Markup-Based Optimization Model Generation for Automated Solver Instantiation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for solving optimization problems require human expertise, leading to inefficiencies and bottlenecks, as OR experts are needed to convert markup documents into model instances, making it difficult to adapt to changes in data or solvers.
Innovation Solution
A system that generates a symbolic model from a markup document using neural networks, allowing for the automatic conversion into a model instance without human intervention, enabling storage and retrieval from a database for flexible use with various solvers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a human OR expert manually converts markup documents into model instances, then the conversion can be performed with high accuracy and understanding of the optimization problem, but the process is time-consuming and creates a bottleneck that limits scalability
Solution Approach 1:
The patent replaces the manual mechanical process of OR expert conversion with an automated neural network-based system. The neural network model learns the conversion process from training data and automatically transforms markup documents into model instances, eliminating the need for human experts to perform repetitive manual conversion tasks while maintaining high accuracy through learned patterns.
Solution Approach 2:
The patent creates a computational copy of the OR expert's knowledge and decision-making process through the neural network. The model is trained on examples of markup documents and their corresponding model instances, effectively copying the conversion expertise into an automated system that can replicate the conversion process without human intervention.
2Reliability
If the conversion process requires human OR expert intervention, then the quality of model instance generation can be maintained, but the accessibility and ease of use for non-experts is reduced
Solution Approach 1:
The patent enables non-expert users to automatically generate model instances without requiring OR expert knowledge. The neural network system handles the complex conversion process autonomously when users provide markup documents, allowing anyone with the ability to create markup documents to access optimization solving capabilities without needing specialized expertise in model formulation.
3Adaptability or versatility
If manual conversion processes are used for each solver, then the models can be precisely adapted to specific solver requirements, but the complexity and time required increases significantly
Solution Approach 1:
The patent creates a universal neural network conversion system that can handle multiple solver formats through a single platform. The system is designed to work with different markup document types and can output model instances compatible with various solvers, eliminating the need for separate manual conversion processes for each solver while maintaining the ability to adapt to specific solver requirements through the unified interface.
Data Source
AI summary
Methods and systems for generating a symbolic model from a markup document, and instantiating a model instance from a symbolic model are described. A markup document containing human language content and mathematical content is parsed into a symbolic model that contains only symbolic code representing an optimization problem. The markup document is parsed to extract a markup declaration, the markup declaration is then processed to a math content span, any metadata entity and any relationship between any metadata entity and the math content span. The math content span is processed into a math content parse tree. The math content parse tree is converted into symbolic code of the symbolic model using any relationship between the metadata entity and the math content span. The symbolic model can be instantiated using data definitions.