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

A network structure deanonymization system and method based on matrix decomposition

A technology of network structure and matrix decomposition, applied in transmission systems, digital transmission systems, data exchange networks, etc., can solve problems such as limited accuracy and lack of global structural information

Active Publication Date: 2021-11-30
CHONGQING UNIV OF POSTS & TELECOMM
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most of the existing network structure deanonymization methods are only based on the anonymous network, and the accuracy is limited, and most of the existing network structure deanonymization methods are only based on the local structure of network nodes to infer sensitive relations, and the global structure information has not been widely used. use

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
  • A network structure deanonymization system and method based on matrix decomposition
  • A network structure deanonymization system and method based on matrix decomposition
  • A network structure deanonymization system and method based on matrix decomposition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0044] In the present embodiment, the present invention proposes a matrix-based network structure to anonymization system, including a dynamic network acquisition data module, a low rank sparse model training module, a construction module, an optimization module of a non-negative matrix decomposition training model, and an optimization module. in:

[0045] Dynamic Network-Static Network Conversion Module for accessing static networks and static networks in a static network and a static network adjacency matrix according to social network data of real-time dynamic network;

[0046] Low rank sparse model training module, used to constrain the local feature structure of each static network while removing noise of each static network;

[0047] The construction module of non-negative matrix decomposition training model is used to capture the intrinsic rules of the network, and portray the potential features of the network;

[0048] Optimization modules for reducing the randomness of th...

Embodiment 2

[0065] This embodiment proposes a matrix-based network structure to go anonymization method, such as figure 1 , Includes the following steps:

[0066] S1, obtain data sources, data sources include the topology structure between the various networks and the topology structure between the relationship between users in the network; figure 2 The dynamic network shown, for example, the friend relationship topology on Tencent Weibo may be similar to the friend relationship topology on Facebook. The two networks have a part of the common user, the relationship between the network, and the network between users. The topology of the relationship is the topological structure of the relationship between users in a network;

[0067] S2, such as figure 2 , The data set of the dynamic social network is divided into T-static network data set according to the time, and the adjacent matrix corresponding to each static network structure information is obtained;

[0068] S3, the adjacent matrix inpu...

Embodiment 3

[0085] As an alternative embodiment, the solving process of the target matrix includes:

[0086]

[0087] Among them, Rank (S t ), Λ represent damping coefficient; e t A noise error matrix of the TF; A t Represents the neighboring matrix of the tip of the tip; t Represents the target matrix of the Type T static network; || E t || 0 Indicates the sparse noise constraint; s represents the target matrix after training after a low rare model; E represents the noise error matrix of the network.

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 network structure deanonymization, in particular to a network structure deanonymization system and method based on matrix decomposition. The topological structure of the relationship between them; the dynamic social network data set is divided into T static network data sets according to time, and the adjacency matrix corresponding to each static network structure information is obtained; the adjacency matrix input is trained according to the low-rank representation model, Calculate the target matrix; input the target matrix into the improved NMF model for training and prediction, and obtain the de-anonymized matrix of the target matrix; the present invention performs low-rank training on the network data, combined with the NMF prediction model, the The structural information of the network and the time series evolution are considered together to improve the prediction accuracy.

Description

Technical field [0001] The present invention relates to the field of network structures to anonymization, and more particularly to a matrix-based network structure to anonymize system and method. Background technique [0002] Social Network is now widely used by third-party consumers (such as researchers and advertisers) to understand user characteristics and behavior. Typically, private information or sensitive information contained in the collected data set is anonymously before publishing network data to prevent personal privacy from being damaged. In order to quantify the guarantee level of the privacy protection mechanism and mitigate the user's concerns, the network to anonymization based on sensitive information is especially important. [0003] In social networks, there is three important privacy risks in social network data distribution due to private networks: content leakage risk, identity leakage risk and link leakage risk. At present, many anonymous methods have been...

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
Patent Type & Authority Patents(China)
IPC IPC(8): H04L12/58H04L12/24G06F17/16
CPCH04L41/147H04L41/145H04L41/142G06F17/16H04L51/52
Inventor 陈幸吴涛先兴平明冠男
Owner CHONGQING UNIV OF POSTS & TELECOMM
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