user-generated content.

Inventive Principle:

Principle #26Copying

3Productivity

If automated tagging is implemented, then time consumption is reduced, but attribute assignment accuracy deteriorates

Engineering Contradiction:

Improvetagging speedVSAvoidattribute assignment accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system uses engagement data as feedback to refine attribute assignments. By analyzing which attributes appear in user queries that lead to item engagements (clicks, purchases, adds to cart), the system identifies the most relevant attributes for each item, ensuring accuracy while maintaining automated speed.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary analysis of query lists and engagement data to pre-determine attribute assignments before items are tagged. By预先 analyzing the relationship between queries and items, the system prepares accurate attribute tags in advance, ensuring both speed and accuracy in the tagging process.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20230135327A1Systems and methods for automated training data generation for item attributes

Publication Date: 2023.05.04 WALMART APOLLO LLC

AI summary

A data generation system can include a computing device that is configured to receive a request to generate a training dataset for an attribute and identify a set of item identifiers from an item database based on an engagement indication. The computing device is further configured to, for each item identifier of the set of item identifiers, obtain a query list including queries resulting in an engagement between the corresponding item identifier and a user and, in response to a portion of queries of the query list including the attribute being above a threshold, assign the corresponding item identifier to the training dataset for the attribute. The computing device is also configured to store the training dataset for the attribute in a training dataset database.