Brand Voice Attribute Extraction for Aligned AI Content Generation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing automated content generation techniques fail to produce content that aligns with a user's preferred brand voice, as existing generative machine learning models are trained on diverse text corpora lacking user-specific brand voices, and users struggle to quantify their brand voice attributes accurately.

Innovation Solution

A method involving a language processing machine learning model to extract brand voice attributes from user-generated content, using prompts to identify narrative, trait classifications, and writing style attributes, and iteratively refine these attributes based on user feedback to generate content tailored to the user's brand voice.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated content generation is performed using existing generative machine learning models, then content can be generated quickly and at low cost, but the content does not align with user-specific brand voice attributes

Engineering Contradiction:

Improvecontent generation efficiencyVSAvoidbrand voice alignment accuracy

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The system performs preliminary extraction of brand voice attributes from user-generated content before content generation. The language processing machine learning model analyzes existing user content to identify narrative style, trait classifications, and writing style attributes, storing these as reference data for subsequent content generation tasks.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback loops where generated content is evaluated against extracted brand voice attributes, and the model iteratively refines its output. User feedback on generated content quality further adjusts the generation process to better align with the brand voice.

Inventive Principle:

Principle #23Feedback

2Manufacturing precision

If users manually define brand voice attributes, then content can be tailored to specific brand voice, but users struggle to accurately quantify their brand voice attributes

Engineering Contradiction:

Improvebrand voice alignment accuracyVSAvoidbrand voice attribute quantification difficulty

Core Design Contradiction:

Manufacturing precisionVSDifficulty of detecting and measuring

Solution Approach 1:

The system enables brand voice attribute extraction to happen automatically without requiring user intervention. The language processing machine learning model autonomously analyzes user-generated content to extract brand voice attributes, eliminating the need for users to manually define or quantify these attributes.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system replaces the manual mechanical process of defining brand voice attributes with an automated machine learning-based extraction process. The language processing model automatically identifies and quantifies brand voice characteristics from content analysis, substituting human cognitive effort with computational analysis.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If existing generative models are used without user-specific training, then resource utilization is low and generation is fast, but content quality does not meet user expectations

Engineering Contradiction:

Improvecontent generation speedVSAvoidcontent quality satisfaction

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary extraction and storage of user-specific brand voice attributes before content generation. This pre-processing step enables the generative model to incorporate user-specific characteristics without requiring time-consuming retraining, maintaining generation speed while improving content quality.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system changes the parameters fed to the generative model by incorporating extracted brand voice attributes (narrative style, trait classifications, writing style) as conditioning inputs. This allows the model to generate content aligned with user preferences without altering the model's fundamental structure or requiring full retraining.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250342835A1Automated extraction of brand voice attributes for generation of content inbrand voice through machine learning

Publication Date: 2025.11.06 INTUIT INC

AI summary

Aspects of the present disclosure provide techniques for automated extraction of brand voice attributes for content generation in a brand voice through machine learning. Embodiments include determining a set of user-generated content items associated with a user from which to extract brand voice attributes and providing the set of user-generated content items to a language processing machine learning model along with a prompt that instructs the language processing machine learning model to extract the brand voice attributes from the set of user-generated content items. Embodiments include receiving the brand voice attributes from the language processing machine learning model in response to the prompt. Embodiments include automatically generating content based on the brand voice attributes. Embodiments include outputting the content for display via a user interface.