Key information extraction method, device and storage medium based on finely-labeled text

A key information and text technology, applied in devices and storage media, based on a method for extracting key information from finely labeled text, in the field of systems, to achieve the effect of reducing the cost of labeling

Active Publication Date: 2020-08-04
ONE CONNECT SMART TECH CO LTD SHENZHEN
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention provides a key information extraction method, system, electronic device and computer-readable storage medium based on finely labeled text, which mainly solves the problem of automatically labeling text segments through the BERT pre-training model and key information extraction model

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
  • Key information extraction method, device and storage medium based on finely-labeled text
  • Key information extraction method, device and storage medium based on finely-labeled text
  • Key information extraction method, device and storage medium based on finely-labeled text

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

[0044] The reading comprehension model is conditioned on the question (or the input is a long text and a question), and the correct answer is marked in the text (the output is the corresponding fragment of the answer in the text); the existing reading comprehension model is to mark in advance Good key sentences / paragraphs are used as the input of the model, while artificially labeling key segments has the disadvantage of being one-sided.

[0045] The present invention uses the BERT (Bidirectional Encoder Representation from Transformer, bidirectional attention neural network model) pre-training model to pre-train the text data, and then inputs the key information extraction model to output the key information in the text data as an answer.

[0046] Specifically, unlike the traditional reading comprehension model that ...

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 present invention relates to the technical field of key information extraction, and proposes a key information extraction method, device, and storage medium based on finely tagged text, wherein the method includes: S110, pre-training text data through a BERT pre-training model to obtain word vectors, and The obtained word vectors are combined into matrix text data; S120, input the matrix text data into a key information extraction model, the key information extraction model is trained using the CMRC data set, and obtains key words according to the matrix text data information; S130, sort the obtained key information according to the preset sorting rules, and output the key information conforming to the set selection rules. The invention solves the problem of automatically labeling text segments, greatly reduces the labeling cost, and achieves the technical effect of providing strong support for downstream tasks.

Description

technical field [0001] The present invention relates to the technical field of key information extraction, in particular to a key information extraction method, system, device and storage medium based on finely marked text. Background technique [0002] Machine reading comprehension refers to allowing machines to answer content-related questions by reading texts. At present, it is more and more widely used to input the questions to be answered and related reading materials into the trained reading comprehension model for artificial intelligence reading comprehension. However, the existing reading comprehension model is based on the condition of the question, marking the correct answer in the text, and using the key sentences / paragraphs marked in advance as the input of the model. The method of outsourcing manual labeling of key sentences / paragraphs will greatly increase the cost of time and money. [0003] In order to achieve the purpose of automatically labeling the fragm...

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): G06F16/33G06F16/338G06N3/04G06N3/08
CPCG06F16/334G06F16/338G06N3/088G06N3/04G06N3/048
Inventor 曹辰捷徐国强
Owner ONE CONNECT SMART TECH CO LTD SHENZHEN
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products