CN106257442A - Computer-aided translation method - Google Patents
Computer-aided translation method Download PDFInfo
- Publication number
- CN106257442A CN106257442A CN201610455156.8A CN201610455156A CN106257442A CN 106257442 A CN106257442 A CN 106257442A CN 201610455156 A CN201610455156 A CN 201610455156A CN 106257442 A CN106257442 A CN 106257442A
- Authority
- CN
- China
- Prior art keywords
- language
- labelled element
- translation
- computer
- template
- 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.)
- Pending
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
- G06F40/42—Data-driven translation
- G06F40/47—Machine-assisted translation, e.g. using translation memory
-
- 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
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
A kind of computer-aided translation method, it is performed by computer, and wherein, described computer-aided translation method comprises following process: pretranslation processes, each language fragments in the statement of first language is processed as the language fragments of second language, and word order is adjusted;And labelled element display processes, processing in the pretranslation result as second language obtained by described pretranslation, at position display labelled element corresponding with the language fragments of described first language.Described computer-aided translation method can also comprise following labelled element removing and process: the instruction for removing labelled element made in response to user, is removed by described labelled element.Thus, user can be made to be easy to identify first language and the corresponding relation of second language.
Description
Technical field
The present invention relates to computer-aided translation (CAT:Computer Aided Translation) method.
Background technology
At present, have based on computer implemented machine translation (MT) technology and can be divided into rule-based machine translation
(RBMT), machine translation (SMT) based on statistics, Case-based design (EBMT), machine translation based on template
(TBMT) machine translation mothod etc..It addition, there is also with translation memory (translation memory, TM) technology as representative
Computer-aided translation technology.
Either machine translation or computer-aided translation, is all that (below, language to be translated is denoted as the first language sometimes
Speech) statement translation become the statement of object language (following, be sometimes denoted as second language).But at current machine translation (bag
Include computer machine translation and computer-aided translation) in product, the corresponding relation between first language and second language
There is many problems in display aspect.
Such as, as it is shown in figure 1, first language is positioned at left side, the second language as translation result is positioned at right side.They divide
It is not made up of multiple language fragments (be the most sometimes called for short and do fragment).So-called language fragments refers to each ingredient of sentence,
They can style be to have the vocabulary of physical meaning and only represent element (such as, the carrying in Japanese etc. of syntactic information
Show " は " of subject) etc..During for example, it is also possible to eated be divided into as the segment portion " eat " representing physical meaning and represent
The segment portion " ed " of state grammer.
As it is shown in figure 1, in existing translated product, such as, if mouse to be placed on certain of the statement of second language
In individual language fragments, then in the statement of first language, show corresponding language fragments with the method for shade.This display packing
There are the following problems, first of all, it is necessary to be switched to show the pattern of corresponding relation by mouse etc., such as, at example shown in Fig. 1
In, need to be placed in certain fragment mouse, be switched to show the model of corresponding relation, just show corresponding relation.If Mus
Mark does not indicate certain fragment, does not the most show the corresponding relation between macaronic daughter element.
Additionally, example the most as shown in Figure 1 is such, corresponding relation can only be shown singly.Additionally, can not show
The change of word order, such as, English is typical SVO (SVO) word order, and Japanese is SOV word order, and the product shown in Fig. 1 can not
The word order position that subunit is original is demonstrated, it is necessary to the display in reference left side could grasp original word order position in second language
Put.Additionally, the syntactic element etc. of the first language being hidden can not be demonstrated in second language.Such as, such as apples
In " s " represent plural number, and Chinese, Japanese etc. the most do not use plural number, therefore, only referring to " Fructus Mali pumilae " in second language, then
Cannot know Fructus Mali pumilae corresponding be " apples " or " apple ".
Summary of the invention
For the problems referred to above, the present invention provides a kind of computer-aided translation method, and it is performed by computer, wherein,
Described computer-aided translation method comprises following process:
Pretranslation processes, and each language fragments in the statement of first language is processed as the language fragments of second language;
And
Labelled element display processes, and is being processed the pretranslation result as second language obtained by described pretranslation
In, at position display labelled element corresponding with the language fragments of described first language.
Process additionally, described computer-aided translation method also comprises following labelled element removing: make in response to user
For removing the instruction of labelled element, described labelled element is removed.
Additionally, corresponding position can include adjusting the position before word order and adjusting the position after word order.
Additionally, corresponding position can include the position occurring to split.
Additionally, described labelled element can be character or picture.
A kind of computer-aided translation program, it makes, and computer execution is following to be processed:
Pretranslation processes, and each language fragments in the statement of first language is processed as the language fragments of second language;
And
Labelled element display processes, and is being processed the pretranslation result as second language obtained by described pretranslation
In, at position display labelled element corresponding with the language fragments of described first language.
Process additionally, described computer-aided translation program also makes computer perform following labelled element removing: in response to
The instruction for removing labelled element that user makes, removes described labelled element.
Additionally, corresponding position can include adjusting the position before word order and adjusting the position after word order.
Additionally, corresponding position can include the position occurring to split.
Additionally, described labelled element can be character or picture.
By with upper type, the sheet of second language can be shown intuitively as in the translation result of second language
Corresponding relation between section and the fragment of first language.
Accompanying drawing explanation
Fig. 1 shows the example for illustrating to show in existing translated product corresponding relation.
Fig. 2 shows the composition structure of supplementary translation program P.
Fig. 3 shows the example of translation template.
Fig. 4 shows the example of the labelling template corresponding with translation template.
Fig. 5 shows the example of the translation result showing labelled element.
Fig. 6 shows the example of the labelling template of additional marking element in dictionary.
Fig. 7 shows the example of the dictionary with part of speech.
Detailed description of the invention
Embodiment 1
The computer-aided translation method of the present invention such as can be performed supplementary translation program by the PC1 as personal computer
P realizes.Supplementary translation program P is such as stored in network or readable medium such as CDROM that computer can access etc.
In.Personal computer PC 1 can be common personal computer, and it has processor, memorizer, display and mouse-keyboard
Deng.Personal computer PC 1 reads auxiliary by downloading from network or having the CDROM medium of supplementary translation program P from storage
Interpretive program P also performs, and carries out translation process.Supplementary translation program P is configured to include that pretranslation module P1, labelled element are aobvious
Show that module P2 and labelled element remove module P3.
Such as, user indicates supplementary translation program P to carry out translation process by keyboard or mouse.In response to this instruction,
In pretranslation module P1, for the statement of first language to be translated, carry out pretranslation based on existing interpretation method.In rollover
In translating, each language fragments in the statement of first language is processed as the language fragments of second language, and carries out to word order
Adjust.
It is for instance possible to use machine translation (TBMT) technology based on template carries out pretranslation.At machine based on template
In device translation, template is to comprise by source language (first language) and target language (second language) both written, comprise change
Amount (concrete syntax fragment) and the character string of variable (variable language fragments).Template also show between macaronic character string
Corresponding translation relation.
Such as, as it is shown on figure 3, preserve such as " I like [A] → private in the translation template storehouse realized by text
は [A] Ga good I In The " such translation template.Word " I " in this template and the constant fragment that " like " is first language
(concrete term fragment), [A] is the variable fragment (variable word fragment) of first language." private は " in this template and " Ga is good
I In The " it is the constant fragment (concrete term fragment) of second language, [A] is variable fragment (the variable word sheet of second language
Section).In this template " → " corresponding relation between bilingual is shown, " I like [A] " " private は [A] Ga will be translated as
Good I In The ".Similarly, " I eat [A] " is translated as " private は [A] The by the template representation shown in " I eat [A] " in Fig. 3
Food べ Ru ".
Such as, for " I like apple " such statement to be translated, in pretranslation module P1, first from translation mould
Plate storehouse selects the template that the degree of approximation is the highest.Such as, approximation can be calculated according to the appearance constant number in statement to be translated
Degree, the constant number degree of approximation the most at most of appearance is the most.In the template shown in Fig. 3, the template only " I " shown in " I eat [A] "
Come across statement to be translated, and " eat " does not comes across statement to be translated, i.e. the degree of approximation is 1.By comparison, in " I like "
" I " and " like " all comes across statement to be translated, the i.e. degree of approximation is 2, therefore the template shown in " I like [A] " and language to be translated
The similarity of sentence is higher than the template shown in " I eat [A] ", and " I " and " like " is corresponding to the constant in template, and " apple "
Corresponding to the variable [A] in template.
It is thus possible, for instance when translating " I like apple ", select the template shown in " I like [A] " as the degree of approximation
High template is translated.According to " I like [A] " shown in template and the relation between " private は [A] Ga good I In The ",
" I like apple " is translated as " private は [A] Ga good I In The ".Then, for the variable [A] corresponding with " apple ", logical
Crossing the methods such as lookup dictionary the most as shown in Figure 7, the corresponding word obtaining apple is " Fructus Mali Asiaticae ".To put as variable [A]
It is changed to " Fructus Mali Asiaticae ", thus " I like apple " is translated as " private は Fructus Mali Asiaticae Ga good I In The " the most at last.
Additionally, for the ease of showing labelled element in subsequent treatment, pretranslation module P1 such as records in internal memory and turns over
Translate each constant in result and the starting position of variable and end position.Such as, in the case of in units of character, constant
Starting position and the end position of " private は " are respectively 1 and 2, the starting position of the variable " [A] " corresponding with " Fructus Mali Asiaticae " and end
Position is respectively 3 and 4, and starting position and the end position of constant " Ga good I In The " are respectively 5 and 9.
In labelled element display module P2, for the result of pretranslation, show labelled element based on labelling template.Example
As, as shown in Figure 4, preserve in the labelling template base realized by text " I like [A] → private は<like>[A] Ga<
Good I In The><like>" such labelling template.Labelling template can be by increasing labelled element on the basis of translation template
Obtain.Labelling template is divided into following three parts: the left part shown in " I like [A] " is corresponding to the in translation template
One verbal portions, it is the title of template, can uniquely identify template as key;" private は<like>[A] Ga<good I In The>"
Shown mid portion is the second language part in translation template to be increased labelled element and obtains (same with translation template
Sample, → corresponding relation of first language part and second language part is shown);Right part shown in "<like>" is for rear
State removing labelled element in process.In the case of there is multiple labelled element to be removed, such as, can use separations such as " | "
Symbol separates each labelled element.
Such as, in the case of " I like apple ", owing to the translation template used in pretranslation module P1 is
Template shown in " I like [A] ", therefore retrieves corresponding labelling template according to " I like [A] ", thus obtain " private は <
Like>[A] Ga<good I In The>".Wherein, the position that "<" and ">" is surrounded is labelled element, in this example, for<like>and
<good I In The>.They are illustrated respectively in "<like>" and "<good I In The>" location display labelled element.
For example, it is possible to show labelled element with bold Italic.So, as it is shown in figure 5, for as pretranslation result
" private は Fructus Mali Asiaticae Ga good I In The ", each constant recorded according to pretranslation module P1 and the starting position of variable and stop bits
Putting and show labelled element, finally, the translation result of " I like apple " becomes " private は like woods after display labelled element
Qin Ga good I In The ".
For the ease of removing labelled element in subsequent treatment, labelled element display module P2 can preserve labelled element
Position.Such as, the starting position of the labelled element shown in<like>and end position are respectively 3 and 6, shown in<good I In The>
The starting position of labelled element and end position are respectively 10 and 13.
Thus, according to translation template, such as, the word order of first language is adjusted to the word order of second language, and as
In the translation result of two language, the position before language fragments is adjusted word order shows with being adjusted the position after word order respectively
Show labelled element so that user can grasp the sheet before the fragment occurring word order to convert and translation intuitively in translation result
Corresponding relation between Duan and relative position.
The corresponding relation that user shows based on labelled element, it is possible to be readily understood by the translation result of pretranslation, and to turning over
Translate result to be modified.Then, if user is satisfied to translation result, supplementary translation program can be indicated by keyboard or mouse etc.
P removes labelled element.For example, it is possible to the operation to " removing labelled element " distributes shortcut F8.When user presses F8, mark
Note element removes the instruction that module P3 is made in response to user, for the labelled element needing to remove in above-mentioned labelling template
Part, starting position based on the labelled element recorded at labelled element display module P2 and end position, by translation result
Labelled element remove.
Such as, in the labelling template shown in " I like [A] → private は<like>[A] Ga<good I In The><like>",
Right part represents to be removed the labelled element shown in<like>.Labelled element according to labelled element display module P2 record
Position, " like " part of 3 to 6 in translation result is labelled element, is therefore removed.And owing to labelling template does not indicates
Removing<good I In The>, therefore, in labelled element removing processes, " good I In The " is retained as translation result.Thus, pin
" like " and " good I In The " to the effect playing labelled element in the example shown in Fig. 5, only " like " part are eliminated,
Thus obtain the translation result after removing labelled element, i.e. " private は Fructus Mali Asiaticae Ga good I In The ".
It is pointed out that labelled element removes module P3 not necessarily.Labelled element can also be omitted and remove module
P3.In this case, user can manually remove labelled element, and the right part of labelling template can also be omitted.
Embodiment 2
In the above description, translation template and labelling template are separately maintained in different file.But, due to labelling mould
Left part and mid portion in plate increase labelled element on translation template and obtain, and therefore can play translation mould
The effect of plate, thus translation template file can be omitted.
Such as, in the case of omitting translation template file, when translating " I like apple ", utilize above-mentioned calculating
The method of the degree of approximation, obtains the template " I like [A] → private は<like>[A] Ga<good I In The><like>" of correspondence.With with
The pretranslation based on translation template of upper explanation processes similarly, uses " I like [A] → private は<like>[A] Ga<good I In
The>" as translation template, " I like apple " is translated as " private は<like>[A] Ga<good I In The>".Then look up word
Allusion quotation, replaces with " Fructus Mali Asiaticae " by variable [A], obtains " private は<like>Fructus Mali Asiaticae Ga<good I In The>".The place of shown below labelling
Manage as broad as long with described above, the most no longer describe.
Additionally, about the calculating of the template degree of approximation, constant illustrated above occurs that the method for number is only example, such as also
Can on the basis of there is number in calculation constant, by comparison constant in a template appearance order with constant in be translated
The similarity i.e. word order similarity etc. of appearance order increases the accuracy the etc. during degree of approximation of calculation template and statement to be translated.
In addition it is also possible to adopt the template selecting to mate most with statement to be translated in other ways.
Variation 1
Above, employ TBMT technology to carry out pretranslation, but be not limited to this, it is also possible to carry out pre-by additive method
Translation.It is for instance possible to use rule-based machine translation mothod carries out pretranslation.Machine at rule-based (RBMT) turns over
Translate under technical situation, in the same manner as described above, set up mark accordingly with each translation rule in the storehouse of storage translation rule
Note template, shows labelled element for rule-based pretranslation result.In the same manner as embodiment 1,2, the process in this example
Comprise pretranslation process, labelled element display process and labelled element removing to process.Due to only pretranslation process and embodiment
1,2 is different, the most only process pretranslation and illustrate.
Such as, there is following regular A1~A5, they represent that the sentence structure of English sentence is (the most sometimes by rule A1
~A5 is referred to as rule A).
Rule A1:S1 → NP VP
Rule A2:NP → NP and NP
Rule A3:NP → N
Rule A4:VP → VP NP
Rule A5:VP → V
Wherein, rule A1 represents that English sentence S1 can be made up of noun phrase NP, auxiliary word は and verb phrase VP.
Rule A2 represents that noun phrase NP can be connected two noun phrase NP by auxiliary word and and constitute.
Rule A3 represents that noun phrase NP can be made up of noun N and (for the purpose of simplifying the description, in this example, is also solved by pronoun
Analysis is noun).
Rule A4 represents that verb phrase VP can be made up of noun phrase NP and verb phrase VP.
Rule A5 represents that verb phrase VP can be made up of verb V.
Following regular B1~B5 represents the Japanese corresponding with rule A1~A5.
Rule B1:S2 → NP は VP
Rule B2:NP → NP と NP
Rule B3:NP → N
Rule B4:VP →<VP>NP The VP
Rule B5:VP → V
Wherein, rule B1 represents that sentence S2 can be made up of noun phrase NP, auxiliary word は and verb phrase VP.
Rule B2 represents that noun phrase NP can be connected two noun phrase NP by auxiliary word と and constitute.
Rule B3 represents that noun phrase NP can be made up of noun N and (for the purpose of simplifying the description, in this example, is also solved by pronoun
Analysis is noun).
Rule B4 represents that verb phrase VP can be constituted The by noun phrase NP, auxiliary word and verb phrase VP is constituted.
Rule B5 represents that verb phrase VP can be made up of verb V.
Additionally, regular A1, B1, rule A2, B2, rule A3, B3, rule A4, B4, rule the most corresponding the turning over of A5, B5
Translate relation.Noun N and verb V as termination word can complete translation, herein, so-called termination by the method searching dictionary
Word, refers to, in rule-based translation processes, to carry out the least unit of rule analysis.With embodiment 1,2 similarly,
Represent labelled element with "<" and ">", thus<VP>in rule B4 is labelled element.
Hereinafter, as a example by " I eat apple ", pretranslation based on RBMT is illustrated.At the dictionary shown in Fig. 7
In, each brief note is made up of English word, corresponding Japanese word and part of speech respectively.
When " I eat apple " is translated by rule-based A, B, first, such as can be based on the dictionary shown in Fig. 7
Analyze the part of speech of each word.It is analyzed based on dictionary, it is known that word I, word apple are noun and meet rule A3,
Word eat is verb and meets rule A5.And then, " eat apple " meets rule A4.And then, " I eat apple " meets rule
Then A1.Based on A1 Yu B1, the corresponding relation of A4 Yu B4, " I eat apple " is translated as " I<eat>apple The eat ".So
After, based on A3 Yu B3, the corresponding relation of A5 Yu B5, by searching dictionary, " I<eat>apple The eat " is processed as " private は<
Eat > Fructus Mali Asiaticae The food べ Ru ".Other display the display process of<eat>of element, delete processing etc. do not have with embodiment 1,2
Difference, therefore omits the description.
In this example, rule-based method carries out pretranslation process and carries out the display process of comparison element, permissible
Obtain and the identical effect of embodiment 1,2.
Additionally, in this example, translation rule can be recorded in template file, it is also possible to is built in program, does not has this
Limit.
Other variation
Additionally, in the above description, labelled element is removed according to the display position of labelled element.But can also be according to mark
Display information display modes etc. such as () content of text, font color, runic, italic, underscores of note element or composite marking
The display information of element and position are purged.Such as, in labelled element display module P2, can showing labelled element
Show that record is in such as internal memory, in order in subsequent treatment, remove labelled element directly access internal memory, i.e. can get to be cleaned
The content of text of labelled element, the information such as display mode, and remove labelled element according to these information.
It is, for example possible to use the special color of labelled element shows labelled element, in display labelling is removed and disposed, logical
Cross and judge whether the color of translation result is that special color is purged.So, both it had been easy to so that user passes through display mode
And easily tell labelled element and the corresponding relation easily determining out between fragment, it is easy to again in follow-up process permissible
Labelled element is removed according to display mode etc..It is likewise possible to by the special font of labelled element, underscore, font size
Labelled element is carried out, as long as labelled element can be made different from the display mode of the other parts of sentence Deng other modes.
Additionally, in the above description, the display packing of labelled element is built-in, but can also be at labelling template file
In, increase for specifying the content how labelled element shows, user come the color of the font that regulation shows, size etc..By
This, user can change into, by editor's labelling template, the display mode oneself liked.
Additionally, in the above description, labelled element is to be realized by the display mode (font color etc.) adjusting text
But it also may use the mode of picture to show.Such as, in the example as shown in fig. 5, can generate and show " like "
The picture of " good I In The ", inserts ground display picture in the position that " like " is corresponding, and in the position of " good I In The " correspondence
Cover ground display picture.In removing processes, the picture playing labelling effect is removed.
Additionally, when setting up labelling template, convert such paired labelled element, except word for such as there is word order
Beyond the mode such as body, color, in labelled element, preferably embody the corresponding relation of semanteme.Such as, right with " good I In The "
In the labelled element answered, it is preferred to use " like " is such can show corresponding semantic display mode.
In the above description, it is shown that as the example of the word order generation adjustment of " like " describing verb, but be not limited to
This.Similarly, other word orders are adjusted and can also be shown by labelling template.Such as, " the 00 To つ い て " of Japanese is corresponding
English be " about 00 ".The language fragments that preposition occurs word order to adjust is comprised, it is also possible to based on " [A] for such
To つ い て → about [A]<To つ い て><To つ い て>" such labelling template shows the " つ corresponding to " about "
い て " word order adjust before position.Process to this template is identical with mode described above, the most no longer describes.
Additionally, for the fragment not occurring word order to adjust, it is also possible to allowed users to by display labelled element
The corresponding relation of the statement fragment before the statement fragment of pretranslation result easy to understand and translation.Such as, the Indo-European languages such as English
The language of system often has single plural form, and Chinese, Japanese are often used without plural number.For the ease of as translation result
The linguistic informations such as the DANFU number showing first language in the statement of second language intuitively, such as, can increase on the basis of dictionary
Label element, definition labelling template as shown in Figure 6.With described above similarly, this labelling template is divided into three parts: left
Side is the key of template;Mid portion is used for showing labelled element;Right side is for removing labelled element in subsequent treatment.
Thus, show such labelled element by labelled element display module P2, such as, " apples " is processed as
" Fructus Mali pumilae s " so that user can be in the statement as the second language of translation result, it is possible to intuitively arrives and turns over conventional
Translate the corresponding relation cannot seen in product, and syntactic element (such as, DANFU number, tense, the moon being hidden or having removed
Property positive etc.), so that translation result can be modified easily by user based on these syntactic elements.
Additionally, similarly, for comprising the complicated long English sentence with subordinate clause of which etc, the usual root when translation
Short sentence easy to understand is split into according to situation.Such as, by as the original which's in the translation result of second language
Show labelled element position embeddedly so that user's original sentence structure easy to understand and modify easily.
Such as, for " Acroponics is another method for growing plants without soil, a
technique in which plants are suspended and the roots misted with a nutrient
Solution " this sentence, in translation result, show " in which " with embedding in the way of italics, tied as follows
Really: " air culture method is another soilless culture method, [in which] this be a kind of plant to be hung up and its root is soaked
Enter the technology of nutritional solution ".In this example, by italics [in which] so that user easily identifies determining in translation result
Corresponding relation between language subordinate clause and first language.
Translation template storehouse based on text illustrated above and labelling template base are merely illustrative, it would however also be possible to employ XML
File, database mode.
Additionally, be illustrated with personal computer above but it also may make this computer (office can be included with network
Territory net, VLAN, the Internet etc.) connect, the clothes that one or more module section of supplementary translation program P are deployed on network
Business device.For example, it is possible to the server that pretranslation module P1, labelled element display module P2 are deployed on the Internet, and will mark
Note element is removed module P3 and is deployed in and can thus constitute, with client, the auxiliary translation system being made up of server and client side.
In addition it is also possible to the server that supplementary translation program P is all deployed on the Internet, for using auxiliary by browser
The function that interpretive program P provides.
Additionally, the above each variation can be with embodiment 1,2 independent assortment.
Industry is applied
Present invention could apply to computer-aided translation method etc..
Claims (4)
1. a computer-aided translation method, it performs by computer, wherein,
Described computer-aided translation method comprises following process:
Pretranslation processes, and each language fragments in the statement of first language is processed as the language fragments of second language;And
Labelled element display processes, and is processing in the pretranslation result as second language obtained by described pretranslation,
Position display labelled element corresponding with the language fragments of described first language.
Computer-aided translation method the most according to claim 1, wherein,
Described computer-aided translation method also comprise following labelled element removing process: in response to user make for removing
The instruction of labelled element, removes described labelled element.
Computer-aided translation method the most according to claim 1, wherein,
Corresponding position includes adjusting the position before word order and adjusting the position after word order.
Computer-aided translation method the most according to claim 1, wherein,
Described labelled element is character or picture.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510357332X | 2015-06-22 | ||
CN201510357332 | 2015-06-22 |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106257442A true CN106257442A (en) | 2016-12-28 |
Family
ID=54498375
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510627948.4A Pending CN105069001A (en) | 2015-06-22 | 2015-09-28 | Computer aided translation method |
CN201610455156.8A Pending CN106257442A (en) | 2015-06-22 | 2016-06-21 | Computer-aided translation method |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510627948.4A Pending CN105069001A (en) | 2015-06-22 | 2015-09-28 | Computer aided translation method |
Country Status (2)
Country | Link |
---|---|
CN (2) | CN105069001A (en) |
WO (1) | WO2016206582A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110291521A (en) * | 2018-01-19 | 2019-09-27 | 株式会社咕嘟妈咪 | Control method, the control program of server and server of server |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105069001A (en) * | 2015-06-22 | 2015-11-18 | 张云鹏 | Computer aided translation method |
CN108664545A (en) * | 2018-03-26 | 2018-10-16 | 商洛学院 | A kind of translation science commonly uses data processing method |
CN108491400A (en) * | 2018-04-09 | 2018-09-04 | 江苏省舜禹信息技术有限公司 | A kind of supplementary translation intelligence system and method |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102708097A (en) * | 2012-04-27 | 2012-10-03 | 曾立人 | Online computer translation method and online computer translation system |
CN104679735A (en) * | 2013-11-30 | 2015-06-03 | 赵会军 | Pragmatic machine translation method |
CN105069001A (en) * | 2015-06-22 | 2015-11-18 | 张云鹏 | Computer aided translation method |
-
2015
- 2015-09-28 CN CN201510627948.4A patent/CN105069001A/en active Pending
-
2016
- 2016-06-21 CN CN201610455156.8A patent/CN106257442A/en active Pending
- 2016-06-21 WO PCT/CN2016/086628 patent/WO2016206582A1/en active Application Filing
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110291521A (en) * | 2018-01-19 | 2019-09-27 | 株式会社咕嘟妈咪 | Control method, the control program of server and server of server |
Also Published As
Publication number | Publication date |
---|---|
WO2016206582A1 (en) | 2016-12-29 |
CN105069001A (en) | 2015-11-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109783826B (en) | Automatic document translation method | |
WO2018207723A1 (en) | Abstract generation device, abstract generation method, and computer program | |
CN107463553B (en) | Text semantic extraction, representation and modeling method and system for elementary mathematic problems | |
JP5452868B2 (en) | Training for text-to-text applications that use string-to-tree conversion for training and decoding | |
Hnátková et al. | The SYN-series corpora of written Czech. | |
Pinnis et al. | Neural machine translation for morphologically rich languages with improved sub-word units and synthetic data | |
CN106257442A (en) | Computer-aided translation method | |
CN112766000B (en) | Machine translation method and system based on pre-training model | |
CN106021224A (en) | Bilingual discourse annotation method | |
JP2017199363A (en) | Machine translation device and computer program for machine translation | |
CN110472040A (en) | Extracting method and device, storage medium, the computer equipment of evaluation information | |
CN107436931B (en) | Webpage text extraction method and device | |
Wax | Automated grammar engineering for verbal morphology | |
KR20040024619A (en) | Third language text generating algorithm by multi-lingual text inputting and device and program therefor | |
JP2016164707A (en) | Automatic translation device and translation model learning device | |
JP2004220266A (en) | Machine translation device and machine translation method | |
Aasha et al. | Machine translation from English to Malayalam using transfer approach | |
CN108701126A (en) | Theme estimating device, theme presumption method and storage medium | |
Soares et al. | A parallel corpus of theses and dissertations abstracts | |
CN105426355A (en) | Syllabic size based method and apparatus for identifying Tibetan syntax chunk | |
Rikters et al. | Combining machine translated sentence chunks from multiple MT systems | |
Sridhar et al. | A Scalable Approach to Building a Parallel Corpus from the Web. | |
JP5453779B2 (en) | Language analysis system, language analysis method, program and machine translation system, language analysis method, program | |
JP2008225566A (en) | Device and method for extracting related information | |
Rajalingam | A rule based iterative affix stripping stemming algorithm for Tamil |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20161228 |
|
WD01 | Invention patent application deemed withdrawn after publication |