Vector Database Query Sharding with Neural Network Parameter Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database query methods for vector databases use the same query parameters for all query vectors, leading to redundant queries and poor performance due to varying query ranges required for different query vectors.
Innovation Solution
Implement a neural network model to determine query parameters for each shard based on the query request, adjusting the query range to avoid redundant queries by clustering inventory data and using sequence learning for incremental data, and aggregating results to improve database performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the same query parameters are used for all query vectors, then the query process is simple and consistent, but it causes redundant queries and poor query performance
Solution Approach 1:
The patent divides the database into multiple shards and assigns different query parameters to each shard based on the query vector's characteristics. This segmentation allows the system to avoid redundant queries by tailoring parameter settings to specific data partitions, thereby improving query performance while maintaining operational simplicity through automated parameter selection.
Solution Approach 2:
The patent applies local quality by determining query parameters individually for each shard based on the query vector's properties. Instead of using uniform parameters across the entire database, the system adjusts parameters locally for each shard to match the specific data distribution and query requirements, eliminating redundancy and optimizing performance.
2Measurement precision
If a larger query range is used to ensure comprehensive search results, then query accuracy improves, but query time and computational resources increase
Solution Approach 1:
The patent dynamically adjusts query parameters based on the query vector's characteristics and the specific shard being queried. The system determines optimal parameters in real-time, allowing the query range to adapt between comprehensive and selective based on the data distribution and query requirements, thereby balancing accuracy and time efficiency.
Solution Approach 2:
The patent changes query parameters dynamically for different shards based on the query vector's properties. By adjusting parameters such as search depth and range individually for each shard, the system achieves high accuracy when needed while reducing query time when the data distribution allows for more selective searching.
3Productivity
If query parameters are adjusted for each shard, then redundant queries are reduced and performance improves, but the system complexity increases
Solution Approach 1:
The patent implements self-service by using the query vector's own characteristics to automatically determine the appropriate query parameters for each shard. The system leverages the query vector's properties (such as dimensionality and distribution) to self-determine optimal parameters without requiring external intervention or complex manual configuration, thereby managing system complexity while improving performance.
Data Source
AI summary
The present disclosure relates to the field of computers, and provides an apparatus, a method, and a storage medium for database query, which are applied to a vector database, by acquiring a query request from a user, determining query parameters corresponding to each of multiple shards based on the query request using a preset neural network model, wherein the query parameters control query complexity by affecting a query range of database data of the query request for the corresponding shard; querying database data in each shard based on the query request and the query parameters to obtain a target query result. Accordingly, corresponding query parameters are adjusted for each shard, so that redundant queries are effectively avoided, and database performance is effectively improved.