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

Text filling method and device

A filling method and text technology, applied in neural learning methods, instruments, biological neural network models, etc., can solve problems such as poor fluency and lack of semantic coherence in text, so as to reduce semantic loss and improve semantic coherence and fluency. , the effect of simple network structure

Inactive Publication Date: 2020-12-11
GUILIN UNIV OF ELECTRONIC TECH
View PDF1 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The current text filling is mainly based on the recurrent neural network, but the existing methods have the problems of lack of semantic coherence and poor fluency of the filled text

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
  • Text filling method and device
  • Text filling method and device
  • Text filling method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0017] The principles and features of the present invention are described below in conjunction with the accompanying drawings, and the examples given are only used to explain the present invention, and are not intended to limit the scope of the present invention.

[0018] figure 1 It is a schematic flowchart of a text filling method provided by an embodiment of the present invention.

[0019] Such as figure 1 As shown, a text filling method includes the following steps:

[0020] S1: Preprocess the input samples to obtain missing samples with multiple blank positions;

[0021] S2: Encoding the missing sample to obtain a coding layer semantic vector;

[0022] S3: Based on the attention mechanism algorithm, a plurality of blank positions in the semantic vector of the coding layer are filled to obtain a semantic vector of the prediction layer and a plurality of filler words used to fill the blank positions, and each of the filling words in the blank positions is respectively fi...

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 a text filling method and device, and the method comprises the steps: preprocessing an input sample to acquire a missing sample with a plurality of blank positions; encoding themissing sample to obtain an encoding layer semantic vector; filling a plurality of blank positions in the coding layer semantic vector based on an attention mechanism algorithm to obtain a predictionlayer semantic vector and a plurality of filling words for filling the blank positions, and respectively marking each word filled to the blank positions to obtain a plurality of marked filling words;and performing loss analysis on the coding layer semantic vector and the prediction layer semantic vector according to the plurality of mark filling words to obtain a filling text. Semantic loss is reduced through iteration, the text long-distance relationship can be captured, the network structure is simple, and compared with the prior art, the semantic coherence and fluency of the filled text can be improved.

Description

technical field [0001] The present invention mainly relates to the technical field of language processing, in particular to a text filling method and device. Background technique [0002] Text filling, that is, automatically filling in the missing part of the text, the purpose of text filling is to use the context information of the missing part to fill in the missing part of the sentence or paragraph, which can be used in many real-world natural language generation scenarios, such as filling blank images, vocabulary constraints Sentence generation, ancient text recovery, acrostic generation. The current text filling is mainly based on the recurrent neural network, but the existing methods have the problems of lack of semantic coherence and poor fluency of the filled text. Contents of the invention [0003] The technical problem to be solved by the present invention is to provide a text filling method and device for the deficiencies of the prior art. [0004] The technic...

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): G06F40/284G06F40/30G06N3/04G06N3/08
CPCG06F40/284G06F40/30G06N3/049G06N3/08G06N3/048G06N3/045
Inventor 蔡晓东田文靖
Owner GUILIN UNIV OF ELECTRONIC TECH
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com