Parallel label propagation-based heterogeneous network community discovery method

A heterogeneous network and label propagation technology, applied in instruments, data processing applications, computing, etc., can solve the problems of slow convergence, affect the overall performance of the community structure, and reduce availability, and achieve the effect of integration

Inactive Publication Date: 2016-06-01
NORTHWESTERN POLYTECHNICAL UNIV
View PDF0 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, community discovery methods based on traditional LPA cannot discover overlapping community structures
COPRA has extended LPA, but there are still three deficiencies: First, although COPRA can discover overlapping community structures, it cannot be used for community discovery for heterogeneous interactive networks; second, COPRA’s iteration end judgment condition It is more harsh, and its convergence speed is slower for some network structures; third, COPRA will lead to th...

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
  • Parallel label propagation-based heterogeneous network community discovery method
  • Parallel label propagation-based heterogeneous network community discovery method
  • Parallel label propagation-based heterogeneous network community discovery method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] Example 1: Comparison of Different Membership Thresholds

[0064] For the community discovery method based on label propagation, the membership threshold θ mainly affects the number of communities v that a node can belong to, that is, θ=1 / v.

[0065] The experimental parameters are set as follows: the value range of the parameter v is [1,20], and the asynchronous update mechanism is selected. The experimental results are shown in the figure.

[0066] According to the experimental results, the following conclusions are drawn:

[0067] The larger the value of the parameter v, that is, the smaller the value of the membership threshold θ, the smaller the number of corresponding communities. The reason is that a smaller membership threshold makes the label more capable of spreading, and it is easier to form a larger community during the iterative process, resulting in a decrease in the number of communities.

[0068] The value of the parameter v has a great influence on t...

Embodiment 2

[0071] Example 2: Comparison of Different Tag Update Mechanisms

[0072] The experimental parameters are set as follows: the value range of the parameter v is [1,20], and the experiment is carried out by using two kinds of label update mechanisms, synchronous and asynchronous, respectively. Experimental results such as Figure 10 shown.

[0073] According to the experimental results, the following conclusions are drawn:

[0074] Under the synchronous and asynchronous tag update mechanisms, HLPA has no significant difference in the two performance indicators of community number and community coverage, and there is only a slight difference when the value of the parameter v is small.

[0075] Under the synchronous and asynchronous tag update mechanisms, HLPA has significant differences in the two performance indicators of modularity and community overlap. When the value of the parameter v is small, the performances of the two label update mechanisms are relatively close; as v ...

Embodiment 3

[0077] Embodiment 3: Comparison of different label attenuation factors

[0078] The experimental parameters are set as follows: v is 3 and 5 respectively, the value range of δ is (0, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40), and the asynchronous label update mechanism is adopted. The related experimental results are shown in the figure.

[0079] According to the experimental results, the following conclusions are drawn:

[0080] With the increase of label decay factor δ, the largest community found by HLPA shows a decreasing trend. When δ increases from 0.00 to about 0.25, the largest community continues to decrease rapidly; and when the value of δ exceeds a certain threshold (about 0.25), the largest community found remains basically stable. This conclusion holds true when the value of the parameter v changes.

[0081] As the label attenuation factor δ increases, the modularity achieved by HLPA also shows a gradual decrease trend. Specifically, when δ increases from 0.00 t...

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 discloses a parallel label propagation-based heterogeneous network community discovery method. The method is characterized by comprising a label initialization step, a label cyclic refreshing step and a community building step. According to the label cyclic refreshing step, based on parallel label propagation, node labels are allowed to be propagated in multiple sub networks of the heterogeneous network in a relatively independent and parallel mode; and the node labels are refreshed through combining parallel propagation results of the multiple sub networks. Compared with a linear combination method LinearComb, the parallel label propagation-based combination method can use heterogeneous interactive information between nodes more effectively, and HLPA is more suitable for heterogeneous network community discovery.

Description

technical field [0001] The invention relates to the technical field of network structure analysis; in particular, it relates to a community discovery method of a heterogeneous network. Background technique [0002] The heterogeneity of social interaction behavior is one of the salient features of mobile social networks, which manifests itself in the presence of both online interactions between users in the virtual space and offline interactions between users in the physical space. Specifically, on the one hand, the online interaction between users forms an online social network, and on the other hand, the offline interaction between users forms an offline social network. Therefore, how to realize the effective integration of online social interaction and offline social interaction, so as to accurately describe the heterogeneous interaction behavior among users, has become the primary challenge to be addressed in the research of heterogeneous mobile social network community d...

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): G06Q50/00
CPCG06Q50/01
Inventor 王柱周兴社於志文郭斌
Owner NORTHWESTERN POLYTECHNICAL 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