Query Proxy System for Client-Specified Data Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data warehousing approaches for data integration are time-consuming and computationally expensive, with tightly coupled data architectures that make it difficult to maintain data freshness and do not allow clients to specify data models for the warehouse.
Innovation Solution
A query proxy system that translates client-specified query expressions into compatible query expressions for multiple source systems, using a client-specified model and mapping metadata to import and transform data, enabling automated data translation without custom programming.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If data warehousing approach is used for data integration, then unified query interface is provided, but data freshness cannot be maintained without multiple ETL processes
Solution Approach 1:
The patent introduces a virtual warehouse layer that acts as an intermediary between source databases and clients. This virtual warehouse uses views and query translation to provide unified access without physically consolidating data, allowing real-time access to fresh data from multiple sources while maintaining a single query interface.
Solution Approach 2:
Instead of physically copying and storing data in a centralized warehouse, the patent creates virtual copies through database views and query translation mechanisms. These virtual representations allow unified querying without actual data movement, ensuring data freshness while providing integration benefits.
2Adaptability or versatility
If data warehousing approach is used, then data integration is achieved, but computational cost and time are high due to ETL processes
Solution Approach 1:
The patent uses virtual data copying through database views rather than physical data movement. Query translation mechanisms convert client queries into source system queries without requiring ETL processes, dramatically reducing computational overhead while maintaining data integration capabilities.
Solution Approach 2:
The patent extracts only the necessary query translation logic from the data integration process, separating it from physical data movement operations. This allows the system to provide data integration through query rewriting rather than through computationally expensive ETL processes.
3Device complexity
If traditional data warehousing is used, then data is centralized, but client cannot specify data model for the warehouse
Solution Approach 1:
The patent implements dynamic data modeling where the virtual warehouse schema is not fixed but adapts based on client specifications. Clients can define their own data models and views over the integrated data, allowing flexibility and customization without changing the underlying data architecture.
Solution Approach 2:
The patent segments the data model specification functionality from the physical data storage architecture. Clients can specify logical data models and relationships at the view layer without affecting the physical organization of data in source systems, allowing model flexibility while maintaining architectural simplicity.
Data Source
AI summary
A system for translating a query expression for a client-specified model (CSM) to a plurality of query expressions compatible with the object model(s) for a plurality of source systems. The source systems comprise a plurality of databases. The databases comprise source information and X sets of response data. The source information comprises mapping metadata and object model metadata. The X sets of response data is data response to a client query. The client comprises client-specified mapping metadata, a CSM, and a query proxy. The query proxy comprises an import module, a client interface module, and a translator module. The import module imports the source information from the plurality of source systems and determines the mapping metadata and the object model metadata. The client interface module receives the CSM, the client-specified mapping metadata, and a query from a client. The translator module parses the query expression and translates a plurality of query expressions compatible with the object model(s) for the plurality of source systems.