ML Expert System for Automated Innovation Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Identifying innovative solutions within complex fields is labor-intensive and requires constant updates to knowledge bases, making it challenging to pinpoint unsolved problems and potential solutions effectively.

Innovation Solution

A machine learning expert system that utilizes trained models to predict innovative solutions by receiving user inputs, defining scenario profiles, and generating candidate innovations that achieve measurable desired outcomes, incorporating human feedback and domain-specific algorithms.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual methods are used to identify innovative solutions by consulting state-of-the-art literature and knowledge bases, then potential solutions can be identified through human analysis, but the process becomes extraordinarily labor intensive and requires constant attention to changes in the knowledge base

Engineering Contradiction:

Improveaccuracy of identifying innovative solutionsVSAvoidtime required for identifying problems and solutions

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual human analysis with an automated machine learning system that processes knowledge bases and identifies innovative solutions. The system uses trained models to automatically analyze technical literature, patents, and domain-specific data, eliminating the need for constant human attention while maintaining or improving identification accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The machine learning system is designed to autonomously monitor and process changes in knowledge bases without requiring constant human intervention. The system self-updates by continuously learning from new data, automatically identifying emerging problems and solutions, and adapting to changes in the domain knowledge base independently.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If comprehensive analysis of massive corpuses of literature and documents is performed to identify unsolved problems and potential solutions, then the quality of innovation identification improves, but the complexity of the system and process increases significantly

Engineering Contradiction:

Improvequality of identifying unsolved problems and solutionsVSAvoidcomplexity of processing massive literature corpuses

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the massive literature corpus into manageable segments and processes them through specialized machine learning models. The system segments analysis tasks into problem identification, solution generation, and validation components, each handled by specific trained models that focus on particular aspects of the knowledge base.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The machine learning system acts as an intermediary between the massive literature corpus and human users. The trained models process and filter the overwhelming amount of information, transforming raw literature data into structured insights about unsolved problems and potential solutions, thereby simplifying the interface between data and user needs.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If multiple options for solving problems are evaluated based on state-of-the-art knowledge, then the likelihood of identifying effective solutions increases, but the labor intensity and resource requirements increase substantially

Engineering Contradiction:

Improvelikelihood of identifying effective solutionsVSAvoidefficiency of evaluating multiple solution options

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent replaces manual evaluation of multiple solution options with automated machine learning models that rapidly assess numerous candidates. The system uses trained models to evaluate the potential effectiveness of different solutions by analyzing their alignment with state-of-the-art knowledge and identified problem requirements, dramatically increasing evaluation efficiency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The machine learning system performs preliminary evaluation and filtering of potential solutions before human review. The trained models pre-assess multiple solution options against established criteria, prioritizing the most promising candidates for further analysis, thereby reducing the overall workload while maintaining high reliability in identifying effective solutions.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11475330B2Machine learning systems and methods for automated prediction of innovative solutions to targeted problems

Publication Date: 2022.10.18 DMASS INC

AI summary

A machine learning server is provided for predicting innovations in one or more scenarios. The machine learning server includes a processor and a memory in communication with the processor. The processor is configured to receive a user input, define a scenario profile based on the user input, apply a first trained machine learning model to the scenario profile to generate at least one target associated with the scenario, wherein the at least one target includes a measurable outcome, prompt for a user selection from the target, apply a second trained machine learning model to the scenario profile and the user selection from the at least one target to generate at least one candidate innovation predicted to achieve the measureable desired outcome, obtain descriptive information related to the at least one candidate innovation, and present the descriptive information related to the at least one candidate innovation.