Indexing method and system of concurrent hash index data structure based on machine learning

A data structure and hash index technology, which is applied to the index of concurrent hash index data structure and the field of efficient concurrent hash index data structure, can solve problems such as incomplete research on hash index, achieve consistency, improve performance, high performance effects

Active Publication Date: 2021-03-02
SHANGHAI JIAO TONG UNIV
View PDF5 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

The technical effect of this patented method improves how well it works for both traditional binary search indexes like LUTS or SHA-1. It also allows for parallel processing without slowdown due to sequential indexing techniques such as merge caching and sparse matrix factorization. Additionally, there are advantages over current methods including improved efficiency and reduced latencies associated with these new systems compared to previous ones. Overall, this improvement makes them more efficient than older technologies at performing complex searches faster and easier.

Problems solved by technology

Technologies described involve improving the performance and scalability of conventional systems like LISAR, XOR logic, etc., particularly relating to identifying objects called tokens. These technical problem addressed by this patents involves efficiently performing parallel searches over these attributes without compromising their accuracy. Additionally, current approaches require prior knowledge about how much space they occupy before selecting relevant attributes, making them slow down even further.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Indexing method and system of concurrent hash index data structure based on machine learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0061] The present invention provides an indexing method for a concurrent hash index data structure based on machine learning. The hash indexing method is divided into foreground threads and background threads. Data indexing operations (read and write requests) are performed in the foreground threads. The expansion operation is performed in the background thread; the invention uses locks and version numbers to control the read and write operations of the index concurrently; the invention uses the RCU technology to ensure the consistency of data when the hash expansion operation updates nodes; A two-stage copy operation is performed during capacity expansion, supports concurrent hash capacity expansion operations and index operations, reduces performance fluctuations caused by blocking of index operations during hash capacity expansion, and at the same time ensures the consistency of index data; the present invention uses a double-layered index structure , while maintaining the ...

Embodiment 2

[0101] Example 2 is a modification of Example 1

[0102] According to a method for indexing a concurrent hash index data structure based on machine learning provided by the present invention, the indexing method includes:

[0103] 1. Group node positioning step: take the key in the request as the input of the machine learning model of the root node, calculate the position of the group node where the key is located, and find the group node where the target key-value pair is located; then execute the data search step (2 ).

[0104] 2. Data search step: in the located group node, use the key in the request as the input of the machine learning model of the old node in the group node, calculate the specific location of the target key value, and then search in the old node data, if If the target key-value pair is found, the completion request step (4) is performed, otherwise, the new node search step (3) is performed in the group node.

[0105] 3. New node data search step: In the...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides an indexing method and system for a concurrent hash index data structure based on machine learning. The indexing method comprises the steps of M1, searching a group node where atarget key value pair is located for from a root node according to a key in a request; M2, in the group nodes obtained through positioning, calculating data positions according to a machine learningmodel of the group nodes; and M3, searching target data according to the data position, and carrying out corresponding operation according to the request type. Compared with an existing machine learning hash index, the system can retrain the machine learning model, hash expansion operation is carried out when the hash conflict probability exceeds a threshold value, the machine learning model is made to adapt to newly inserted data, and the high performance of the index is kept.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Owner SHANGHAI JIAO TONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap