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

Machine translation automatic optimization method and system based on self-supervision

A technology of automatic optimization and machine translation, applied in the field of machine translation, it can solve the problems such as the effect of the translation system plummeting, and achieve the effect of high parallelism

Active Publication Date: 2020-08-28
HARBIN INST OF TECH
View PDF6 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The common feature of the above-mentioned APE models is that they are all system-dependent, that is, for a specific machine translation system, the APE model that uses the bilingual sentence pairs generated by it for pre-training has a good performance on this system, but it is not good for other translation systems. Effect plummets

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
  • Machine translation automatic optimization method and system based on self-supervision
  • Machine translation automatic optimization method and system based on self-supervision
  • Machine translation automatic optimization method and system based on self-supervision

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

[0047] In the traditional post-editing task of machine translation, the labeling process of the post-editing dataset is relatively cumbersome, and the requirements for the quality of the labeling personnel are also high, which makes the post-editing task very expensive, so the scale of the post-editing dataset has always been small , which limits the development of traditional post-editing methods. The invention breaks through the limitations of traditional ...

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 relates to a machine translation automatic optimization method and system based on self-supervision, and belongs to the field of machine translation. The method comprises the following steps of 1, training a model by using a large-scale parallel corpus to enable the model to learn replacement operation; step 2, training the model by using artificially constructed pseudo data to enable the model to learn an insertion operation to obtain a trained model; and 3, performing candidate word prediction on each word and an inter-word gap in the to-be-optimized translation by using the trained model to finish post-editing operation of replacement and insertion. According to the method, the bilingual parallel corpus is used as the input of model pre-training, so that the model can obtain good optimization performance on different machine translation systems, and the model is a universal machine translation automatic post-editing model.

Description

technical field [0001] The invention relates to a method and system for automatic optimization of machine translation translation based on self-supervision, belonging to the field of machine translation. Background technique [0002] Machine translation is the technique of using a computer program to translate text or speech from one natural language to another. The quality of machine translation has been greatly improved in recent years, especially after the emergence of neural machine translation, but its overall quality is still far behind that of human translation. Machine translation needs to be optimized before it can be used as a qualified translation. . The optimization technology of machine translation translation can be roughly divided into two types: one is to introduce other translation results on the basis of the original translation for fusion optimization of the translation, such as introducing the translation results of multiple machine translation systems; ...

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): G06F40/166G06F40/232G06F40/58
CPCG06F40/166G06F40/232G06F40/58Y02T10/40
Inventor 杨沐昀徐冰王佳麒赵铁军朱聪慧曹海龙赵恩博唐煜
Owner HARBIN INST OF TECH
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com