Database Query Graph Optimization for Combined ER-Relational Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Database management systems with combined entity-relationship and relational data models face performance issues and maintenance overhead due to complex queries that result in accessing duplicate data fields, leading to inefficient processing and increased server load.
Innovation Solution
The ReadiNow Intelligent Engines (RIE) method optimizes query processing by parsing and converting query texts into query graphs, combining equivalent nodes, encoding relationship types and data fields into state transition and field rules, and preparing structured queries to efficiently access data fields in a combined ER- and relational-model database management system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual queries are used to determine exact data fields and relationship types to access, then query precision is improved, but server processing time increases substantially
Solution Approach 1:
The system performs preliminary analysis of the query graph to pre-determine the exact set of data fields and relationship types that need to be accessed before executing the actual query. This preliminary action includes analyzing the query graph structure, identifying all target entities, and determining the complete set of required data fields in advance, thus avoiding multiple iterative queries and reducing server processing time substantially
Solution Approach 2:
The query processing is segmented into distinct phases: query graph construction, query graph analysis to determine required data fields, and final query execution. By segmenting the process, the system can determine the exact data fields needed before execution without performing multiple sequential queries, thereby improving both precision and time efficiency
2Reliability
If multiple queries are executed to access data fields, then data access completeness is improved, but server processing overhead increases
Solution Approach 1:
The system merges multiple individual data field access operations into a single consolidated query execution. By analyzing the query graph beforehand to determine all required data fields and relationship types, the system can construct one comprehensive query that retrieves all necessary data in a single operation, thereby maintaining data access completeness while significantly reducing server processing overhead
Solution Approach 2:
The system performs preliminary analysis of the query graph to pre-determine the complete set of data fields and relationship types that need to be accessed. This preliminary action enables the construction of a single comprehensive query that retrieves all required data, eliminating the need for multiple sequential queries and reducing server processing overhead while maintaining data access completeness
3Quantity of substance
If complex queries traverse multiple relationship type paths, then data field coverage is improved, but duplicate data access increases
Solution Approach 1:
The system performs preliminary analysis of the query graph to identify all target entities and the complete set of required data fields before query execution. By determining in advance which data fields are actually needed, the system can construct queries that retrieve only the necessary data, avoiding duplicate data access while maintaining comprehensive data field coverage across multiple relationship paths
Solution Approach 2:
The system extracts and identifies the exact set of unique data fields that need to be accessed by analyzing the query graph structure and relationship paths. By extracting only the necessary data fields and excluding duplicates before query execution, the system maintains comprehensive data field coverage while eliminating redundant data access and reducing energy loss
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
Disclosed is a method of accessing a network of relationship instances and data field values of entity instances in a combined entity relationship- and relational-model database management system. The method comprises the steps of receiving a query that natively references relationship types including data fields to be accessed; parsing the query into a parse tree; converting the parse tree to a query graph; optimising the query graph by consolidating equivalent nodes in the query graph; analysing the nodes in the optimised query graph; codifying each node of the optimised query graph into a first table of state transition rules; codifying the data fields to be accessed into a second table of field rules; preparing a query that includes the first and second tables; identifying entity instances to be accessed by the prepared query by iteratively following relationship instances according to the first table; and loading data fields of the identified entity instances according to the second table.