Ontology-Based Data Translation for Disparate Database Querying
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Companies face difficulties in querying and updating security clearances across disparate databases with different column identifiers, making it challenging to determine customer impact or update security policies efficiently, especially when dealing with large numbers of lines of business.
Innovation Solution
The implementation of R2RML mappings and machine learning algorithms to identify common elements across databases, coupled with the creation of a semantic ontology, allows for the translation of relational data into RDF format, enabling queries and updates across diverse database systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional ETL methods with SQL statements are used to query relational databases, then data can be retrieved from structured databases, but it becomes difficult and inefficient to query across disparate databases with different column identifiers and schemas
Solution Approach 1:
The patent introduces an intermediary layer (RDF graph database and ontology mapping system) between the heterogeneous relational databases and the query interface. This intermediary translates diverse database schemas into a unified RDF model, enabling consistent queries across different database types without modifying the source databases or requiring complex ETL logic for each query scenario.
Solution Approach 2:
The system creates a universal query interface that works across multiple database types (relational, flat file, graph) by translating all data sources into a common RDF representation. This universal layer allows the same query language and logic to be applied regardless of the underlying database structure, eliminating the need for database-specific query logic.
2Productivity
If manual methods are used to identify common elements across databases with different column identifiers, then security clearance updates can be performed, but the process becomes time-consuming and inefficient especially with large numbers of lines of business
Solution Approach 1:
The system employs machine learning algorithms that automatically analyze database schemas, column names, and data patterns to self-identify common elements and relationships across disparate databases. This automated self-service approach eliminates manual mapping efforts, allowing the system to autonomously build and update the ontology and R2RML mappings even as databases evolve.
Solution Approach 2:
The patent performs preliminary actions by pre-translating and pre-processing database schemas into RDF format and pre-building the ontology mappings before queries are executed. This preliminary transformation of data structures and relationships enables rapid querying without requiring real-time analysis or manual intervention during actual security clearance updates.
3Adaptability or versatility
If disparate database schemas with different column identifiers are maintained, then each database can preserve its own structure, but determining customer impact and updating security policies becomes challenging
Solution Approach 1:
The system changes the parameter representation by translating concrete database column identifiers (which vary across schemas) into abstract semantic concepts defined in the ontology. This parameter transformation allows the system to maintain awareness of the original diverse schemas while operating with a unified semantic model that preserves the meaning and relationships of the data regardless of its source structure.
Data Source
AI summary
A method may include receiving a semantic query, wherein the semantic query is formatted using a first ontology; retrieving a transitive mapping between the first ontology and a second ontology; translating a first element of the semantic query from the format of the first ontology to a second element of the second ontology in accordance with the transitive map; parsing a relational database to Resource Description Format (R2RML) mapping to determine a set of database tables that include the second element; directing an execution of the semantic query to the set of database tables in accordance with the R2RML mapping; and storing the results of the executed query in a tripe format, the stored results formatted using the first ontology.