Component Classification Using Geometry and Metadata Confidence Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing component management services face challenges in organizing components effectively across libraries and project teams due to inconsistent or incomplete metadata, leading to inefficient keyword searching and time-consuming manual searches.
Innovation Solution
A component management service utilizes a combination of machine-learning (ML) model-based geometry and text classification to group related components, calculating an overall confidence score for each classification, enabling quicker and more accurate component selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If keyword searching is used on component metadata, then users can search for components using simple keywords, but the search results are incomplete and inaccurate due to inconsistent or incomplete metadata
Solution Approach 1:
The patent introduces an intermediary classification system that acts as a mediator between the user's keyword search and the component database. Instead of directly searching component metadata (which is inconsistent), the system first classifies components into standardized categories using multiple attributes (geometry, metadata, relationships). This intermediary classification layer enables accurate search results even when individual metadata fields are incomplete or inconsistent across different libraries and content packs.
2Ease of operation
If components are organized using individual library or content pack conventions, then each library maintains effective internal organization, but there is no effective organization across the entire collection of components
Solution Approach 1:
The patent creates a universal classification system that serves multiple functions simultaneously: it maintains the internal organization structure of individual libraries and content packs while also providing cross-library organization. The system uses a multi-attribute classification approach (geometry, metadata, relationships) that can accommodate different library conventions while establishing a unified framework for organizing the entire component collection across all libraries and content packs.
Solution Approach 2:
The patent segments the component organization system into multiple independent classification dimensions (geometry-based classification, metadata-based classification, relationship-based classification). Each dimension can operate independently and maintain its own conventions, while the combination of these segmented classifications creates a comprehensive cross-library organization system. This segmentation allows different libraries to maintain their internal structures while being integrated into a unified whole.
3Measurement precision
If users manually page through component libraries to find suitable components, then they can ensure thorough review, but it wastes considerable time
Solution Approach 1:
The patent performs preliminary classification and organization of components into standardized categories based on multiple attributes (geometry, metadata, relationships) before the user needs to search. This preliminary action creates a pre-organized structure that enables rapid retrieval of relevant components. When users search, the system can quickly filter and present relevant components from the pre-classified groups, eliminating the need for manual paging through entire libraries while ensuring accurate results.
Data Source
AI summary
In example embodiments, a component management service identifies related components using a combination of ML model geometry-based classification and text-based classification. The component management service accesses a plurality of components, wherein each component is associated with a geometry mesh and textual metadata. It classifies each component based on geometric similarity by providing the geometry mesh as input to one or more ML models. A geometry classification confidence score is produced for each classification. The component management service also classifies each component based on textual similarity by providing the textual metadata as input to the one or more ML models. A textual classification confidence score is produced for each classification. It calculates an overall confidence score for each classification by combining the geometry classification confidence score and the textual metadata classification confidence score. It displays components together that have an overall confidence score that exceeds a threshold for a same classification.