CN109241543A - The preconditioning technique of consistency translationese - Google Patents
The preconditioning technique of consistency translationese Download PDFInfo
- Publication number
- CN109241543A CN109241543A CN201811079665.0A CN201811079665A CN109241543A CN 109241543 A CN109241543 A CN 109241543A CN 201811079665 A CN201811079665 A CN 201811079665A CN 109241543 A CN109241543 A CN 109241543A
- Authority
- CN
- China
- Prior art keywords
- word
- module
- field
- data library
- contrasting data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/40—Processing or translation of natural language
- G06F40/58—Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The invention proposes a kind of preconditioning techniques of consistency translationese, material progress terminological analysis is translated for treating in translation process, the translationese that wherein may need especially to treat is extracted first, after classifying to it, it is compareed in sample terminology bank corresponding with the material to be translated, to accurately filter out the term for needing the being consistent property in translation process.Using this method, those special words for corresponding to Material Field to be translated can not only be extracted, the popular word for having particular meaning in special linguistic context can also be extracted, the special word and popular word may make up " the consistency translationese " for needing the being consistent property in translation process.
Description
Technical field
The invention belongs to translation technology field more particularly to the preconditioning techniques of consistency translationese;Particularly, it is related to
A kind of classification of translationese, control, screening technique and system.
Background technique
Heavy due to translation duties in current translation, usual one female translation duties needs are split into
Multiple and different subtasks, translator or translation team to distribute to different translate, and obtain respective subtask
Translation result, then will splice after multiple subtask translation results carry out necessary processing, and be combined into and final correspond to female translation
The translation result of task.
Why to carry out " necessary processing " rather than directly combination can just obtain final translation result, one of them
Important reason is that, even for same a translation material, the translation result that different translators/translation team obtains is all
Will not be identical, especially for the translation of certain specific terms.For example, same in different subtasks for appearing in
The translation of a specific term, different translators/translation team can all provide different translation results;Even for certain normal
The term seen, due to context of co-text difference, it should have different meanings in different contexts, but just same at last
A translator/translation team, should have some the generic term of different meanings in different context, since thinking is fixed
The influence of gesture can also provide identical translation result.If directly combined without necessary processing, it will cause final
Translation result is obviously inaccurate, for example, the translation result of the same specific term is in the obvious disunity of different piece;It is certain common
Term should have different meanings in different context, but just the same as the result is shown.To avoid similar phenomenon, currently turn over
During translating, usually after each translator/translation team all completes respective task, above-mentioned ask is handled concentrating in together
Topic, to reach different translators/translation team result consistency.
In the present invention, defining such term is " consistency term ", means that each translator/translation team should be
Keep translation criteria consistent in translation process.
However, most of current processing technique is based on artificial check and correction technology, or automatically extracting based on high frequency words
Technology;The former efficiency is lower, has dragged slowly whole translation efficiency significantly;For the latter, either extracted before translation is completed high
Frequency word extracts high frequency words after the completion of still translating, the biggest problems are that, due to the limitation of participle technique, when female translation duties
When quantity is larger, a large amount of high frequency words are often extracted, for example, every ten thousand word proposes the high frequency words of K word or more, this rank
High frequency words cause subsequent processing task heavier;In order to reduce high frequency words number of levels, generally require to mention " high frequency "
Standard, but in this way, will missing inspection special word, and have in the special linguistic context of part the conventional word of particular meaning
Converge, after all only using the frequency as standard, the quantity of such vocabulary be it is less, cannot still guarantee the accuracy of translation result.
That is, using current method, the accurate extraction for efficiently realizing " consistency term " of having no idea, to lead
Cause final translation result that can not be consistent the translation of " consistency term ", to bring to translational action serious
Obstacle.
Therefore, how to guarantee the accuracy and consistency of the translation result of related translationese in translation process, always
Since be technical problem that translation person faces.
Summary of the invention
To solve the above problems, the invention proposes a kind of preconditioning techniques of consistency translationese, for translating
Treat in the process translate material carry out terminological analysis, extract the translationese that wherein may need especially to treat first, to its into
It after row classification, is compareed in sample terminology bank corresponding with the material to be translated, to accurately filter out needs in translation process
In being consistent property term.Using this method, those special words for corresponding to Material Field to be translated can be not only extracted,
The popular word for having particular meaning in special linguistic context, the special word and popular word can also be extracted, it can structure
At " the consistency translationese " for needing the being consistent property in translation process.Accurately extracting " the consistency translation art
After language ", different translators/translation team can reach an agreement before translation starts on translation criteria, thus complete
Behind respective subtask, corresponding sub- translation result is directly combined, so that the final translation result of female translation duties is quickly obtained,
And maintain the consistency of translation result.
In the first aspect of the invention, a kind of extracting method of consistency translationese is provided, this method includes such as
Lower step:
1) from wait translate the corpus for randomly selecting the first predetermined ratio in material, it is carried out at the participle based on natural language processing
Reason;
2) based on the word segmentation processing as a result, determining at least one field term contrasting data library;
It is characterized in that, after determining at least one field term contrasting data library,
Further include following steps:
(I) it treats and translates material progress word segmentation processing, material to be translated is converted to translation word repertorie;
(II) classify to described each word to translation word repertorie, obtain different word's kinds;
(III) it is based on field term contrasting data library, determines the reference standard of the word of the different classifications;
(IV) based on the reference standard and field term contrasting data library, the word of the different classifications is sieved
Choosing, to obtain consistency translationese.
Point as a further improvement of the present invention is different from the prior art, and there is no directly treat to translate material by the present invention
All corpus directly carry out word segmentation processing, but use the method randomly selected, this greatly reduces workload;
Certainly, the premise for reducing workload is to ensure that the representativeness and accuracy of word segmentation processing result, if having no purpose list
Pure randomly selects, and does not ensure that accuracy.
At least one field term contrasting data library of the determination, comprising:
(21) corpus for randomly selecting the second predetermined ratio backward from the beginning part of material to be translated carries out word segmentation processing to it,
Field term contrasting data library is determined based on word segmentation result;
And/or;
(22) corpus for randomly selecting third predetermined ratio forward from the tail portion of material to be translated carries out word segmentation processing to it,
Field term contrasting data library is determined based on word segmentation result.
Point as a further improvement of the present invention, the of the invention purpose of randomly selecting is very strong, select must to
The beginning part for translating material is randomly selected backward, and/or, it is randomly selected forward from the tail portion of material to be translated, such extraction
Mode is one of innovative point of the invention;Its specific improvement effect will be shown in specific embodiment part;
Wherein, field term contrasting data library, to establish and being constantly updated in translation process related to field in advance
Specific term database.
The specific term database is established as follows:
(a) it is collected for more than the corpus document of some specific area of predetermined quantity;
(b) word segmentation processing is carried out to the corpus document, corpus document is converted into lexicon;
(c) semantic analysis is carried out to the lexicon, for each vocabulary, the dimension such as comprehensive word frequency, distribution, part of speech, capital and small letter
Degree is given a mark, if score value meets first condition, which is included in specific term database.
Particularly, inventor is especially focused on the distribution dimension of vocabulary during establishing the specific term database
With capital and small letter dimension.As one of important improvement of the invention, the marking considers emphatically the capital and small letter and distributing position of vocabulary
Associativity.This is one of important innovations point of the invention, and importance will be embodied in subsequent specific embodiment part;
According to the method described above, the present invention can establish the field term contrasting data library in multiple and different fields in advance.
Next, it needs to be determined that being screened using which kind of field term contrasting data library to word segmentation processing result.
Since the present invention has extracted the corpus for having representative the beginning part and tail portion of material to be translated, to its into
After row word segmentation processing, word segmentation processing result can be compareed with existing field term contrasting data library, determine material to be translated
Field.
This control can be carried out using existing semantic analysis technology, for example, semantic similarity judgement etc..
So far, the present invention has determined that wait translate technical field described in material.It is worth noting that, material to be translated may
Corresponding more than one technical field, therefore, there is also determining multiple fields term contrasting data library, the present invention to this not
It is limited.
Next, the present invention starts to handle all materials to be translated, it is finally screened out from it consistency translation art
Language.Here include:
It treats and translates material progress word segmentation processing, material to be translated is converted to translation word repertorie;
Classify to described each word to translation word repertorie, obtains different word's kinds;
Based on field term contrasting data library, the reference standard of the word of the different classifications is determined;
Based on the reference standard and field term contrasting data library, the word of the different classifications is screened, from
And obtain consistency translationese.
Point as a further improvement of the present invention, is different from the prior art, and the present invention is not to carry out for all words
Screening, but be classified first.It based on different classification, is compared using different reference standards, this process
Comparison and screening operation amount can be substantially reduced.
The classification, including word is divided into function word and notional word, wherein notional word includes common solid word, variation entity word
With particular entity word.In the present invention, function word directly skips over;For notional word, then emphasis considers particular entity word and variation entity
Word.Due to using classification policy, subsequent comparison and screening operation amount averagely reduce 78%, while also assuring accuracy.
In the second aspect of the invention, a kind of automatic extraction system of consistency translationese is disclosed, for realizing
The said extracted method of the consistency translationese.The system can use computer implemented corresponding system functional module
Form is realized, comprising:
Field determining module, the field determining module include randomly selecting module and corpus word segmentation module;It is described to randomly select
Module is used for from wait translate the corpus for randomly selecting the first predetermined ratio in material;The corpus word segmentation module is used for described random
The corpus that abstraction module extracts carries out the word segmentation processing based on natural language processing;The field determining module is based on the corpus
The word segmentation result of word segmentation module determines the field of the material to be translated, so that it is determined that at least in conjunction with field term contrasting data library
One field term contrasting data library;
Corpus conversion module, the corpus conversion module are converted to the material to be translated to translation word repertorie;
Categorization module, for the word to translation word repertorie to be classified, and output category result;
Standard determining module determines the screening criteria of different classifications word;
Screening module is based on the screening criteria, filters out consistency translationese.
Further, module is established in the field term contrasting data library with the automatic abstraction module cooperation, for establishing
Field term contrasting data library;Field term contrasting data library is established in advance and is constantly updated in translation process
Specific term database relevant to field.
It further includes feedback and update module that module is established in field term contrasting data library, for receiving translation result feedback,
Field term contrasting data library is updated.
It can be referred to using computer process in the extracting method of the third aspect of the invention, the consistency translationese
It enables and realizing, described instruction executes on computers, for realizing the method.The present invention can also provide a kind of computer-readable
Storage medium is stored thereon with computer executable instructions, for realizing the extracting method of the consistency translationese.
The more explicit embodiment of technical solution of the present invention, will be in conjunction with attached drawing part further body in a particular embodiment
It is existing.
Detailed description of the invention
Fig. 1 is the flow chart of the method described in the present invention
Fig. 2 is system framework figure of the present invention
Fig. 3 is the time cost table of comparisons using the specific implementation procedure of method of the invention.
Specific embodiment
The most typical application scenarios of the present invention occur when multiple translation team need to be jointly processed by a female translation project,
By means of the present invention the consistency vocabulary in translation project can be confirmed in the 1st time.Translating team can be in advance to knowledge
Not Chu vocabulary carry out the pretreatment of unified translation, the final translation quality for promoting whole project avoids customer complaint and anti-
The phenomenon that work of returning.
Referring to Fig.1, method of the invention is generally divided into two parts:
(1) at least one field term contrasting data library is determined;
(2) consistency translationese is filtered out.
When specific implementation, (1) partially includes following sub-step:
1) from wait translate the corpus for randomly selecting the first predetermined ratio in material, it is carried out at the participle based on natural language processing
Reason;
2) based on the word segmentation processing as a result, determining at least one field term contrasting data library.
In the present embodiment, corpus to be translated is about 58700 words, extracts the corpus of material the beginning part 15% to be translated (about
500 words), and extract the corpus (about 300 word) of ending 10%;
Based on above-mentioned extraction corpus, in conjunction with existing field term contrasting data library, judge that corpus to be translated belongs to firearms field.
Enter (2) part as a result, and execute following steps:
(I) it treats and translates material progress word segmentation processing, material to be translated is converted to translation word repertorie;
(II) classify to described each word to translation word repertorie, obtain different word's kinds;
(III) it is based on field term contrasting data library, determines the reference standard of the word of the different classifications;
(IV) based on the reference standard and field term contrasting data library, the word of the different classifications is sieved
Choosing, to obtain consistency translationese.
In the present embodiment, it treats after translating corpus word segmentation processing, obtains about 23700 words;Classification knowledge is carried out to it
Not, comprising:
(1) function word 3700, such as is/if/then/when/how/ ... etc., this kind of vocabulary do not have entity meaning;
(2) notional word: 20000, for the word in addition to above-mentioned function word, further progress is classified:
(21) common solid word, about 10800: referring to have the usual vocabulary of ordinary meaning, such vocabulary and specific field
Unrelated, meaning is fixed, for example, time/computer/ ... etc.;
(22) it makes a variation entity word 500, about: such vocabulary looks like conventional vocabulary, but due to being in specific context
Or field, thus have particular meaning, and such as: magzine, common meaning are magazine, but in firearms field, contain benefit performance
Become " magazine ";
(23) particular entity word, about 8800: the exclusive proper noun in the field, for example, Bullet(bullet).
For the word of different classifications, according to following screening:
(1) function word and notional word, directly skip over, because the meaning of this kind of word is fixed, not will cause ambiguity or translation inaccuracy;
(2) make a variation entity word: such vocabulary is given a mark according to dimensions such as distribution, capital and small letters, if score value meets second condition,
Then exported as consistency candidate terms;
(3) particular entity word: investigating its capital and small letter and frequency is given a mark, if score value meets third condition, is used as one
The candidate terms output of cause property.
Concept), then the score of the word is higher, herein, its score: Aiming can be measured using following formula
(Bindon as an example, if some particular entity word frequency of occurrence is higher and big write state, such as BAC
M1 = exp(x)+lg(Y);
M2=exp(X)+lg(Z);
Wherein, X is capital and small letter state description, and X takes 1 when capitalization, and when small letter takes 0;Y is frequency, states some word in overall vocabulary
The frequency of appearance in library is percentage;
Z is distributing position, characterizes the word in the position that overall lexicon occurs and the sum of the typical value of corresponding number, the representative
It is worth it according to following standard value:
Preceding 1/10 or rear 1/10 part for appearing in lexicon, occur z1 times, then typical value is exp (z1^10);
{ 1/10,1/5 } part is appeared in, is occurred Z2 times, then typical value is exp (z2^5);
{ 1/5,9/10 } part is appeared in, frequency of occurrence is z3 times, then typical value is exp (z3^9)
The second condition may is that M2 score value is greater than e+3;The third condition may is that M1 is greater than e+9.
Wherein, e is the natural logrithm truth of a matter, and exp () indicates that z^num indicates the num power of z using e as the index at bottom.
Above formula has fully considered the priority of capital and small letter, rather than only using frequency as sole indicator, and existing
Technology only considers that high frequency is compared, and accuracy is higher.
Referring to fig. 2, it is the functional frame composition for realizing the computer system of the method for the invention, specifically includes that
Field determining module, the field determining module include randomly selecting module and corpus word segmentation module;It is described to randomly select
Module is used for from wait translate the corpus for randomly selecting the first predetermined ratio in material;The corpus word segmentation module is used for described random
The corpus that abstraction module extracts carries out the word segmentation processing based on natural language processing;The field determining module is based on the corpus
The word segmentation result of word segmentation module determines the field of the material to be translated, so that it is determined that at least in conjunction with field term contrasting data library
One field term contrasting data library;
Corpus conversion module, the corpus conversion module are converted to the material to be translated to translation word repertorie;
Categorization module, for the word to translation word repertorie to be classified, and output category result;
Standard determining module determines the screening criteria of different classifications word;
Screening module is based on the screening criteria, filters out consistency translationese.
Further, module is established in the field term contrasting data library with computer system cooperation, for establishing neck
Domain term contrasting data library;Field term contrasting data library, in advance establish and constantly updated in translation process with
The relevant specific term database in field.
It further includes feedback and update module that module is established in field term contrasting data library, for receiving translation result feedback,
Field term contrasting data library is updated.For example, shown in dotted arrow in Fig. 1.In Fig. 2, categorization module, mark
Quasi- determining module is executing classification and mark accurate timing, can refer to existing field term contrasting data library;Meanwhile
Classification results and final the selection result can be used for feeding back update module for updating field term contrasting data library.
It is using the respective processing workload of method and art methods of the invention referring to Fig. 3.Wherein, of the invention
It only needs to do word segmentation processing for 500+300 and determines field, then for word segmentation result, it is only necessary to for 500+800 therein
Do Screening Treatment;And in the prior art, after needing to carry out word segmentation processing to all words, high frequency time identification is carried out one by one, is known
The threshold value that other precision depends on high frequency is set, and belongs to subjective setting, once setting is inaccurate, will be obtained uncontrollable consequence, and be turned over
Translating team often seems and has no way of doing it in face of so many " high frequency words ", and it is that can not think that it is big, which to arrange vocabulary bring workload,
As.
According to Fig. 3 as it can be seen that method of the invention greatly reduces workload, while it ensure that accuracy.
In conclusion the present invention has abandoned the technology prejudice for excessively focusing on high frequency words in conventional translation process, distinguish
Function word, common solid word, variation a variety of different types of vocabulary such as entity word and particular entity word.
Although will recognize that entity vocabulary more needs accurate in translation process in previous production practices
It translating and, this is extremely important to the promotion of whole translation quality, but " high frequency " has been solely focused in traditional technology, without
Once classification processing is made.
The present invention promotes project quality to reach whole by the preextraction and pretreatment to term, shortens the project cycle
Effect, the thinking of true kernel is " doing subtraction ".According to the empirical analysis of observation and practical application to industry pain spot, separately
The maintenance one " field term contrasting data library " for warding off path, further does preliminary word segmentation processing result by the link and subtracts
Method is performed for more than original lexical data 90% " compression ", as much as possible to extract to the translation helpful effective word of team
It converges, filters out meaningless, inaccurate or too common interference vocabulary.Finally achieve the effect for promoting translation efficiency and quality.
Claims (9)
1. a kind of computer implemented translation consistency term pretreatment system, the system comprises field determining module with it is consistent
Property term filtering module, wherein the field determining module include randomly select module and corpus word segmentation module;The random pumping
Modulus block is used for from wait translate the corpus for randomly selecting the first predetermined ratio in material;The corpus word segmentation module be used for it is described with
The corpus that machine abstraction module extracts carries out the word segmentation processing based on natural language processing;The field determining module is based on institute's predicate
The word segmentation result for expecting word segmentation module determines the field of the material to be translated, so that it is determined that extremely in conjunction with field term contrasting data library
A few field term contrasting data library;
The consistency term filtering module includes corpus conversion module, categorization module and standard determining module, and the corpus turns
Mold changing block is converted to the material to be translated to translation word repertorie;Categorization module is used to classify the word to translation word repertorie,
And output category result;Standard determining module, for determining the screening criteria of different classifications word;
The consistency term filtering module is based on the screening criteria, filters out consistency translationese.
2. the system as claimed in claim 1 establishes module cooperation, the field term pair with field term contrasting data library
According to Database module for establishing field term contrasting data library;Field term contrasting data library, to establish in advance
And the specific term database relevant to field constantly updated in translation process.
3. system as claimed in claim 2, wherein it further includes feedback and update mould that module is established in field term contrasting data library
Block is updated field term contrasting data library for receiving translation result feedback.
4. system as claimed in claim 3, wherein the categorization module, standard determining module are executing classification and standard
When determining, it is based on existing field term contrasting data library;Also, classification results and final the selection result for feed back with
Update module updates field term contrasting data library.
5. a kind of extracting method of consistency translationese, which comprises this method comprises the following steps:
From wait translate the corpus for randomly selecting the first predetermined ratio in material, it is carried out at the participle based on natural language processing
Reason;
Based on the word segmentation processing as a result, determining at least one field term contrasting data library;
It is characterized in that, after determining at least one field term contrasting data library,
Further include following steps:
(I) it treats and translates material progress word segmentation processing, material to be translated is converted to translation word repertorie;
(II) classify to described each word to translation word repertorie, obtain different word's kinds;
(III) it is based on field term contrasting data library, determines the reference standard of the word of the different classifications;
(IV) based on the reference standard and field term contrasting data library, the word of the different classifications is sieved
Choosing, to obtain consistency translationese.
6. method as claimed in claim 5, at least one field term contrasting data library of the determination, comprising:
(21) corpus for randomly selecting the second predetermined ratio backward from the beginning part of material to be translated carries out word segmentation processing to it,
Field term contrasting data library is determined based on word segmentation result;
And/or
(22) corpus for randomly selecting third predetermined ratio forward from the tail portion of material to be translated carries out word segmentation processing to it,
Field term contrasting data library is determined based on word segmentation result.
7. method as claimed in claim 5, field term contrasting data library, for establish in advance and in translation process not
The disconnected specific term database relevant to field updated.
8. the method for claim 7, the specific term database is established as follows:
It is collected for more than the corpus document of some specific area of predetermined quantity;
Word segmentation processing is carried out to the corpus document, corpus document is converted into lexicon;
Semantic analysis is carried out to the lexicon, for each vocabulary, comprehensive word frequency, distribution, part of speech, capital and small letter dimension are carried out
Marking, if score value meets first condition, is included in specific term database for the vocabulary.
9. a kind of computer readable storage medium includes computer executable instructions, is held by processor and memory thereon
Row described instruction, for realizing the described in any item methods of claim 5-8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811079665.0A CN109241543B (en) | 2018-09-19 | 2018-09-19 | Preprocessing technique for consistent translation terms |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811079665.0A CN109241543B (en) | 2018-09-19 | 2018-09-19 | Preprocessing technique for consistent translation terms |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109241543A true CN109241543A (en) | 2019-01-18 |
CN109241543B CN109241543B (en) | 2023-05-30 |
Family
ID=65058555
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811079665.0A Active CN109241543B (en) | 2018-09-19 | 2018-09-19 | Preprocessing technique for consistent translation terms |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109241543B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110807338A (en) * | 2019-11-08 | 2020-02-18 | 北京中献电子技术开发有限公司 | English-Chinese machine translation term consistency self-correcting system and method |
CN111126086A (en) * | 2019-12-23 | 2020-05-08 | 传神语联网网络科技股份有限公司 | Block chain system fusing translation term authentication and modification traceability |
CN111144133A (en) * | 2019-12-23 | 2020-05-12 | 传神语联网网络科技股份有限公司 | Translation document keyword modification traceability system and method based on block chain |
CN111382579A (en) * | 2020-01-13 | 2020-07-07 | 中船第九设计研究院工程有限公司 | Data preprocessing verification platform of ship pipeline manufacturing execution system |
CN111597826A (en) * | 2020-05-15 | 2020-08-28 | 苏州七星天专利运营管理有限责任公司 | Method for processing terms in auxiliary translation |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102662490A (en) * | 2012-04-08 | 2012-09-12 | 北京语智云帆科技有限公司 | Inputting method having assistant translation function |
CN103678280A (en) * | 2013-12-30 | 2014-03-26 | 武汉传神信息技术有限公司 | Translation task fragmentization method |
CN107870905A (en) * | 2017-12-04 | 2018-04-03 | 语联网(武汉)信息技术有限公司 | A kind of recognition methods of specific vocabulary |
-
2018
- 2018-09-19 CN CN201811079665.0A patent/CN109241543B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102662490A (en) * | 2012-04-08 | 2012-09-12 | 北京语智云帆科技有限公司 | Inputting method having assistant translation function |
CN103678280A (en) * | 2013-12-30 | 2014-03-26 | 武汉传神信息技术有限公司 | Translation task fragmentization method |
CN107870905A (en) * | 2017-12-04 | 2018-04-03 | 语联网(武汉)信息技术有限公司 | A kind of recognition methods of specific vocabulary |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110807338A (en) * | 2019-11-08 | 2020-02-18 | 北京中献电子技术开发有限公司 | English-Chinese machine translation term consistency self-correcting system and method |
CN110807338B (en) * | 2019-11-08 | 2022-03-04 | 北京中献电子技术开发有限公司 | English-Chinese machine translation term consistency self-correcting system and method |
CN111126086A (en) * | 2019-12-23 | 2020-05-08 | 传神语联网网络科技股份有限公司 | Block chain system fusing translation term authentication and modification traceability |
CN111144133A (en) * | 2019-12-23 | 2020-05-12 | 传神语联网网络科技股份有限公司 | Translation document keyword modification traceability system and method based on block chain |
CN111144133B (en) * | 2019-12-23 | 2023-05-26 | 传神语联网网络科技股份有限公司 | System and method for modifying and tracing translation manuscript keyword based on block chain |
CN111126086B (en) * | 2019-12-23 | 2023-05-26 | 传神语联网网络科技股份有限公司 | Blockchain system integrating translation term authentication and modification traceability |
CN111382579A (en) * | 2020-01-13 | 2020-07-07 | 中船第九设计研究院工程有限公司 | Data preprocessing verification platform of ship pipeline manufacturing execution system |
CN111597826A (en) * | 2020-05-15 | 2020-08-28 | 苏州七星天专利运营管理有限责任公司 | Method for processing terms in auxiliary translation |
CN111597826B (en) * | 2020-05-15 | 2021-10-01 | 苏州七星天专利运营管理有限责任公司 | Method for processing terms in auxiliary translation |
Also Published As
Publication number | Publication date |
---|---|
CN109241543B (en) | 2023-05-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109241543A (en) | The preconditioning technique of consistency translationese | |
US11151182B2 (en) | Classification model training method and apparatus | |
CN109299480B (en) | Context-based term translation method and device | |
CN105095190B (en) | A kind of sentiment analysis method combined based on Chinese semantic structure and subdivision dictionary | |
CN105022725A (en) | Text emotional tendency analysis method applied to field of financial Web | |
CN105956179B (en) | Data filtering method and device | |
CN103678565B (en) | Domain self-adaption sentence alignment system based on self-guidance mode | |
US20180336272A1 (en) | Generation of natural language processing events using machine intelligence | |
CN108363691A (en) | A kind of field term identifying system and method for 95598 work order of electric power | |
CN111563377A (en) | Data enhancement method and device | |
CN115168615A (en) | Knowledge graph big data processing method and system combining data visualization | |
CN109508557A (en) | A kind of file path keyword recognition method of association user privacy | |
CN113468317A (en) | Resume screening method, system, equipment and storage medium | |
CN112667819A (en) | Entity description reasoning knowledge base construction and reasoning evidence quantitative information acquisition method and device | |
Siddique et al. | Bilingual word embeddings for cross-lingual personality recognition using convolutional neural nets | |
CN111460147A (en) | Title short text classification method based on semantic enhancement | |
CN116304046A (en) | Dialogue data processing method and device, storage medium and electronic equipment | |
CN111736804B (en) | Method and device for identifying App key function based on user comment | |
CN109947942B (en) | Bayesian text classification method based on position information | |
CN114328903A (en) | Text clustering-based customer service log backflow method and device | |
CN110069780B (en) | Specific field text-based emotion word recognition method | |
CN112541352A (en) | Policy interpretation method based on deep learning | |
CN102982029A (en) | Identification method and device for searching requirement | |
CN112329478A (en) | Method, device and equipment for constructing causal relationship determination model | |
CN117556049B (en) | Text classification method of regular expression generated based on large language model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |