Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Distributed dynamic graph management system oriented to large graph segmentation

A management system, distributed technology, applied in the field of distributed dynamic graph management system

Active Publication Date: 2016-12-07
SICHUAN UNIV
View PDF2 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the graph segmentation technology has been developed for a long time, how to efficiently segment large-scale graph data and establish new massive graph data processing models and algorithms is still a new research problem.

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
  • Distributed dynamic graph management system oriented to large graph segmentation
  • Distributed dynamic graph management system oriented to large graph segmentation
  • Distributed dynamic graph management system oriented to large graph segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0083] image 3 It is a functional block diagram of the distributed graph management system provided by Embodiment 1 of the present invention. Such as image 3 As shown, the system includes 1 Master host and k Slave hosts.

[0084] The master host is provided with a distribution logic module, a routing logic module, and a global information storage module.

[0085]In this embodiment, the allocation method of the allocation logic module is a greedy allocation method based on local minimum. Specifically, the allocation logic module is configured to: read a piece of vertex information in the large graph data stream, and Allocation is performed, and the allocation result is stored in the corresponding Slave host. In this embodiment, the distribution logic module is specifically configured to: read a piece of vertex information in the large graph data stream; find the minimum degree vertex v among all the vertices of the vertex information min ; will be compared with vertex v ...

Embodiment 2

[0097] Embodiment 2 also provides a distributed graph management system, the schematic diagram of which is the same as that of Embodiment 1, and the only difference lies in the allocation method adopted by the allocation logic module.

[0098] In particular, in this embodiment, the allocation logic module adopts a greedy allocation method based on local maximum degree. Specifically, the allocation logic module is configured to: read a piece of vertex information in the large graph data stream; Find the maximum degree vertex v in max ; will be compared with v max Relevant edge e is taken out; edge allocation is performed according to the following third allocation formula:

[0099] Index2 hc = arg max i∈{1,…,k} {|P t (i)∩Γ(v max )|w(t,i)},

[0100] Among them, Index2 hc is the assigned partition, w(t,i) is the weight penalty function: ω is the preset penalty function parameter, P t (i) is the vertex set of i partition at time t, Γ(v min ) is the vertex v max The set...

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 relates to the field of graph data mining, and provides a distributed graph management system. The system comprises at least one Master host and a plurality of Slave hosts, wherein the Master host is provided with a distribution logic module, and each Slave host is independently provided with a dynamic balance module. By use of the distributed graph management system which is put forward by the invention, the introduction, the segmentation and the access of the graph are realized, the visualization of a segmentation result is realized, and a user can observe a graph segmentation result at a glance and know whether segmentation is good or bad. In addition, the dynamic balance modules in the system determine the partition of the main vertex of each vertex according to a vertex reading behavior in the graph and then determine a copying situation of a shadow vertex according to a vertex activity behavior, network loads are lightened, and storage space occupied by vertex information is saved.

Description

technical field [0001] The invention belongs to the field of graph data mining, in particular to a distributed dynamic graph management system oriented to large graph segmentation. Background technique [0002] In the face of large-scale graph data, the conventional processing method is to place it on distributed multi-machine nodes for parallel processing, and the solution to the graph segmentation problem is the premise of adopting this scheme. As early as the 1970s, the problem of graph segmentation has become a hot topic in the field of graph theory research. After more than 40 years of development, traditional graph segmentation algorithms have approached maturity. Only by dividing the entire graph can it be analyzed on the distributed graph computing platform. However, the quality of the graph segmentation algorithm will directly affect the performance of the distributed computing platform. The main reasons include the load balancing problem of each machine and the t...

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
IPC IPC(8): G06F17/30G06T7/00
CPCG06F16/51G06F16/56
Inventor 李川王昂
Owner SICHUAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com