Current Inventory Image Matching with Predicted User Preferences

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Online systems fail to present images of items that accurately reflect user preferences and the inventory at retailer locations, leading to dissatisfaction and suboptimal ordering experiences.

Innovation Solution

An online system uses machine learning to predict user preferences based on collected data and matches inventory images with these preferences, selecting items for display in the user interface based on similarity scores.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If default images are used for items, then the user interface is simple and easy to operate, but the images do not reflect user preferences or actual inventory, leading to user dissatisfaction

Engineering Contradiction:

Improveuser interface simplicityVSAvoidimage accuracy to user preferences

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The patent applies local quality by customizing images based on individual user preferences and characteristics. Instead of using a single default image for all users, the system selects images from the actual inventory that match each user's specific preferences (e.g., marbling amount for steaks, color for fruits), thereby improving image accuracy while maintaining interface simplicity.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system changes the parameter of image selection based on user preference attributes. By analyzing user data and predicting preferences for specific item attributes (such as meat marbling, fruit color, vegetable ripeness), the system dynamically selects images that match these parameters, resolving the contradiction between simplicity and accuracy.

Inventive Principle:

Principle #35Parameter changes

2Reliability

If images are selected to match user preferences, then user satisfaction improves, but the system complexity increases due to machine learning models and inventory matching processes

Engineering Contradiction:

Improveimage accuracy to user preferencesVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system uses copies of actual inventory images rather than creating new synthetic images. By selecting and displaying copies of real items from the retailer location's inventory that match user preferences, the system achieves high image accuracy without the complexity of generating artificial images, thereby managing system complexity while improving reliability.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent introduces machine learning models as intermediaries between user data and image selection. These models process user preferences and inventory information to automatically determine matching images, reducing the need for complex manual curation processes while maintaining high accuracy in representing user preferences.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If the system displays all available inventory items, then the user interface is comprehensive, but it becomes difficult to present items that specifically match user preferences

Engineering Contradiction:

Improveinventory coverageVSAvoidpreference matching capability

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The system segments the inventory display by category and applies preference filtering within each segment. Instead of presenting all inventory items uniformly, the system divides items into categories (meat, produce, dairy, etc.) and then selects specific items within each category that match user preferences, maintaining comprehensiveness while improving preference matching capability.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by selecting only the portion of inventory items that are most relevant to user preferences rather than displaying everything. By using machine learning to identify and present only the most suitable items (partial selection), the system maintains ease of operation while still providing comprehensive coverage of relevant inventory options.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20250307896A1Matching Images of Current Inventory with Machine Learning Predictions of User Preferences to Customize User Interface

Publication Date: 2025.10.02 MAPLEBEAR INC

AI summary

An online system retrieves user data for a user and applies a machine-learning model to predict a measure of preference of the user associated with an item category based on the user data. For an item included in the item category, the system receives information describing an inventory of the item at a retailer location and a request from a client device of the user to access a user interface describing items included among the inventory at the retailer location. The system determines a measure of similarity between the information describing the inventory of the item and the predicted measure of preference of the user and computes a score based on the measure of similarity. The system selects items to include in the user interface based on the score, generates the user interface including information describing the selected items, and sends the user interface to the client device.