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

BERT pre-training model-based text abstract generation method

A pre-training and model technology, applied in neural learning methods, biological neural network models, instruments, etc., can solve problems such as obstacles in the knowledge acquisition process, and achieve the effect of improving text quality, excellent grades, and improving quality

Active Publication Date: 2021-07-16
CHONGQING UNIV OF POSTS & TELECOMM +1
View PDF10 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This creates a huge barrier to the knowledge acquisition process

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
  • BERT pre-training model-based text abstract generation method
  • BERT pre-training model-based text abstract generation method
  • BERT pre-training model-based text abstract generation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0061] The technical solutions in the embodiments of the present invention will be described clearly and in detail below with reference to the drawings in the embodiments of the present invention. The described embodiments are only some of the embodiments of the invention.

[0062] The technical scheme that the present invention solves the problems of the technologies described above is:

[0063] In this embodiment, a method for generating abstracts based on the BERT pre-trained model is performed in the following steps.

[0064] Step 1: Preprocessing the text data set (removing special characters, converting animated expressions, replacing date tags, hyperlink URLs, replacing numbers, and replacing English tags);

[0065] (1) Special characters: Remove special characters, mainly including punctuation marks and commonly used stop particles and transition words, including: "「,",￥,..."Ah, hey, and;

[0066] (2) Convert the label content in brackets into words, such as [happy],...

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 requests to protect a text abstract generation method based on a BERT pre-training model. The method comprises the following steps: preprocessing a Chinese short text data set; obtaining the global information by utilizing the BERT bidirectional coding characteristic, and sending the data into a BERT pre-training model for training; inputting an original text of which the abstract actually needs to be obtained into the BERT pre-training model, and performing training by using the trained parameters to obtain an optimal word vector; sending the obtained high-quality word vector to the improved LeakGAN model; and training the text in the improved LeakGAN, and finally abstract output is obtained. According to the invention, the generator can generate a more accurate abstract, and the accuracy and fluency of the abstract are improved.

Description

technical field [0001] The invention belongs to the field of natural language processing text generation, and relates to a method for generating abstracts based on a BERT pre-training model. Background technique [0002] With the progress of the times and the development of information technology, the Internet has become an increasingly important social, entertainment and even work platform in human life, and it is the main channel for people to obtain various knowledge resources. The Internet has increasingly become an essential part of people's lives and has penetrated into every aspect of life. [0003] However, while the Internet provides convenient and fast services for human beings, it also brings about the inevitable problem of information overload. With the rapid increase in the amount of information data, the form of information is also showing a trend of diversification, mainly including text, sound, image and so on. As the most basic form of information on the I...

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 Applications(China)
IPC IPC(8): G06F40/284G06F40/30G06F40/253G06N3/04G06N3/08
CPCG06F40/284G06F40/30G06F40/253G06N3/08G06N3/047G06N3/044G06N3/045
Inventor 文凯周玲玉杨航王宗文
Owner CHONGQING UNIV OF POSTS & TELECOMM