Configurable Ontology to Data Model Transformation Preserving Metadata
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data modeling tools face limitations in transforming large ontologies like the Financial Industry Business Ontology (FIBO) into Physical Data Models, particularly in handling anonymous classes and losing metadata for object properties, and lack configurability in transformation rules and naming standards.
Innovation Solution
The Configurable Ontology to Data Model Transformation (CODT) system uses RDF Query Language (SPARQL) to extract ontology metadata, transforming it into standardized Metadata Sets that provide a holistic view, allowing for configurable transformation rules and naming standards, rather than relying on procedural algorithms or parsing source files.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional parsing approaches are used to transform ontology files, then individual elements can be processed, but the approach reaches its limits with very large ontologies like FIBO and loses metadata for object properties with particular design patterns
Solution Approach 1:
The patent introduces an intermediary layer (the transformation system with configurable rules and metadata sets) between the ontology platform and data modeling tool. This intermediary preserves metadata by intercepting the transformation process, applying configurable rules that maintain object property information, and translating ontology elements into data model elements without losing critical metadata.
Solution Approach 2:
The system changes parameters by using configurable transformation rules and metadata sets that can be adjusted to preserve different types of metadata. The approach transforms how ontology parameters are handled by introducing configurable mapping rules that maintain object property characteristics during the transformation from ontology to data model.
2Adaptability or versatility
If existing data modeling tool imports are used, then RDF/OWL files can be transformed, but the transformation does not enable users to change mapping and transformation rules or apply naming standards
Solution Approach 1:
The system implements dynamics by making transformation rules and metadata sets configurable and adaptable. Users can dynamically adjust mapping rules, transformation parameters, and naming standards without changing the underlying system architecture. This allows the same transformation engine to adapt to different ontology types and data modeling requirements.
Solution Approach 2:
The patent creates a universal transformation system that can handle multiple ontology formats (RDF, OWL) and transform them into various data model representations. The configurable metadata sets and transformation rules make the system multi-functional, capable of serving different data modeling needs while maintaining a unified approach to the transformation process.
3Ease of manufacture
If ontology object properties are transformed into data model relationships by default, then the transformation is straightforward, but metadata for object properties with particular design patterns is lost
Solution Approach 1:
The system applies preliminary action by establishing configurable transformation rules before the actual transformation occurs. These pre-configured rules identify object properties with particular design patterns and apply appropriate transformation strategies to preserve their metadata, preventing information loss before it occurs.
Solution Approach 2:
The patent applies local quality by treating different object properties differently based on their specific characteristics. Instead of a blanket transformation approach, the system analyzes each object property's design pattern and applies targeted transformation rules that preserve the specific metadata relevant to that property type, maintaining local optimality in the transformation process.
Data Source
AI summary
A computer system, storage medium, and method are disclosed for transforming an ontology into a data model. A user may configure transformation rules and ontology to data model mapping.In the first embodiment, the system comprises components for extraction from a source ontology, transformation into an entity-relationship model, load into particular data modeling tools. Other embodiments comprise an extended configuration, analytics, and user interface component.The storage medium holds standardized metadata sets for source ontology, a generic entity-relationship model representation, and data modeling tool tool-specific metadata, with machine-readable instructions to self-populate.The method may use SPARQL to extract ontology metadata, 4GL language to transform ontology into data model metadata sets, and import files or direct access to load metadata into the data modeling tool.The system, storage medium, and method can operate in reverse, transforming a data model into an ontology.