Causal Formulation Network Model for Natural Language Explanations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern product formulation is complex and often lacks a centralized source of institutional knowledge, leading to formulation data gaps, anomalies, and conflicts that reduce efficiency, accuracy, and quality.
Innovation Solution
A computer-implemented method using a causal-based formulation network model within a web-based GUI, which maps formulation parameters and outcomes to a multi-dimensional product-agnostic formulation backbone, generating natural language explanations of how changes in formulation parameters impact desired outcomes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If formulation data is distributed across multiple sources without centralization, then institutional knowledge is preserved in various forms, but formulation data gaps, anomalies, and conflicts occur that reduce efficiency and accuracy
Solution Approach 1:
The patent combines multiple distributed formulation data sources into a single integrated formulation network model. This model unifies ingredient nodes, product nodes, and formulation data from various sources (internal databases, external databases, expert knowledge) into a cohesive structure, eliminating data gaps and conflicts while preserving the diversity of original sources through proper integration and relationship mapping.
Solution Approach 2:
The formulation network model acts as an intermediary layer between distributed data sources and the formulation system. It standardizes and harmonizes data from multiple sources through a unified ontology and data structure, resolving anomalies and conflicts before presenting integrated information to users, thus maintaining data integrity without requiring changes to source systems.
2Productivity
If expert formulator knowledge is not memorialized in a transferable manner, then individual expertise is retained, but formulation efficiency and quality suffer due to lack of systematic knowledge transfer
Solution Approach 1:
The patent captures and stores expert formulator knowledge in the formulation network model through structured data representations. Expert insights, formulation rules, and best practices are encoded as relationships and attributes within the network, creating transferable digital copies of tacit knowledge that can be systematically applied across multiple formulation projects without relying on individual experts' availability.
Solution Approach 2:
The system performs preliminary knowledge capture and structuring during the model building phase, organizing expert knowledge into reusable formats before formulation tasks begin. This advance preparation of knowledge structures enables rapid retrieval and application during actual formulation work, significantly improving efficiency without requiring experts to be present during each formulation process.
3Reliability
If a comprehensive formulation network model is created to integrate all formulation data, then a single source of truth is achieved, but the system complexity and implementation difficulty increase
Solution Approach 1:
The formulation network model is segmented into distinct modular components: ingredient nodes with properties, product nodes with specifications, formulation records with relationships, and knowledge entities with rules. This segmentation allows the complex system to be built, managed, and queried in manageable sections while maintaining the integrity of the unified data structure and single source of truth.
4Ease of operation
If complex formulation data is presented without explanation, then complete information is provided, but user understanding and decision-making efficiency are reduced
Solution Approach 1:
The system introduces an explanation layer that acts as an intermediary between the complex formulation network model and the user interface. This layer translates complex network relationships, causal connections, and data dependencies into human-readable explanations and insights, preserving complete underlying information while presenting it in an easily understandable format that aids user decision-making.
Data Source
AI summary
A computer-implemented method for generating natural language explanations of product formulations includes implementing a causal-based formulation network model within a web-based graphical, activating a target causal path of the causal-based formulation network model based on subscriber input; constructing a formulation impact explanation prompt based on a formulation outcome node and a sequence of interconnected formulation parameter nodes of the target causal path; generating, by a large language model, a natural language explanation of the target causal path based on an input of the formulation impact explanation prompt; and surfacing, by the web-based graphical user interface, the natural language explanation of the target casual path.