Clustered Data Retrieval for Speed–Precision Balance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image retrieval systems face challenges in balancing retrieval speed and precision due to the increasing volume of data, leading to inefficiencies in processing large datasets.
Innovation Solution
A data retrieval method that classifies data into first-type and second-type data, using clustering algorithms to optimize retrieval by employing different methods for each type, ensuring precision and speed through a combination of narrowed retrieval ranges and brute-force searches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If brute-force retrieval is used on all data to ensure precision, then retrieval precision is improved, but retrieval speed deteriorates
Solution Approach 1:
The patent divides the database into multiple clusters based on data similarity, creating a hierarchical structure. This segmentation allows the system to first perform retrieval within a specific cluster (narrowed range) and only perform brute-force retrieval on remaining candidates, thus reducing the overall computation while maintaining precision.
Solution Approach 2:
The patent applies different retrieval strategies to different data regions: dense regions (clusters) use narrowed-range retrieval while sparse regions use brute-force retrieval. This local quality approach optimizes the balance between speed and precision for different parts of the data space.
2Speed
If narrowed retrieval range is used to increase speed, then retrieval speed is improved, but retrieval precision deteriorates
Solution Approach 1:
The patent performs preliminary clustering of data before retrieval operations. This preliminary action organizes data into clusters that can be efficiently searched, allowing narrowed-range retrieval to be both fast and accurate by confining the search to relevant clusters.
Solution Approach 2:
The system uses feedback from the clustering structure to guide the retrieval process. By analyzing cluster densities and distributions, the system dynamically adjusts the retrieval strategy, applying narrowed-range search where appropriate and brute-force search where needed to maintain precision.
3Quantity of substance
If data volume increases to improve comprehensiveness, then data completeness is improved, but processing efficiency deteriorates
Solution Approach 1:
The patent segments large volumes of data into manageable clusters, enabling efficient processing of comprehensive datasets. This segmentation allows the system to handle increasing data volumes without proportionally increasing processing time by operating on cluster-level abstractions.
Solution Approach 2:
The patent changes the parameter of data organization from flat storage to hierarchical clustering. This parameter change enables the system to efficiently manage and process large volumes of data by transforming the data structure to support scalable operations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This discloses a data retrieval method, specifically including two parts: a data preparation process and a data retrieval process. In the data preparation process, raw data in a database is classified into first-type data and second-type data. In the data retrieval process, a first retrieval range is determined in the first-type data, and to-be-retrieved data is retrieved in the first retrieval range, to obtain a first retrieval result; the to-be-retrieved data is retrieved from all of the second-type data, to obtain a second retrieval result; and finally, a final retrieval result is determined from the first retrieval result and the second retrieval result. In this way, a problem of a retrieval speed and retrieval precision in the data retrieval process is resolved.