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

Translation check method and system

An inspection method and inspection system technology, applied in the field of translation inspection methods and systems, can solve problems such as inability to judge whether they are correct or not.

Active Publication Date: 2015-04-15
新方正控股发展有限责任公司 +2
View PDF5 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The technical problem to be solved by the present invention is that the translation detection method in the prior art depends to a large extent on bilingual experts, and at the same time cannot judge whether it is correct or not, but can only judge the quality of the translation, so as to provide a learning corpus, A translation checking method that trains a bigram model and automatically filters out "incorrect" or "wrong" phrase translations in a large number of related translations

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
  • Translation check method and system
  • Translation check method and system
  • Translation check method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0115] This embodiment provides a translation checking method, the method flow chart is as follows figure 1 As shown, it includes initialization processing and checking evaluation processing.

[0116] The initialization process is based on a batch of corpus, the corpus is the corpus of a certain field, and the binary grammar model is obtained by counting the probability information of the binary segmentation entries of the corpus in the field, as the inspection scoring model, for the inspection and judgment processing Provide a basis for scoring the translation.

[0117] The inspection and evaluation process uses the inspection scoring model to score domain translations, compares the translation score with a preset judgment threshold, and judges a translation with a score smaller than the preset judgment threshold as "wrong", otherwise judges it as "correct" .

[0118] The steps of the initialization process are as follows:

[0119] S11: Obtain a batch of text sets D in a c...

Embodiment 2

[0167] In this embodiment, except that step S13 is different from embodiment 1, other steps are the same as embodiment 1. In the step S13, w i at the beginning w j The probability of occurrence f p (w i ,w j ) method is:

[0168] to get all w from list L i The first binary segmentation entry, the second character w in the binary segmentation entry j Join the string S;

[0169] Store each character in the string S into a set T;

[0170] Count the length n of the string S, for each character w in the set T j , count the character w j The number of occurrences m in the string S;

[0171] then take w i at the beginning w j The probability of occurrence f p (w i ,w j )for

[0172] f p (w i ,w j )=m / n

[0173] Among them, the initial value of m, n is zero.

[0174] First count the length n of the string S and each character w in the set T j , count the character w j The number of occurrences in the string S is m, and then the ratio of n and m is used as w i at...

Embodiment 3

[0176] In this embodiment, except that step S23 is different from embodiment 1, other steps are the same as embodiment 1, and the method for scoring and evaluating the translation in step S23 is as follows:

[0177] Score 译 =avg{Score i ,i=1,2,...,n-1}

[0178] Among them, Score i is the score of a binary segmentation entry, Score i =f p (w i ,w j ), f p (w i ,w j ) is the binary segmentation entry in the translation (w i w j ) corresponds to the value in the model.

[0179] In the translation inspection method provided in this embodiment, the method of scoring the translation adopts the average value of each binary item in the translation, which can effectively avoid erroneous scoring caused by some binary items not included in the model.

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 provides a method and device for checking a translation, comprising initialization and determination processing. Said initialization processing: training a bigram model on the basis of a batch of corpora, the bigram model serving as a scoring model and providing a basis for scoring a translation during said determination processing. Said determination processing: scoring a translation by means of said scoring model, comparing the translation score to a preset determination threshold, and determining a translation having a score of less than said preset threshold value to be "incorrect", otherwise, determining the translation to be "correct". The described technical solution effectively avoids the problem in the prior art of translation testing methods relying heavily on bilingual specialists or high-quality manual translations for reference, while also being unable to determine whether a translation is correct or not and only being able to judge the extent to which a translation is good or bad.

Description

technical field [0001] The invention relates to a translation checking method and system thereof, in particular to a translation checking method and system based on a binary grammar model, and belongs to the technical field of electronic digital data processing methods. Background technique [0002] In recent years, the application of machine translation (Machine Translation) has become more and more extensive, and the quality requirements for machine translation translations are getting higher and higher. In the field of translation, even the best translators can hardly meet the highest standards of "credibility, expressiveness, and elegance" required by the translation industry. Therefore, the evaluation of machine translation has become an important and difficult subject. [0003] Since one needs to know at least two languages ​​to evaluate the translation quality, the translation quality evaluation has become a very difficult intellectual activity. Therefore, the evaluat...

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): G06F17/27G06F17/28
CPCG06F40/51
Inventor 叶茂王元龙金立峰汤帜徐剑波
Owner 新方正控股发展有限责任公司
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