Semantic Knowledge Graph for Legacy Data Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Integrating data from various legacy applications into a unified system is challenging, especially when data is not directly stored in the knowledge graph but in files, databases, or data lakes, making it difficult to search and query efficiently.
Innovation Solution
A semantic knowledge graph is used to model and integrate data across legacy applications, allowing for the selection of relevant data patterns, traversal reasoning, and automatic creation of new subgraphs and datasets, enabling querying by semantics and relationships without moving master data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If data is stored in legacy applications (files, databases, data lakes) rather than in the knowledge graph, then data can be retained in existing systems without migration, but data searching and querying becomes difficult and inefficient
Solution Approach 1:
The patent introduces a knowledge graph as an intermediary layer between legacy applications and query interfaces. The knowledge graph models data from multiple legacy systems (files, databases, data lakes) using semantic relationships, enabling efficient querying without moving the actual master data from its original storage locations. This mediator approach resolves the contradiction by maintaining data in legacy systems while providing efficient access through the knowledge graph.
Solution Approach 2:
The patent creates semantic copies of data relationships and structures in the knowledge graph without physically moving the actual data. The knowledge graph contains modeled representations of data from legacy applications, allowing queries to be executed efficiently on these models while the original data remains in its source systems. This copying approach enables efficient querying while preserving the original data storage architecture.
2Adaptability or versatility
If a semantic knowledge graph is used to model and integrate data across legacy applications, then data can be queried by semantics and relationships without moving master data, but the system complexity increases due to graph pattern identification and subgraph creation
Solution Approach 1:
The patent segments the complex task of knowledge graph integration into manageable components: graph pattern identification, subgraph selection, subgraph creation, and knowledge graph modification. Each component handles a specific aspect of the integration process, making the overall system more manageable despite the inherent complexity of semantic data integration across multiple legacy applications.
Solution Approach 2:
The patent performs preliminary actions by pre-identifying relevant graph patterns and pre-selecting subgraphs based on query requirements before executing the full integration and querying process. This preliminary preparation reduces the complexity of the main processing by having key decisions and structures ready in advance, streamlining the overall system operation.
3Loss of information
If subgraphs are selected and created based on missing data elements and inferred knowledge data, then the knowledge graph can be enhanced with inferred information, but the processing time and computational resources increase
Solution Approach 1:
The patent applies partial action by selectively creating subgraphs only for missing data elements and inferred knowledge that are relevant to specific queries, rather than comprehensively inferring and adding all possible knowledge to the entire graph. This approach balances information completeness with processing efficiency by performing inference operations only where and when needed.
Solution Approach 2:
The knowledge graph system performs self-service by automatically identifying missing data elements and creating subgraphs with inferred knowledge data without requiring manual intervention. The system autonomously enhances itself by detecting gaps in the knowledge graph and filling them through inference, reducing the need for external processing while maintaining information completeness.
Data Source
AI summary
Various embodiments are provided for identifying relevant graph patterns in a knowledge graph in a computing environment by a processor. Data elements may be identified from a knowledge graph and associated datasets that is related to one or more nodes of the knowledge graph and external to the knowledge graph. One or more subgraphs may be selected and created based on missing data elements and inferred knowledge data. The knowledge graph may be modified with one or more subgraphs.