Knowledge Graph Compiler for Data Hydration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for building, managing, and using knowledge graphs are labor-intensive, cost-intensive, and often source-specific, application-specific, and domain-specific, making them inefficient for data hydration and integration across multiple source systems.
Innovation Solution
The use of a compiler associated with a knowledge graph that includes a source graph, a domain graph, and a mapping profile, allowing for the conversion of vendor-specific source datasets into domain-specific datasets using a single compiler and knowledge graphs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional systems are used for building and managing knowledge graphs, then data transformation can be performed, but the process becomes labor-intensive and cost-intensive
Solution Approach 1:
The system performs preliminary actions by pre-defining source graphs, domain graphs, and mapping profiles that establish the transformation framework in advance. These pre-configured components enable automated data hydration without requiring manual intervention during actual data transformation operations, thereby improving productivity and reducing time loss.
Solution Approach 2:
The knowledge graph system enables self-service by automatically performing data transformation and hydration operations using the predefined compiler and mapping profiles. The system serves itself by autonomously converting source data to domain-specific data without requiring external manual processing, thus eliminating labor-intensive operations and reducing both time and cost.
2Adaptability or versatility
If multiple source-specific systems are used for data integration, then data from different vendors can be processed, but the system complexity increases
Solution Approach 1:
The system implements universality by creating a single compiler that can handle multiple source systems through the use of domain graphs and mapping profiles. Instead of requiring separate compilers for each vendor, the universal compiler adapts to different data sources by applying predefined mapping rules, thereby maintaining versatility while reducing system complexity to a single unified component.
Solution Approach 2:
The knowledge graph acts as an intermediary layer between diverse source systems and the target domain. The source graphs and mapping profiles serve as mediators that translate between different vendor-specific data formats and the unified domain model, enabling multiple source systems to be integrated without increasing overall system complexity.
3Productivity
If conventional data hydration methods are used, then data can be converted, but the process is source-specific, application-specific, and domain-specific
Solution Approach 1:
The system achieves universality by designing domain graphs and mapping profiles that are reusable across multiple sources and applications. The same domain graph can be applied to different source systems by simply changing the mapping profile, enabling rapid data conversion without redeveloping the entire transformation logic, thus improving both productivity and adaptability.
Solution Approach 2:
The system segments the data transformation process into distinct reusable components: source graphs, domain graphs, and mapping profiles. This segmentation allows each component to be independently developed and reused across different scenarios, enhancing versatility while maintaining efficient data conversion through modular assembly of these components.
Data Source
AI summary
Systems and methods for data hydration are provided. In some examples, a method includes accessing a compiler. In some examples, the compiler is associated with a source graph, a domain graph, and a mapping profile. In some examples, the domain graph includes one or more domain data schemas. In some examples, the method further includes receiving a source dataset from a source system, and applying the compiler to the source dataset from the source system to generate a domain dataset. In some examples, the domain dataset uses at least one of the one or more domain data schemas. In some examples, the method is performed using one or more processors.