Graph Index Construction from Unmodified RDBMS Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Relational database management systems (RDBMSs) face challenges in efficiently processing graph data due to the lack of support for vertex and edge identifiers, which limits the construction of graph indexes from existing tables and constrains users to create dedicated tables with specific key columns, thereby restricting the reusability of existing data models.
Innovation Solution
The implementation of Compressed Sparse Row extended with Edge (CSRE) data structures that map primary key values of vertex/edge table rows to integer identifiers, enabling the construction of graph indexes from unmodified RDBMS tables and providing transparent, transactionally consistent vertex/edge identifiers for efficient graph query and algorithm execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If graph indexes are constructed from existing RDBMS tables, then the reusability of existing data models is improved, but the lack of vertex/edge identifier support in traditional RDBMS limits the efficiency of graph query processing
Solution Approach 1:
The patent introduces an intermediary mapping layer between existing RDBMS table rows and graph index structures. This mapping mechanism translates existing table rows into vertex and edge identifiers without modifying the original tables, enabling both reusability of existing data models and efficient graph query processing through the intermediary identification system
Solution Approach 2:
The patent creates a universal identifier system that works across different RDBMS table structures. The vertex and edge identifiers serve multiple functions: they enable graph index construction, support efficient array indexing for graph traversals, and maintain compatibility with existing RDBMS tables through the mapping mechanism
2Productivity
If dedicated tables with specific key columns are created for graph support, then graph query efficiency is improved, but the complexity of data model management increases
Solution Approach 1:
The patent extracts the identifier generation and mapping functionality from the core graph processing logic. By separating the identification mechanism (vertex/edge IDs generated through the mapping system) from the graph index construction, the solution achieves efficient graph query processing without requiring complex modifications to existing data models
3Productivity
If vertex and edge identifiers are added to existing tables, then graph index construction is enabled, but constraints are imposed on existing table structures
Solution Approach 1:
The mapping mechanism acts as an intermediary that enables graph index construction without directly modifying existing table structures. It translates existing table rows into graph identifiers through a mapping layer, preserving table flexibility while enabling graph processing capabilities
Solution Approach 2:
The patent creates a virtual copy of the table row identification system through the mapping mechanism. Instead of adding physical identifier columns to existing tables, it generates vertex and edge identifiers by copying and transforming existing row identifiers through the mapping system, avoiding structural constraints on original tables
Data Source
AI summary
Vertex/edge table rows are mapped to unique integer identifiers, to enable construction of in-memory representation of a graph from existing, unmodified RDBMS tables. The unique integer identifiers are based on an encoding of primary key values of the table rows. The unique integer identifiers are used as graph indexes of the in-memory representation.