Iterative Content Generation for User-Intent Alignment

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing AI systems struggle to accurately generate output content that conforms to user intentions, particularly in enhancing content through iterative processes.

Innovation Solution

A device and method utilizing a neural network model to receive natural language input, determine user intentions, set a target area in base content, generate output content, and iteratively refine it based on similarity calculations to ensure alignment with user intent.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If a neural network model generates output content based on natural language input, then the content generation capability is improved, but the accuracy of aligning with user intentions deteriorates

Engineering Contradiction:

Improvecontent generation capabilityVSAvoidalignment accuracy with user intentions

Core Design Contradiction:

Extent of automationVSMeasurement precision

Solution Approach 1:

The patent implements an iterative generation process where the system generates output content, calculates similarity between the generated content and user intentions, and uses this similarity feedback to refine and regenerate content until the alignment meets acceptable accuracy thresholds

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs continuous iterative generation and refinement of output content rather than a single-pass generation, maintaining the useful action of content improvement through repeated cycles of generation, evaluation, and refinement

Inventive Principle:

Principle #20Continuity of useful action

2Measurement precision

If iterative generation process is implemented to improve content accuracy, then the alignment with user intentions is improved, but the processing time and complexity increase

Engineering Contradiction:

Improvealignment accuracy with user intentionsVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system uses similarity calculation as feedback to determine when to stop iterating, allowing the process to terminate early when acceptable accuracy is achieved rather than always completing a fixed number of iterations

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs iterative generation only to the extent necessary to achieve acceptable alignment accuracy, avoiding unnecessary additional iterations once the content meets the required quality threshold

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12380887B2Device, method, and program for enhancing output content through iterative generation

Publication Date: 2025.08.05 SAMSUNG ELECTRONICS CO LTD

AI summary

A method of improving output content through iterative generation is provided. The method includes receiving a natural language input, obtaining user intention information based on the natural language input by using a natural language understanding (NLU) model, setting a target area in base content based on a first user input, determining input content based on the user intention information or a second user input, generating output content related to the base content based on the input content, the target area, and the user intention information by using a neural network (NN) model, generating a caption for the output content by using an image captioning model, calculating similarity between text of the natural language input and the generated output content, and iterating generation of the output content based on the similarity.