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

Keyword extraction method based on depth map neural network

A neural network and keyword technology, applied in the field of keyword extraction based on deep graph neural network, to achieve the effect of simple calculation process and improved effect

Pending Publication Date: 2020-09-15
NANCHANG UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] However, compared with unsupervised graph-based ranking methods, existing end-to-end methods only treat documents as word sequences and do not benefit from the global information of the graph structure.

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
  • Keyword extraction method based on depth map neural network
  • Keyword extraction method based on depth map neural network
  • Keyword extraction method based on depth map neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0039] Further detailed explanation through specific implementation mode below:

[0040] The embodiment is basically as attached figure 1 And attached figure 2 Shown:

[0041] The node set W={w for a given source document 1 ,w 2 ,...,w n}, waiting keyword set K={k 1 , k 2 ,...,k m};

[0042] A method for extracting keywords based on a deep graph neural network, comprising the following steps:

[0043] S1, organize the source documents, build an adjacency matrix consistent with the shape of the source documents, and define with diagram The corresponding adjacency matrix is and word w i ∈W to word w j The edge weight of ∈W is

[0044]

[0045]

[0046] where P(w i ) for the word w i the position of p i A set of , the edge weight shows the degree of association between two nodes;

[0047] S2, the directed graph corresponding to the document to be extracted by keywords: the set of keywords to be extracted is represented as a graph count The corres...

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 field of computer document retrieval, and particularly relates to a keyword extraction method based on a depth map neural network, which comprises the following steps of:performing matrix weighting on a document to form a directed graph, and combining the most suitable keywords from adjacent vocabularies by using a graph convolution encoder and a graph convolution decoder; in the period, in order to ensure the stability of the data, carrying out a regularization mode to ensure the stability of the data. Therefore, the scheme is improved on the basis of the end-to-end keyword extraction method of the graph neural network, and the keyword extraction efficiency is greatly improved.

Description

technical field [0001] The invention belongs to the field of computer document retrieval, and in particular relates to a keyword extraction method based on a deep graph neural network. Background technique [0002] The keywords of an article are usually several words or phrases, which serve as a summary of the main content of the document. The use of keywords enables people to quickly understand the content of the document and grasp the topic of the article. Nowadays, key extraction technology is widely used in information retrieval, information management and other fields. [0003] Traditional keyword extraction methods are all unsupervised methods. Unsupervised methods usually use some heuristics to identify candidate keywords first, and then rank the candidate keywords according to their importance scores. Along this direction, the state-of-the-art algorithms are graph-based sorting methods. However, such methods are completely unsupervised, they rely heavily on manua...

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): G06F40/30G06N3/04G06N3/08G06F16/33
CPCG06F40/30G06N3/088G06F16/3344G06N3/048G06N3/045
Inventor 段文影
Owner NANCHANG UNIV