Natural-Language QUBO Code Generation Without Expert Modeling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing optimization models, such as QUBO, require subject matter expertise to fit real-world problems, creating a bottleneck for timely delivery and innovation by necessitating expert intervention.

Innovation Solution

A method that allows non-experts to generate QUBO source code using natural language inputs, leveraging a vector database and semantic search to transform natural language problem descriptions into executable QUBO code, reducing the need for subject matter expertise.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If expert intervention is used to formally define use cases as optimization models, then the quality and accuracy of the optimization models are improved, but the time required to deliver optimization models and the cost increase

Engineering Contradiction:

Improvequality of optimization modelsVSAvoidtime to deliver optimization models

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent introduces an intermediary system comprising a vector database and semantic search engine that mediates between the user's natural language problem description and the QUBO model generation. This intermediary automatically translates natural language into formal optimization models, eliminating the bottleneck of expert intervention while maintaining model quality through semantic understanding and vector-based problem matching.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system enables users to self-generate QUBO models by directly inputting problem descriptions in natural language without requiring expert assistance. The automated pipeline including semantic search, vector database querying, and template-based code generation allows stakeholders to independently create and solve optimization problems, significantly reducing delivery time and cost.

Inventive Principle:

Principle #25Self-service

2Reliability

If expert intervention is required to fit use cases into QUBO format, then the correctness of the optimization formulation is improved, but the ease of operation and accessibility for non-experts deteriorates

Engineering Contradiction:

Improvecorrectness of optimization formulationVSAvoidaccessibility for non-experts

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The patent employs an intermediary system that bridges the gap between natural language problem descriptions and formal QUBO formulations. The semantic search engine and vector database act as mediators that automatically interpret user intent, retrieve relevant problem templates, and generate correct optimization models, making the process accessible to non-experts while maintaining formulation correctness.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical process of manual expert formulation with an automated computational system. Instead of experts manually translating problems into QUBO format, the system uses natural language processing, semantic search, and template-based generation to automatically create correct optimization formulations, significantly improving ease of operation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Adaptability or versatility

If manual expert formulation of QUBO problems is used, then the adaptability to complex problem requirements is improved, but the productivity and speed of generating optimization models deteriorates

Engineering Contradiction:

Improveadaptability to complex problem requirementsVSAvoidspeed of generating optimization models

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent implements preliminary action by pre-storing a vector database of problem templates and formulations that have been previously defined and validated. When a user submits a new problem, the system performs semantic search to quickly retrieve relevant pre-defined templates and adapt them to the specific requirements, significantly accelerating the generation process while maintaining adaptability to complex problems.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system handles adaptability through parameter changes by allowing users to modify specific parameters and constraints in the generated QUBO models based on their complex problem requirements. The template-based approach with configurable parameters enables rapid adaptation of standard formulations to specific complex scenarios without requiring complete manual re-formulation.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20260003578A1Guided QUBO code generation from natural language

Publication Date: 2026.01.01 DELL PROD LP

AI summary

One example method for generating QUBO (quadratic unconstrained binary optimization) problem source code, includes receiving a natural language description of a target optimization problem to be solved, based on the natural language description, performing a query to obtain a list that includes problems similar to the optimization problem, receiving a user selection of one of the problems in the list, either, retrieving penalties associated with the selected problem, or receiving a user indication that the penalties do not adequately conform with the target optimization problem, and when there is no receipt of the user indication, using the penalties, and the selected problem, to generate source code for a target QUBO problem that represents the target optimization problem.