Property Graph Data Model for System Architecture Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current system architecture development tools lack a unified and efficient method to analyze and quantify the consistency and robustness of system architectures, relying on manual and subjective expert evaluations that are inconsistent and lack repeatability, especially in comparing physical and information exchange interfaces.
Innovation Solution
A property graph data model is generated to represent system architectures, allowing for computational analysis of modularity, robustness, and resiliency by transforming data from disparate sources into a unified, scalable format, and a weighted part-to-part coupling graph is used to evaluate consistency and robustness through graph algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual expert evaluation is used to analyze system architecture consistency and robustness, then flexibility in assessment is maintained, but repeatability and objectivity are lost
Solution Approach 1:
The patent replaces manual expert evaluation (mechanical human analysis) with automated computational analysis using graph algorithms. The system transforms system architecture data into graph representations and applies algorithms like PageRank and betweenness centrality to objectively measure consistency and robustness, eliminating subjectivity while maintaining analytical flexibility through configurable parameters.
Solution Approach 2:
The patent introduces quantifiable parameters (graph metrics such as node centrality, edge connectivity, path lengths) to transform qualitative expert assessments into measurable objective data. These parameters enable repeatable analysis while allowing flexibility through parameter adjustment to match different assessment criteria and system types.
2Adaptability or versatility
If data is dispersed across multiple disparate tools with different languages and standards, then domain-specific development capabilities are maintained, but unified analysis capability is lost
Solution Approach 1:
The patent introduces an intermediary layer (graph transformation module) that sits between disparate tool data sources and the analysis engine. This intermediary translates various data formats and standards into a unified graph representation, preserving domain-specific capabilities while enabling unified analysis through a common intermediate format.
Solution Approach 2:
The patent creates a universal graph-based representation that can accommodate multiple data types and sources. The graph model serves as a multi-functional framework that can represent different system architecture aspects (physical hierarchy, information exchange interfaces) from various domains in a unified structure, enabling comprehensive analysis without sacrificing domain specificity.
3Measurement precision
If computational analysis is performed only on various aspects of system design rather than overall architecture, then detailed component analysis is improved, but holistic system analysis is lost
Solution Approach 1:
The patent segments the overall system architecture into graph components (nodes representing system elements, edges representing relationships) while maintaining the complete system view. This segmentation enables precise analysis of individual components through localized graph operations while preserving holistic analysis capability through global graph metrics that capture overall system properties.
Data Source
AI summary
A system and method generates a property graph data model to represent a system architecture. The model includes a plurality of vertices of different vertex types representing different design dimensions of the system architecture, a plurality of compositional edges that connect vertices of a same vertex type to represent hierarchical directed-composition and hierarchical directed-decomposition of the connected vertices with respect to one another, and a plurality of connecting edges that connected vertices of different vertex types to represent connections between different elements of the different design dimensions. Information about the different design dimensions is extracted from one or more source databases to generate the property graph data model.