Canonical Data Model for Distributed Metadata Exchange
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database management systems face challenges in unifying and standardizing data across disparate systems and applications due to unique data models, leading to difficulties in data discovery and integration, which slows down digital transformation and innovation.
Innovation Solution
A canonical model-driven active metadata exchange framework that uses a Canonical Data Model (CDM) to create a unified mapping of data sources and apps/services, enabling system and app/service-agnostic metadata discovery and standardization, thereby facilitating data computation and integration across multiple metadata repositories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple unique data models are used across different systems and applications, then each system can be customized for specific business requirements, but data discovery and unification across departments and systems becomes difficult
Solution Approach 1:
The patent introduces a canonical data model as an intermediary layer between diverse source systems and the data catalog. This canonical model serves as a standardized representation that enables data discovery and unification without requiring changes to the underlying source systems, thus maintaining their customization while enabling cross-system data access.
Solution Approach 2:
The patent segments the data model into multiple layers: source system-specific data models, a canonical data model layer for standardization, and a data catalog layer for discovery. This segmentation allows each layer to serve its specific purpose while working together to resolve the contradiction between customization and discoverability.
2Productivity
If custom code and solutions are created to connect disparate applications and systems, then data integration can be achieved, but the process slows innovation and leads to brittle integrations
Solution Approach 1:
The patent creates a canonical copy of data models from source systems, storing these canonical representations in a data catalog. This copying approach eliminates the need for custom integration code by providing a standardized interface that can be reused across multiple applications, thereby reducing integration development time and improving productivity.
Solution Approach 2:
The canonical data model serves multiple functions: it acts as a standardized representation for data discovery, a mapping layer for integration, and a governance framework. This multi-functionality replaces the need for separate custom solutions for each integration scenario, reducing both time and complexity.
3Adaptability or versatility
If enterprise data is heavily customized for specific business requirements or stored in raw form, then business-specific needs are met, but data standardization across databases and applications is lost
Solution Approach 1:
The patent applies local quality by maintaining source system-specific data models with their unique customizations intact while introducing a canonical data model layer that provides standardization. Each source system retains its local characteristics and business-specific customizations, while the canonical layer ensures enterprise-wide standardization for data discovery and integration.
Data Source
AI summary
Systems, methods, and computer-readable media are provided for data catalogs, metadata repositories, data discovery, and data governance, and in particular, for a canonical model-driven active metadata exchange for distributed data catalogues. Disclosed implementations include an application independent metadata repository with a Canonical Data Model (CDM). The CDM maintains a single set of use case agnostic mappings between data sources and the CDM. The physical mappings of a particular entity in the CDM are used to determine the different applications that are mapped to it and what objects or data structures that are exposed by that particular data source. Other embodiments may be described and/or claimed.