Database Attribute Fetching via Denormalized Property Bags
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accessing specific attributes in a hierarchical class-based object stored in a database can be time-consuming due to the hierarchical structure of database tables, where each attribute is stored in a column of a database table, occupying memory space and requiring traversal from the base table to the attribute's location.
Innovation Solution
Storing attributes in both a primary table at a lower level of the hierarchy and a secondary table at a higher level, allowing for faster access by treating a set of attributes as a whole, reducing the number of columns needed and improving storage efficiency, and using a 'property bag' to store multiple attributes in a single database location.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If attributes are stored in a hierarchical database table structure where each attribute occupies a separate column, then database functionality and search capability are enabled, but access time to retrieve specific attributes increases due to traversal from the base table
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing flattened representations of hierarchical object attributes in a denormalized table structure. This allows the system to bypass hierarchical traversal during attribute access, as the flattened data is already prepared in an accessible format. The preprocessing step creates a trade-off where initial storage complexity is increased to eliminate retrieval complexity.
Solution Approach 2:
The patent creates a copy of the hierarchical attribute data in a denormalized, flattened table structure. This copy maintains the same information as the hierarchical source but organizes it for faster access without requiring traversal through multiple table levels. The copy enables direct attribute retrieval while the original hierarchical structure remains intact for data integrity and update operations.
2Quantity of substance
If each attribute of each object instance is allocated a separate location in database table columns, then attribute storage is enabled, but database memory space is consumed
Solution Approach 1:
The patent merges multiple attribute columns that would otherwise occupy separate database locations into a single denormalized table structure. By combining related attributes from different hierarchical levels into unified columns, the system reduces the total number of column definitions and associated metadata overhead, thereby optimizing database memory space utilization while maintaining full attribute storage capacity.
3Loss of time
If attributes are stored in a denormalized table with flattened hierarchy, then access time is reduced, but the number of columns in the table increases
Solution Approach 1:
The patent segments the denormalized table into multiple specialized tables, each handling specific groups of attributes. This segmentation approach distributes the column complexity across multiple manageable tables rather than consolidating all attributes in a single wide table. Each segmented table contains a subset of columns, making the schema more maintainable and queries more efficient while still providing fast attribute access compared to traditional hierarchical traversal.
Data Source
AI summary
An attribute of a hierarchical class-based object is received. The attribute in a database is stored both in a first location in a first table at a first level of a hierarchy of database tables which table hierarchy is based at least in part on a class hierarchy associated with the hierarchical class-based object and in a second location in a second table at a second level of the table hierarchy which second level is higher than the first level.