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

Entity knowledge automatic extraction method, computer device and computer readable medium

An automatic extraction and entity recognition technology, applied in the field of text processing, can solve problems such as difficult optimization and low accuracy, and achieve the effect of improving feature extraction capabilities

Pending Publication Date: 2022-04-15
YOZOSOFT
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The current automatic entity knowledge extraction method still has the disadvantages of low accuracy and difficulty in optimization.

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
  • Entity knowledge automatic extraction method, computer device and computer readable medium
  • Entity knowledge automatic extraction method, computer device and computer readable medium
  • Entity knowledge automatic extraction method, computer device and computer readable medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0011] In order to further illustrate the technical means and efficacy of the present invention to achieve the object of the present invention, the specific embodiments, structures, features, and efficacy of the present invention will be described in detail below with reference to the drawings and preferred embodiments.

[0012] Here is the extraction of entity knowledge in the official document as an example. Firstly, the following entity categories are summarized and defined by the field of document text based on statistical methods and rule-based methods:

[0013]

[0014] According to the analysis of the data set, in the sentence, the entity category defined in this paper has a mutually associated role. If the sentence is held in Beijing, the 19th National Fifth Plenary Session is held in Beijing, ** issued an important speech. "You can extract the entity" "Beijing", "Beijing", "Beijing", among which the entity "19th Fifth Plenary Session" is "event activities", the entity "...

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 provides an automatic entity knowledge extraction method, a computer device and a computer readable medium. The entity knowledge automatic extraction method comprises the steps that an input text H0 is input into the first K layers of a BERT model composed of N layers to be processed, and a context representation vector HK is output in the Kth layer; the HK is input into the remaining N-K layers of the BERT model for the first time due to a first task to be processed, at the moment when a context representation vector is output on the Nth layer, each layer of the remaining N-K layers processes input on the basis of a first mask matrix, and a first layer entity is obtained on the basis of first layer entity recognition; and inputting the HK into the residual N-K layers of the BERT model again for processing due to a second task, processing the input of each layer of the residual N-K layers on the basis of a second mask matrix at the moment when the context representation vector is output at the Nth layer, and performing second layer entity identification to obtain a second layer entity. Therefore, the feature extraction capability of the BERT model on the entity knowledge can be improved.

Description

Technical field [0001] The present invention relates to the field of text processing, and more particularly to an entity knowledge automatic extraction method and a computer device, a computer readable medium. Background technique [0002] As an important task in natural language processing, information extraction is a text processing technology for extracting meaningful entities, attributes, relationships, events such as original unstructured natural language text. In official document writing, its physical knowledge has an important role, can assist official writing, such as content review. The current entity knowledge automatic extraction method still has a low accuracy and difficulty optimization. Therefore, it is necessary to propose an improved entity knowledge automatic extraction method. Inventive content [0003] It is an object of the present invention to provide an entity knowledge automatic extraction method and a computer device, a computer readable medium, which ca...

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): G06F16/36G06F40/295
CPCG06F16/36G06F40/295Y02D10/00
Inventor 夏振涛谈辉李艳朱立烨石雁
Owner YOZOSOFT
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