Constant-Size Index Data Block Merging via Cardinality Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large relational databases face inefficiencies in data access due to sequential searches, and constant-space indexes, while reducing false positive rates, can become inefficient as the database size increases, necessitating a method to improve data block merging for enhanced performance.
Innovation Solution
A computer-implemented method that calculates cardinality differences and combined sizes for data block pairs in a relational database, selects suitable pairs, merges them to form new blocks, and updates the database, maintaining a constant-size index with improved efficiency by rearranging data blocks based on cardinality differences and threshold values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If sequential search is used to access data blocks, then the database can be accessed without additional index structures, but the search efficiency deteriorates significantly especially on slow storage disks
Solution Approach 1:
The patent segments the data blocks into groups and creates compact indexes for each group rather than requiring a full sequential scan. This segmentation allows the system to quickly locate relevant data blocks without searching the entire database, thereby improving search efficiency while maintaining relatively simple index structures.
Solution Approach 2:
The patent performs preliminary organization of data blocks by grouping them based on cardinality characteristics before query execution. This preliminary action creates optimally grouped data blocks that can be efficiently accessed during queries, eliminating the need for sequential searching while keeping index structures compact.
2Reliability
If data blocks are merged to reduce false positive rates in constant-space indexes, then query accuracy improves, but the time and computational resources required for merging increase
Solution Approach 1:
The patent changes the parameter of data block organization by grouping blocks based on cardinality differences rather than simply merging all blocks. This parameter change allows the system to achieve low false positive rates by creating groups with similar cardinality characteristics, reducing the merging time required while maintaining high query accuracy.
Solution Approach 2:
The patent applies local quality optimization by creating different groupings for different portions of the database based on their specific cardinality characteristics. Each group is optimized locally with appropriate merging based on its cardinality profile, achieving overall low false positive rates without requiring extensive global merging operations.
3Productivity
If more data blocks are merged into fewer groups, then the index size is reduced and search speed improves, but the false positive rate increases
Solution Approach 1:
The patent optimizes the balance between search speed and false positive rate by changing the grouping parameter based on cardinality differences. It creates groups with optimal sizes that achieve fast search speeds while maintaining low false positive rates through intelligent cardinality-based segmentation rather than simple consolidation.
Data Source
AI summary
In one aspect, a computer-implemented method includes calculating cardinality differences and combined sizes for one or more data block pairs in a relational database. The method also includes selecting a data block pair from the one or more data block pairs based on a calculated cardinality difference and a calculated combined size associated with the selected data block pair. The method further includes merging the selected data block pair to form a merged data block, and replacing the selected data block pair in the relational database by the merged data block.