Unlock instant, AI-driven research and patent intelligence for your innovation.

Social media data classification method and device based on n-side dfs subgraph lightweight unsupervised graph representation learning

A social media and data classification technology, applied in the field of social media data classification, can solve the problems of not being applicable to large-scale graph data sets, and the inability to comprehensively extract the structural information of graphs, and achieve the effect of improving classification accuracy

Active Publication Date: 2022-06-03
PLA STRATEGIC SUPPORT FORCE INFORMATION ENG UNIV PLA SSF IEU
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problem that the existing graph representation learning method cannot comprehensively extract the structural information of the graph and is not suitable for large-scale graph data sets, the present invention proposes a light-weight unsupervised graph representation learning method based on N-side DFS subgraphs Social media data classification method and device

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
  • Social media data classification method and device based on n-side dfs subgraph lightweight unsupervised graph representation learning
  • Social media data classification method and device based on n-side dfs subgraph lightweight unsupervised graph representation learning
  • Social media data classification method and device based on n-side dfs subgraph lightweight unsupervised graph representation learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] like figure 1 As shown, a social media data classification method based on N-edge DFS subgraph lightweight unsupervised graph representation learning, for the convenience of expression, referred to as Substructure2vec, including:

[0064] Step S101: Traverse the N-edge DFS subgraph structure in each graph of the atlas, and extract the N-edge DFS subgraph of each graph; including:

[0065] Step S1011: adopt the depth-first subgraph search algorithm, use the minimum DFSCode to uniquely identify the subgraph, and convert the subgraph into a text representation;

[0066] Step S1012: Perform N-edge DFS sub-graph extraction sequentially for each graph in the atlas, first generate an initial 1-edge sub-atlas, and then sequentially perform N-edge DFS sub-graph mining on the generated initial 1-edge sub-atlas:

[0067] Generate an initial edge set, and gradually expand from each initial edge. When expanding, first determine whether the current subgraph has the smallest DFSCode....

Embodiment 2

[0135] like Figure 8 As shown, a social media data classification device based on N-edge DFS subgraph lightweight unsupervised graph representation learning, including:

[0136] The subgraph extraction module 201 is used to traverse the N-edge DFS subgraph structure in each graph of the atlas, and extract the N-edge DFS subgraphs of each graph;

[0137] The subgraph collection module 202 is used to collect the extracted N-edge DFS subgraphs to form a subgraph set of each graph;

[0138] The graph vector representation module 203 is configured to input the sub-atlas into the neural network model for training, and obtain the vector representation of each graph.

[0139] Specifically, the subgraph extraction module 201 includes:

[0140] The minimum DFSCode identification submodule 2011 is used for adopting the depth-first subgraph search algorithm, using the minimum DFSCode to uniquely identify the subgraph, and converting the subgraph into a textual representation;

[0141]...

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 belongs to the technical field of graphics processing, and discloses a social media data classification method and device based on N-side DFS sub-graph lightweight unsupervised graph representation learning. The method includes: step 1: traversing N Edge DFS subgraph structure, extracting N-edge DFS subgraphs of each graph; Step 2: Collecting the extracted N-edge DFS subgraphs to form a sub-graph set of each graph; Step 3: Input the sub-graph sets The neural network model is used for training to obtain the vector representation of each graph; the device includes: a sub-graph extraction module; a sub-graph collection module; and a graph vector representation module. The present invention is applicable to large-scale graph data sets, and can extract sub-graph structures more comprehensively.

Description

technical field [0001] The invention belongs to the technical field of graphics processing, and in particular relates to a social media data classification method and device based on N-edge DFS subgraph lightweight unsupervised graph representation learning. Background technique [0002] In real life, a graph is a ubiquitous data structure that can simulate the connection between almost all things, such as the communication relationship between users in a communication network, the connection between computers in a network topology diagram, social networking The relationship between users and users in the network, etc. Usually each entity in the system is mapped to a node in the graph, and the connection between entities is mapped to an edge in the graph. The graph structure can easily reflect the connection between things in real life. Among them, the similarity calculation between graphs is a popular research area. Comparing the similarity between graphs has a wide range ...

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): G06F16/901G06N3/08G06K9/62
CPCG06F16/9024G06N3/088G06F18/241
Inventor 刘琰冯昊周资乔陈静刘楝赵艺张琦
Owner PLA STRATEGIC SUPPORT FORCE INFORMATION ENG UNIV PLA SSF IEU