Machine-Learning Digital Standards for Accurate Information Retrieval

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for converting paper standards to digital formats are inefficient and inaccurate, leading to inconsistent and time-consuming information retrieval, as they rely on manual rule-based classification that often misses important information and lacks a standardized format for digital presentation.

Innovation Solution

A machine-learning model is trained to extract and classify conceptual units from underlying standards using a schema, converting them into digital standards that provide a consistent and functional digital layout, enabling intuitive information retrieval and interaction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If manual rule-based classification is used to convert paper standards to digital formats, then the conversion process is simple to implement, but the accuracy and consistency of information retrieval deteriorates

Engineering Contradiction:

Improveease of implementationVSAvoidaccuracy of information retrieval

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent replaces manual rule-based classification (mechanical system) with a machine-learning model (intelligent system) that automatically extracts and classifies conceptual units from standards. The model learns patterns from training data and applies them to convert paper standards to digital formats, significantly improving accuracy while maintaining ease of implementation through automation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the fundamental parameter of the classification system from fixed manual rules to dynamic machine-learning models that can adapt and learn from data. This parameter change enables the system to improve accuracy over time while maintaining implementation simplicity through automated processes.

Inventive Principle:

Principle #35Parameter changes

2Device complexity

If manual rule-based classification is used to convert standards, then the implementation process is straightforward, but the time required for conversion increases

Engineering Contradiction:

Improvecomplexity of conversion processVSAvoidconversion speed

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent replaces manual rule-based classification with automated machine-learning models that process standards much faster. The models automatically extract conceptual units, classify them, and generate digital formats without human intervention, dramatically increasing conversion speed while keeping the process straightforward through automation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent implements preliminary training of machine-learning models on annotated standards before actual conversion. This preliminary action prepares the models to quickly and accurately convert new standards without requiring complex manual rule setup for each conversion task, thereby increasing productivity while maintaining process simplicity.

Inventive Principle:

Principle #10Preliminary action

3Ease of manufacture

If manual rule-based methods are used for standard conversion, then the system is easy to implement, but information completeness and accuracy deteriorates

Engineering Contradiction:

Improveease of implementationVSAvoidinformation completeness

Core Design Contradiction:

Ease of manufactureVSLoss of information

Solution Approach 1:

The patent replaces manual rule-based classification with machine-learning models that are trained on annotated standards containing complete information. The models learn to extract all relevant conceptual units and their relationships, ensuring information completeness while maintaining ease of implementation through automated processing.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent uses annotated standards as training data, providing feedback to the machine-learning models during training. This feedback mechanism enables the models to learn from correct classifications and improve their ability to extract complete and accurate information from standards, preventing information loss while keeping the system easy to implement.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12475387B2Generation of digital standards using machine-learning model

Publication Date: 2025.11.18 SAE INT

AI summary

One embodiment provides a method for generating a digital standard utilizing a trained machine-learning model, the method including: training at least one machine-learning model to generate digital standards from underlying standards utilizing a schema, wherein the training includes: receiving, for unstructured information within the underlying standards, a plurality of annotated underlying standards including a set of underlying standards having annotations identifying a classification of conceptual units within the set of underlying standards and corresponding to the schema; and teaching, for structured information within the underlying standards, the at least one machine-learning model patterns delineating information as belonging to conceptual units within the schema; and deploying the at least one trained machine-learning model to convert a second set of underlying standards to the digital standards, wherein the second set of underlying standards is different than the set of underlying standards. Other aspects are described and claimed.