CN101271451A - Computer aided translation method and device - Google Patents
Computer aided translation method and device Download PDFInfo
- Publication number
- CN101271451A CN101271451A CNA2007100891631A CN200710089163A CN101271451A CN 101271451 A CN101271451 A CN 101271451A CN A2007100891631 A CNA2007100891631 A CN A2007100891631A CN 200710089163 A CN200710089163 A CN 200710089163A CN 101271451 A CN101271451 A CN 101271451A
- Authority
- CN
- China
- Prior art keywords
- mentioned
- languages
- sentence
- translation
- example sentence
- 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
Images
Landscapes
- Machine Translation (AREA)
Abstract
The present invention provides a computer-aided translating method and a computer-aided translating device. The computer-aided translating method provided by the present invention includes: a similar first-language example sentence is searched in an aligned bilingual library for a first-language sentence to be translated; wherein, the aligned bilingual library includes pairs of corresponding example sentences in the first language and the second language, as well as the alignment information of each pair of example sentences; compared with the similar first-language example sentence, the first-language sentence to be translated provided with at least a same part and a different part respectively prepares a corresponding translation to each part of at least one different part; a reference translation is generated by combining the corresponding second-language example sentence of the similar first-language example sentence with the translation of each part of at least one different part; the similar first-language example sentence and the alignment information of the corresponding second-language example sentence are utilized to indicate the relationship between each part of at least one same part and/or at least one different part in the first-language sentence to be translated and the corresponding part in the reference translation.
Description
Technical field
The present invention relates to the information processing technology, particularly, relate to computer-aided translation (Computer Aided Translation, CAT) technology based on the bilingual alignment technology.
Background technology
Computer aided translation system is a kind of system that utilizes computing machine to help human translation.Sentence that auxiliary translation system uses translation memory (Translation Memory) technology to store to have translated and translation thereof are as translation instance (Translation Example).In the process that the user translates, system can search example sentence similar to importing sentence and translation thereof automatically, and the translation of the example sentence found is offered the user as the reference translation, thereby human translation is offered help.But the user must determine final correct translation according to the difference edit reference translation between input sentence and the example sentence.
At present, machine-aided translation system is imported sentence and highlighted with reference to the same section in the source language sentence of example sentence (or difference part) meeting when showing with reference to example sentence.But these parts that highlight corresponding translation part in the translation of example sentence is not identified or is identified.So, in translation, comprised unwanted translation fragment.Therefore, in the process of editor's reference translation, the translator must contrast the identical or difference part in input sentence and the example sentence source language sentence repeatedly, determines required translation fragment.The translator need browse wholely with reference to example sentence in this process, has reduced translation efficiency.
Summary of the invention
In order to solve above-mentioned problems of the prior art, the invention provides the method for computer-aided translation and the device of computer-aided translation.
According to an aspect of the present invention, a kind of method of computer-aided translation is provided, comprise: be the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence; When the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation respectively; The translation combination of the example sentence of second languages that will be corresponding with the example sentence of above-mentioned similar first languages and each of above-mentioned at least one difference part generates reference translation; And the alignment information of utilizing the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
According to another aspect of the present invention, a kind of method of computer-aided translation is provided, comprise: be the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence; Whether the similarity of sentence of judging the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated is less than a predetermined threshold; When the similarity of the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated during less than above-mentioned predetermined threshold, utilize the right combination of a plurality of example sentences in the above-mentioned bilingual example sentence storehouse of having carried out alignment to obtain reference translation, the example sentence of first languages of each that wherein above-mentioned a plurality of example sentences are right comprises at least one fragment of the sentence of above-mentioned first languages to be translated; And utilize the right alignment information of above-mentioned a plurality of example sentence, point out each fragment in the sentence of above-mentioned first languages to be translated and the corresponding relation of the corresponding fragment in the above-mentioned reference translation.
According to another aspect of the present invention, a kind of device of computer-aided translation is provided, comprise: search unit (searching unit), be used to the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence; Translation unit (translating unit), be used for when the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation respectively; Assembled unit (combining unit) is used for the translation combination with each of the example sentence of second languages corresponding with the example sentence of above-mentioned similar first languages and above-mentioned at least one difference part, generates reference translation; And Tip element (indicating unit), be used to utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
According to another aspect of the present invention, a kind of device of computer-aided translation is provided, comprise: search unit (searching unit), be used to the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence; Judging unit (determining unit), whether the similarity of sentence that is used to judge the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated is less than a predetermined threshold; Assembled unit (combining unit), when being used for similarity when the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated less than above-mentioned predetermined threshold, utilize the right combination of a plurality of example sentences in the above-mentioned bilingual example sentence storehouse of having carried out alignment to obtain reference translation, the example sentence of first languages of each that wherein above-mentioned a plurality of example sentences are right comprises at least one fragment of the sentence of above-mentioned first languages to be translated; And Tip element (indicating unit), be used to utilize the right alignment information of above-mentioned a plurality of example sentence, point out each fragment in the sentence of above-mentioned first languages to be translated and the corresponding relation of the corresponding fragment in the above-mentioned reference translation.
Description of drawings
Believe by below in conjunction with the explanation of accompanying drawing, can make people understand the above-mentioned characteristics of the present invention, advantage and purpose better the specific embodiment of the invention.
Fig. 1 is the process flow diagram of the method for computer-aided translation according to an embodiment of the invention;
Fig. 2 is the synoptic diagram according to the demonstration reference translation of the embodiment of Fig. 1;
Fig. 3 is the process flow diagram of the method for computer-aided translation according to another embodiment of the invention;
Fig. 4 is according to the process flow diagram of the best example sentence of searching of the embodiment of Fig. 3 to the method for combination;
Fig. 5 is the synoptic diagram according to the demonstration reference translation of the embodiment of Fig. 3;
Fig. 6 is the process flow diagram that shows and edit according to the embodiment of Fig. 3;
Fig. 7 is the synoptic diagram that highlights edit segment according to the embodiment of Fig. 3;
Fig. 8 is the synoptic diagram according to the right alignment information of the example sentence of the embodiment of Fig. 3;
Fig. 9 is the block scheme of the device of computer-aided translation according to another embodiment of the invention; And
Figure 10 is the block scheme of the device of computer-aided translation according to another embodiment of the invention.
Embodiment
Below just in conjunction with the accompanying drawings each embodiment of the present invention is described in detail.
Fig. 1 is the process flow diagram of the method for computer-aided translation according to an embodiment of the invention.As shown in Figure 1, at first,, import the sentence of first languages to be translated in step 101.
For example: Mary played a small role in the drama.
Then, in step 102, sentence for first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, the bilingual example sentence storehouse of having carried out alignment is the bilingual example sentence storehouse of having been carried out word alignment by professional's (for example, the translator) craft or computing machine automatically, it comprise many to translation each other corresponding first languages and the alignment information between the example sentence of second languages and the every pair of example sentence.The similar example sentence that for example finds to and alignment information as follows.
Then, in step 103, the example sentence of the sentence of first languages to be translated and similar first languages is compared, if the sentence of first languages to be translated is incomplete same with the example sentence of the first similar languages, find out the difference part between them, and can utilize word alignment information between the example sentence of the example sentence of the first similar languages and second languages corresponding with it, determine described difference part position in the example sentence of second languages of correspondence.
For above-mentioned example, the difference between sentence to be translated and the similar example sentence is Mary and John, and small and critical.These two positions of difference part in the Chinese example sentence corresponding with similar example sentence are respectively primary importance and the 5th position.
Then, in step 104, for corresponding translation is partly prepared in the difference between the example sentence of the sentence of first languages to be translated and similar first languages.Here, can search the translation of described difference part in dictionary, or preferably search from above-mentioned bilingual example sentence storehouse, the present invention is to this not restriction.
Particularly, the translation that can obtain Mary and small is respectively " Mary " and " little ".
Then, in step 105, the example sentence of second languages that will be corresponding and the translation combination of above-mentioned difference part with the example sentence of similar first languages, generate reference translation, wherein according to the position of difference part in the example sentence of second languages of correspondence of in step 103, determining, with the difference part in the example sentence of the translation replacement of described difference part second languages corresponding with the example sentence of similar first languages.
In this step, use " John " at the primary importance place in the Chinese example sentence of " Mary " replacement, and replace " important " of the 5th position in the Chinese example sentence, thereby obtain following reference translation with " little ".
At last, in step 106, utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages second languages corresponding with it, point out same section in the sentence of above-mentioned first languages to be translated and/or difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.Here, the method of prompting difference part and/or same section can be varied, for example utilize square frame, underscore, different colours, different fonts or its combination etc., show corresponding difference part with a kind of pattern, and showing corresponding same section with another kind of pattern, the present invention is to this not restriction.
Particularly, as shown in Figure 2, utilize the first pattern C1 to be presented at " John " and " Mary " in the reference translation in the right target language sentence of " John ", similar example sentence in the right source language sentence of " Mary " in the sentence to be translated, similar example sentence, utilize the second pattern C2 to be presented at " important " and " little " in the reference translation in the right target language sentence of " critical ", similar example sentence in the right source language sentence of " small " in the sentence to be translated, similar example sentence, and show corresponding same section with normal mode.
The method of the computer-aided translation by present embodiment is utilized the bilingual example sentence storehouse structure reference translation of having carried out alignment, can not comprise unwanted translation fragment in reference translation, thereby improve translation efficiency.In addition, for sentence to be translated, similar example sentence to reference translation in corresponding part show with same pattern, make the user very clear to composition, source and the translation quality of reference translation.Simultaneously, the user only need pay close attention to the similar example sentence centering part relevant with sentence to be translated, thereby has further improved translation efficiency.
In addition, the method for the computer-aided translation of present embodiment also allows the user that reference translation is edited, and for example increases, deletes, revises translation, or edits translation easily by directly pulling the translation fragment.When the user edits an any when part in the reference translation, its part in sentence to be translated, similar example sentence centering correspondence can be simultaneously highlighted.These will be described in detail with reference to figure 6 and 7 in the following embodiments.
In addition, the alignment information that the method for the computer-aided translation of present embodiment also allows the user to treat between translation of the sentence and the reference translation is set, and the reference translation after the user can being confirmed and sentence to be translated and the alignment information between them be saved in the above-mentioned bilingual example sentence storehouse, thereby expanded above-mentioned bilingual example sentence storehouse of having carried out alignment.
Under same inventive concept, Fig. 3 is the process flow diagram of the method for computer-aided translation according to another embodiment of the invention.Below just in conjunction with this figure, present embodiment is described.For those parts identical, suitably omit its explanation with front embodiment.
As shown in Figure 3, at first,, import the sentence of first languages to be translated in step 301.
For example:
There?is?a?red?jacket?on?the?bed.
Perhaps
Mary?played?a?small?role?in?the?drama.
Then, in step 302, sentence for first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, the bilingual example sentence storehouse of having carried out alignment is the bilingual example sentence storehouse of having been carried out word alignment by professional's (for example, the translator) craft or computing machine automatically, it comprise many to translation each other corresponding first languages and the alignment information between the example sentence of second languages and the every pair of example sentence.For example, the similar example sentence that finds for above-mentioned two examples to and alignment information as follows respectively.
Perhaps
Then, in step 303, judge that whether similarity between similar example sentence and the sentence to be translated is greater than pre-set threshold φ.
The method of calculating two character string similaritys is for example referring to Levenshtein, V. document " Binary codes capable of corresting deletions; insertions; and reversals. " Soviet Physics-Doklady 10,10 (1996), 707-710 is incorporated herein its whole contents as a reference.Wherein, utilize " editing distance " to weigh the similarity of two character strings.Editing distance is the number of operations from the needed minimum insertion of former character string converting into target character string, deletion and replacement.The similarity situation of how much determining two character strings by number of operations.If former character string and target string are in full accord, editing distance is zero so, otherwise editing distance is the number of times of described operation.
For example, if s=" test ", t=" test ", the editing distance between s and the t equals 0 so, because two sentences are just the same.If s=" test ", t=" tent ", the editing distance between s and the t equals 1 so, because once replacement operation " s " → " n " takes place.
Usually, use normalized editing distance as the similarity between two character strings:
d’(s，t)＝1-d(s，t)/max(|s|，|t|)
Wherein, d (s t) is editing distance, d ' (s t) is normalized editing distance, | s| is the length of character string s, and | t| is the length of character string t.The value of normalized editing distance is between 0-1, if former character string and target string are different fully, editing distance is 0 so, if former character string and target string are in full accord, editing distance is 1 so.In the case, the threshold value φ of similarity can be redefined for any numerical value between the 0-1, and for example 0.6, it can be set according to user's demand, does not limit at this.
Therefore, when sentence to be translated is " There is a red jacket on the bed. ", can calculate the editing distance d (s between itself and the similar example sentence " There is such a thing in agitation. ", t)=25, the length of sentence to be translated | s|=32, the length of similar example sentence | t|=34, then normalized editing distance d ' (s, t)=1-25/max (32,34)=0.265.
If φ=0.6, then the similarity between step 303 judgement sentence to be translated and similar example sentence is less than pre-set threshold φ, and method proceeds to step 304.
In step 304, utilize above-mentioned bilingual example sentence storehouse of having carried out alignment to obtain best example sentence to combination, wherein best example sentence comprises at least one fragment of the sentence of above-mentioned first languages to be translated to each the example sentence of first languages of combination.
Obtaining best example sentence in above-mentioned bilingual example sentence storehouse of having carried out alignment can have multiplely to the method for combination, present embodiment adopts dynamic programming algorithm, and concrete process flow diagram describes to obtain the detailed process of best example sentence to combination as shown in Figure 4 below with reference to Fig. 4.
Dynamic programming algorithm starts from step 401, for sentence to be translated from i
ThWord is to j
ThContinuous fragment [the w of word
i, w
j], in step 401, calculating comprises continuous fragment [w
i, w
j] example sentence to the combination score δ (i, j).
Then, in step 402, import continuous fragment [w
i, w
j].
Then, in step 403, maximum score MaxScore is set at 0.
Then, in step 404, judge in above-mentioned bilingual example sentence storehouse of carrying out alignment, whether to exist to comprise continuous fragment [w
i, w
j] example sentence right.
If being judged as in step 404 is that then method proceeds to step 405, wherein, calculates continuous fragment [w
i, w
j] weight Ψ (i, j is fre) as above-mentioned maximum score MaxScore.This weight is based on continuous fragment [w
i, w
j] the right alignment information of length, the number of times that in described bilingual example sentence storehouse, occurs or example sentence calculate.Wherein, continuous fragment [w
i, w
j] length long more, weight is big more; Continuous fragment [w
i, w
j] number of times that occurs in described bilingual example sentence storehouse is many more, weight is big more; And if in the fragment of the alignment information cutting example sentence right, comprise continuous fragment [w to acquisition according to example sentence
i, w
j], then weight is big, otherwise weight is little.
Then, in step 413, return MaxScore and example sentence is right, and finish in step 414.
Comprise continuous fragment [w if in step 404, can not find
i, w
j] example sentence right, then method proceeds to step 406.
In step 406, judge whether to exist m
i, m
jSatisfy i＜m
i＜m
j＜j.
If be judged as not in step 406, then finish in step 414.
If have m in step 406 judgement
i, m
jSatisfy i＜m
i＜m
j＜j is then with continuous fragment [w
i, w
j] be divided into continuous fragment [w
i, m
1], [m
1, m
2] and [m
2, w
j], and method proceeds to step 407.
In step 407, calculating comprises continuous fragment [w respectively
i, m
1] example sentence to the combination score δ (i, m
1), and comprise continuous fragment [m
2, w
j] example sentence to the combination score δ (m
2, j).
Then, in step 408, calculate penalty factor γ (m
1, m
2), there is not the word that covers for two overlapping or middle existence of fragment, γ will be to a minus penalty value so.
Then, in step 409, Score=δ (i, m count the score
1)+δ (m
2, j)+r (m
1, m
2).
Then, in step 410, judge that whether score Score is greater than maximum score MaxScore, if score Score is greater than maximum score MaxScore, then give maximum score MaxScore with score Score, and proceed to step 412, if score Score less than maximum score MaxScore, then is directly to step 412.
In step 412, judge whether to exist different m
i, m
jSatisfy i＜m
i＜m
j＜j if exist, then continues continuous fragment [w
i, w
j] cut apart, and turn back to step 407, if there is no, then method proceeds to step 413, returns MaxScore and example sentence is right, and finishes in step 414.
At last, the example sentence combination that score is the highest to combination, for example comprises three pairs of example sentences as required best example sentence:
First example sentence is right
Second example sentence is right
The 3rd example sentence is right
For sentence to be translated " There is a red jacket on the bed. ", " the There is " in the right source language sentence of first example sentence is identical with the fragment [1..2] of sentence to be translated, and its corresponding translation is " having "; " a red jacket " in the right source language sentence of second example sentence is identical with the fragment [3..5] of sentence to be translated, and its corresponding translation is " jacket of a redness "; " on the bed " in the right source language sentence of the 3rd example sentence is identical with the fragment [6..8] of sentence to be translated, and its corresponding translation is " in bed ".
Turn back to step 304, obtain best example sentence to after making up in above-mentioned bilingual example sentence storehouse of carrying out alignment, method proceeds to step 307, and in this step, to combination, the translation that system merges all fragment correspondences generates reference translation according to this best example sentence.That is to say that the translation that merges fragment [1..2], fragment [3..5] and the fragment [6..8] of sentence to be translated " has ", " jacket of a redness " and " in bed ", obtain reference translation and be the jacket of a redness " have in bed ".
At last,, utilize the right alignment information of above-mentioned a plurality of example sentence, point out each fragment in the sentence of above-mentioned first languages to be translated and the corresponding relation of the corresponding fragment in the above-mentioned reference translation in step 308.Here, the method for the corresponding relation of prompting respective segments can be varied, for example utilizes square frame, underscore, different colours, different fonts or its combination etc., shows the corresponding relation of each fragment respectively with different patterns, and the present invention is to this not restriction.
Particularly, as shown in Figure 5, utilize pattern C1 to show " Thereis " in the sentence to be translated respectively, " There is " in the right source language sentence of first example sentence, " having " and " the having " in the reference translation in right object statement of first example sentence, the pattern C2 of utilization shows " a red jacket " in the sentence to be translated, " a red jacket " in the right source language sentence of second example sentence, " jacket of a redness " and " jacket of a redness " in the reference translation in right object statement of second example sentence, the pattern C3 of utilization shows " the on the bed " in the sentence to be translated, " on the bed " in the right source language sentence of the 3rd example sentence, " in bed " and " in bed " in the reference translation in right object statement of the 3rd example sentence.
In addition, if right source language sentence and the target speech sentence of example sentence is long, example sentence centering does not have the part of reference value not to be revealed so, and for example, the first example sentence centering in Fig. 5 has only the previous section of example sentence to be revealed.If the user wishes to browse whole example sentence or understands the right alignment situation of example sentence, use the click example sentence so, this example sentence and alignment information are revealed, as shown in Figure 8.
In addition, the method for the computer-aided translation of present embodiment also allows the user that reference translation is edited, and for example increases, deletes, revises translation, or edits translation easily by directly pulling the translation fragment, as shown in Figure 6.When the user edits an any when part in the reference translation, its part corresponding in sentence to be translated, similar example sentence can be simultaneously highlighted, as shown in Figure 7.
Particularly, Fig. 6 is the process flow diagram that shows and edit according to the embodiment of Fig. 3, in Fig. 6, unit 608 is fragment index tables, and this fragment index table has been stored each fragment corresponding word (sentence in Fig. 5 in this table) in its sentence.Unit 601 is operations of click fragment; Unit 602 is operations that the user edits fragment in the reference translation; Fragment in the unit 603 expressions sentence to be translated; The fragment of unit 604 expression example sentence centerings; Fragment in the unit 605 expression reference translations; Unit 606 is that homologous segment is searched module; Unit 607 is that fragment highlights module.When user's performance element 601 was clicked in the unit 603,604,605 any one or performance element 602 edit cells 605, the information of pairing other associated clip of operated fragment was searched in unit 606 in unit 608.Then unit 607 sentence to be translated, example sentence to reference translation in highlight operated fragment and its corresponding fragment.
Fig. 7 is the synoptic diagram that highlights edit segment according to the embodiment of Fig. 3, in Fig. 7, the pattern C4 of utilization highlights " in bed " and " in bed " in the reference translation in right object statement of " on the bed ", the 3rd example sentence in the right source language sentence of " on the bed ", the 3rd example sentence in the sentence to be translated.
Fig. 8 is the synoptic diagram according to the right alignment information of the example sentence of the embodiment of Fig. 3, in Fig. 8, show among Fig. 5 first example sentence to and alignment information.
Turn back to step 303, when sentence to be translated is " Mary played a small role in thedrama. ", can calculate the editing distance d (s between itself and the similar example sentence " John played a critical role in thedrama. ", t)=12, the length of sentence to be translated | s|=38, the length of similar example sentence | t|=41, then normalized editing distance d ' (s, t)=1-12/max (38,41)=0.707.
If φ=0.6, then the similarity between step 303 judgement sentence to be translated and similar example sentence is greater than pre-set threshold φ, and method proceeds to step 305.
In step 305, the example sentence of the sentence of first languages to be translated and similar first languages is compared, if the sentence of first languages to be translated is incomplete same with the example sentence of the first similar languages, find out the difference part between them, and can utilize word alignment information between the example sentence of the example sentence of the first similar languages and second languages corresponding with it, determine described difference part position in the example sentence of second languages of correspondence.
For above-mentioned example, the difference between sentence to be translated and the similar example sentence is Mary and John, and small and critical.These two positions of difference part in the Chinese example sentence corresponding with similar example sentence are respectively primary importance and the 5th position.
Then, in step 306, for corresponding translation is partly prepared in the difference between the example sentence of the sentence of first languages to be translated and similar first languages.Here, can search the translation of described difference part in dictionary, or preferably search from above-mentioned bilingual example sentence storehouse, the present invention is to this not restriction.
Particularly, the translation that can obtain Mary and small is respectively " Mary " and " little ".
Then, in step 307, the example sentence of second languages that will be corresponding and the translation combination of above-mentioned difference part with the example sentence of similar first languages, generate reference translation, wherein according to the position of difference part in the example sentence of second languages of correspondence of in step 305, determining, with the difference part in the example sentence of the translation replacement of described difference part second languages corresponding with the example sentence of similar first languages.
In this step, use " John " at the primary importance place in the Chinese example sentence of " Mary " replacement, and replace " important " of the 5th position in the Chinese example sentence, thereby obtain following reference translation with " little ".
At last, in step 308, utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages second languages corresponding with it, point out same section in the sentence of above-mentioned first languages to be translated and/or difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.Here, the method of prompting difference part and/or same section can be varied, for example utilize square frame, underscore, different colours, different fonts or its combination etc., show corresponding difference part with a kind of pattern, and showing corresponding same section with another kind of pattern, the present invention is to this not restriction.
Particularly, as shown in Figure 2, utilize the first pattern C1 to be presented at " John " and " Mary " in the reference translation in the right target language sentence of " John ", similar example sentence in the right source language sentence of " Mary " in the sentence to be translated, similar example sentence, utilize the second pattern C2 to be presented at " important " and " little " in the reference translation in the right target language sentence of " critical ", similar example sentence in the right source language sentence of " small " in the sentence to be translated, similar example sentence, and show corresponding same section with normal mode.
The method of the computer-aided translation by present embodiment, the bilingual example sentence storehouse structure reference translation of alignment has been carried out in utilization, in reference translation, do not comprise unwanted translation fragment, and can in bilingual example sentence storehouse, can not find under the situation of similar example sentence, utilize example sentence composite construction reference translation, thereby improved translation efficiency.In addition, for sentence to be translated, example sentence to reference translation in corresponding part show with same pattern, make the user very clear to composition, source and the translation quality of reference translation.Simultaneously, the user only need pay close attention to the similar example sentence centering part relevant with sentence to be translated, thereby has further improved translation efficiency.
In addition, the method for the computer-aided translation of present embodiment also allows the user that reference translation is edited, and for example increases, deletes, revises translation, or edits translation easily by directly pulling the translation fragment.When the user edits an any when part in the reference translation, its part in sentence to be translated, example sentence centering correspondence can be simultaneously highlighted.
In addition, the alignment information that the method for the computer-aided translation of present embodiment also allows the user to treat between translation of the sentence and the reference translation is set, and the reference translation after the user can being confirmed and sentence to be translated and the alignment information between them be saved in the above-mentioned bilingual example sentence storehouse, thereby expanded above-mentioned bilingual example sentence storehouse of having carried out alignment.
Under same inventive concept, Fig. 9 is the block scheme of the device of computer-aided translation according to another embodiment of the invention.Below just in conjunction with this figure, present embodiment is described.For those parts identical, suitably omit its explanation with front embodiment.
The computer-aided translation device 900 of present embodiment comprises: search unit (searching unit) 901, be used to the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence; Translation unit (translating unit) 902, be used for when the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation respectively; Assembled unit (combining unit) 903 is used for the translation combination with each of the example sentence of second languages corresponding with the example sentence of above-mentioned similar first languages and above-mentioned at least one difference part, generates reference translation; And Tip element (indicating unit) 904, be used to utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
Preferably, above-mentioned translation unit 902 is configured to search the translation of partly each of above-mentioned at least one difference from above-mentioned bilingual example sentence storehouse.
Preferably, Tip element 904 comprises display unit (displaying unit), be used to utilize square frame, underscore, different colours, different fonts or its combination etc., show corresponding difference part with a kind of pattern, and showing corresponding same section with another kind of pattern, the present invention is to this not restriction.
Particularly, as shown in Figure 2, utilize the first pattern C1 to be presented at " John " and " Mary " in the reference translation in the right target language sentence of " John ", similar example sentence in the right source language sentence of " Mary " in the sentence to be translated, similar example sentence, utilize the second pattern C2 to be presented at " important " and " little " in the reference translation in the right target language sentence of " critical ", similar example sentence in the right source language sentence of " small " in the sentence to be translated, similar example sentence, and show corresponding same section with normal mode.
By using the computer-aided translation device 900 of present embodiment, utilize the bilingual example sentence storehouse structure reference translation of having carried out alignment, in reference translation, can not comprise unwanted translation fragment, thereby improve translation efficiency.In addition, for sentence to be translated, similar example sentence to reference translation in corresponding part, utilize display unit to show with same pattern, make the user very clear to composition, source and the translation quality of reference translation.Simultaneously, the user only need pay close attention to the similar example sentence centering part relevant with sentence to be translated, thereby has further improved translation efficiency.
In addition, computer-aided translation device 900 can also comprise edit cell (editing unit), be used to allow the user to edit the reference translation that generates by combinations thereof unit 903, for example increase, delete, revise translation, or edit translation easily by directly pulling the translation fragment.When the user edits an any when part in the reference translation, its part in sentence to be translated, similar example sentence centering correspondence can be highlighted by above-mentioned display unit simultaneously.
In addition, computer-aided translation device 900 can also comprise setup unit (setting unit) and preserve unit (storing unit), the alignment information that is used to allow the user to treat between translation of the sentence and the reference translation is set, and the reference translation after the user can being confirmed and sentence to be translated and the alignment information between them be saved in the above-mentioned bilingual example sentence storehouse, thereby expanded above-mentioned bilingual example sentence storehouse of having carried out alignment.
Under same inventive concept, Figure 10 is the block scheme of the device of computer-aided translation according to another embodiment of the invention.Below just in conjunction with this figure, present embodiment is described.For those parts identical, suitably omit its explanation with front embodiment.
The computer-aided translation device 1000 of present embodiment comprises: search unit (searching unit) 1001, be used to the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence; Judging unit (determining unit) 1002, whether the similarity of sentence that is used to judge the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated is less than a predetermined threshold; Assembled unit (combining unit) 1003, when being used for similarity when the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated less than above-mentioned predetermined threshold, utilize the right combination of a plurality of example sentences in the above-mentioned bilingual example sentence storehouse of having carried out alignment to obtain reference translation, the example sentence of first languages of each that wherein above-mentioned a plurality of example sentences are right comprises at least one fragment of the sentence of above-mentioned first languages to be translated; And Tip element (indicating unit) 1004, be used to utilize the right alignment information of above-mentioned a plurality of example sentence, point out each fragment in the sentence of above-mentioned first languages to be translated and the corresponding relation of the corresponding fragment in the above-mentioned reference translation.
Preferably, above-mentioned Tip element 1004 comprises display unit (displaying unit), be used to utilize square frame, underscore, different colours, different fonts or its combination etc., show the corresponding relation of each fragment respectively with different patterns, the present invention is to this not restriction.
Particularly, as shown in Figure 5, utilize pattern C1 to show " Thereis " in the sentence to be translated respectively, " There is " in the right source language sentence of first example sentence, " having " and " the having " in the reference translation in right object statement of first example sentence, the pattern C2 of utilization shows " a red jacket " in the sentence to be translated, " a red jacket " in the right source language sentence of second example sentence, " jacket of a redness " and " jacket of a redness " in the reference translation in right object statement of second example sentence, the pattern C3 of utilization shows " the on the bed " in the sentence to be translated, " on the bed " in the right source language sentence of the 3rd example sentence, " in bed " and " in bed " in the reference translation in right object statement of the 3rd example sentence.
In addition, if right source language sentence and the target speech sentence of example sentence is long, example sentence centering does not have the part of reference value not to be revealed so, and for example, the first example sentence centering in Fig. 5 has only the previous section of example sentence to be revealed.If the user wishes to browse whole example sentence or understands the right alignment situation of example sentence, use the click example sentence so, this example sentence and alignment information are revealed, as shown in Figure 8.
Preferably, combinations thereof unit 1003 can also comprise cutting unit (dividing unit) and computing unit (calculating unit), and be configured to utilize dynamic programming algorithm to obtain the right combination of above-mentioned a plurality of example sentence, detailed process is identical with above-mentioned embodiment with reference to figure 4, is not giving unnecessary details at this.
Preferably, the computer-aided translation device 1000 of present embodiment also comprises: translation unit (translating unit), be used for similarity when the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated greater than above-mentioned predetermined threshold, and when the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation; Combinations thereof unit 1003 further is configured to, and the translation combination of the example sentence of second languages that will be corresponding with the example sentence of above-mentioned similar first languages and each of above-mentioned at least one difference part generates reference translation; And above-mentioned Tip element 1004 further is configured to, utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
Preferably, above-mentioned translation unit further is configured to search the translation of above-mentioned difference part from above-mentioned bilingual example sentence storehouse.
Preferably, above-mentioned Tip element 1004 comprises display unit (displaying unit), be used to utilize square frame, underscore, different colours, different fonts or its combination etc., show corresponding difference part with a kind of pattern, and showing corresponding same section with another kind of pattern, the present invention is to this not restriction.
Particularly, as shown in Figure 2, utilize the first pattern C1 to be presented at " John " and " Mary " in the reference translation in the right target language sentence of " John ", similar example sentence in the right source language sentence of " Mary " in the sentence to be translated, similar example sentence, utilize the second pattern C2 to be presented at " important " and " little " in the reference translation in the right target language sentence of " critical ", similar example sentence in the right source language sentence of " small " in the sentence to be translated, similar example sentence, and show corresponding same section with normal mode.
Computer-aided translation device 1000 by present embodiment, the bilingual example sentence storehouse structure reference translation of alignment has been carried out in utilization, in reference translation, do not comprise unwanted translation fragment, and can in bilingual example sentence storehouse, can not find under the situation of similar example sentence, utilize example sentence composite construction reference translation, thereby improved translation efficiency.In addition, for sentence to be translated, example sentence to reference translation in corresponding part, utilize display unit to show with same pattern, make the user very clear to composition, source and the translation quality of reference translation.Simultaneously, the user only need pay close attention to the similar example sentence centering part relevant with sentence to be translated, thereby has further improved translation efficiency.
In addition, the computer-aided translation device 1000 of present embodiment can also comprise edit cell (editing unit), be used to allow the user that reference translation is edited, for example increase, delete, revise translation, or edit translation easily by directly pulling the translation fragment.When the user edits an any when part in the reference translation, its part in sentence to be translated, example sentence centering correspondence can be highlighted by above-mentioned display unit simultaneously.
In addition, the computer-aided translation device 1000 of present embodiment can also comprise setup unit (setting unit) and preserve unit (storing unit), the alignment information that is used to allow the user to treat between translation of the sentence and the reference translation is set, and the reference translation after the user can being confirmed and sentence to be translated and the alignment information between them be saved in the above-mentioned bilingual example sentence storehouse, thereby expanded above-mentioned bilingual example sentence storehouse of having carried out alignment.
Though more than described the method for computer-aided translation of the present invention and the device of computer-aided translation in detail by some exemplary embodiments, but above these embodiment are not exhaustive, and those skilled in the art can realize variations and modifications within the spirit and scope of the present invention.Therefore, the present invention is not limited to these embodiment, and scope of the present invention only is as the criterion by claims.
Claims (38)
1. the method for a computer-aided translation may further comprise the steps:
Sentence for first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence;
When the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation respectively;
The translation combination of the example sentence of second languages that will be corresponding with the example sentence of above-mentioned similar first languages and each of above-mentioned at least one difference part generates reference translation; And
Utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
2. the method for computer-aided translation according to claim 1, wherein, above-mentioned prompting step comprise above-mentioned at least one same section in the sentence that shows above-mentioned first languages to be translated respectively accordingly and/or above-mentioned at least one difference part each with and part in above-mentioned reference translation.
3. the method for computer-aided translation according to claim 1 and 2 further comprises: allow the user to edit above-mentioned reference translation.
4. according to the method for any one described computer-aided translation among the claim 1-3, further comprise: allow the user to set the sentence of above-mentioned first languages to be translated and the alignment information between the above-mentioned reference translation.
5. according to the method for any one described computer-aided translation among the claim 1-4, further comprise: the sentence and the alignment information between them of the reference translation of second languages after the user is confirmed and above-mentioned first languages to be translated are saved in the above-mentioned bilingual example sentence storehouse.
6. according to the method for any one described computer-aided translation among the claim 1-5, wherein, the step of preparing corresponding translation respectively for each of above-mentioned at least one difference part comprises and searching from above-mentioned bilingual example sentence storehouse and this difference translation of the second corresponding languages partly.
7. the method for a computer-aided translation may further comprise the steps:
Sentence for first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence;
Whether the similarity of sentence of judging the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated is less than a predetermined threshold;
When the similarity of the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated during less than above-mentioned predetermined threshold, utilize the right combination of a plurality of example sentences in the above-mentioned bilingual example sentence storehouse of having carried out alignment to obtain reference translation, the example sentence of first languages of each that wherein above-mentioned a plurality of example sentences are right comprises at least one fragment of the sentence of above-mentioned first languages to be translated; And
Utilize the right alignment information of above-mentioned a plurality of example sentence, point out each fragment in the sentence of above-mentioned first languages to be translated and the corresponding relation of the corresponding fragment in the above-mentioned reference translation.
8. the method for computer-aided translation according to claim 7, wherein, above-mentioned prompting step comprises each fragment and its fragment in above-mentioned reference translation in the sentence that shows above-mentioned first languages to be translated respectively accordingly.
9. according to the method for claim 7 or 8 described computer-aided translation, wherein, utilize dynamic programming algorithm to obtain the right combination of above-mentioned a plurality of example sentence.
10. the method for computer-aided translation according to claim 9, wherein, above-mentioned dynamic programming algorithm may further comprise the steps:
The sentence of above-mentioned first languages to be translated is divided into a plurality of fragments; And
A plurality of fragments of cutting apart for each is above-mentioned, it is right to search the example sentence that comprises this fragment in above-mentioned bilingual example sentence storehouse.
11. the method for computer-aided translation according to claim 10, wherein, above-mentioned dynamic programming algorithm is further comprising the steps of:
If in above-mentioned bilingual example sentence storehouse, do not comprise that the example sentence of above-mentioned fragment is right, above-mentioned fragment be divided into a plurality of fragments once more; And
A plurality of fragments of cutting apart once more for each is above-mentioned, it is right to search the example sentence that comprises this fragment in above-mentioned bilingual example sentence storehouse.
12., further comprising the steps of according to the method for claim 10 or 11 described computer-aided translation:
Calculate each weight of above-mentioned a plurality of fragments;
According to the weight of aforementioned calculation, calculate the score of the right combination of above-mentioned a plurality of example sentence;
Wherein, the right combination of example sentence that score is the highest is as the right combination of above-mentioned a plurality of example sentences.
13. the method for computer-aided translation according to claim 12, wherein, the step of the weight of each of the above-mentioned a plurality of fragments of described calculating is carried out based on the length of each fragment, the number of times that occurs in described bilingual example sentence storehouse or the right alignment information of above-mentioned a plurality of example sentence.
14. the method for computer-aided translation according to claim 7 also comprises:
When the similarity of the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated greater than above-mentioned predetermined threshold, and when the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation;
The translation combination of the example sentence of second languages that will be corresponding with the example sentence of above-mentioned similar first languages and each of above-mentioned at least one difference part generates reference translation; And
Utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
15. the method for computer-aided translation according to claim 14, wherein, above-mentioned prompting step comprise above-mentioned at least one same section in the sentence that shows above-mentioned first languages to be translated respectively accordingly and/or above-mentioned at least one difference part each with and part in above-mentioned reference translation.
16. method according to claim 14 or 15 described computer-aided translation, wherein, the step of preparing corresponding translation respectively for each of above-mentioned at least one difference part comprises and searching from above-mentioned bilingual example sentence storehouse and this difference translation of the second corresponding languages partly.
17. the method according to any one described computer-aided translation among the claim 7-16 further comprises: allow the user to edit above-mentioned reference translation.
18. the method according to any one described computer-aided translation among the claim 7-17 further comprises: allow the user to set the sentence of above-mentioned first languages to be translated and the alignment information between the above-mentioned reference translation.
19. the method according to any one described computer-aided translation among the claim 7-18 further comprises: the sentence and the alignment information between them of the reference translation of second languages after the user confirmed and above-mentioned first languages to be translated are saved in the above-mentioned bilingual example sentence storehouse.
20. the device of a computer-aided translation comprises:
Search unit (searching unit), be used to the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence;
Translation unit (translating unit), be used for when the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation respectively;
Assembled unit (combining unit) is used for the translation combination with each of the example sentence of second languages corresponding with the example sentence of above-mentioned similar first languages and above-mentioned at least one difference part, generates reference translation; And
Tip element (indicating unit), be used to utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
21. the device of computer-aided translation according to claim 20, wherein, above-mentioned Tip element comprises display unit (displaying unit), each that is used for showing accordingly respectively above-mentioned at least one same sections of sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and part in above-mentioned reference translation.
22. according to the device of claim 20 or 21 described computer-aided translation, also comprise edit cell (editing unit), be used to allow the user to edit the reference translation that generates by the combinations thereof unit.
23. device according to any one described computer-aided translation among the claim 20-22, also comprise setup unit (setting unit), be used to allow the user to set the sentence of above-mentioned first languages to be translated and the alignment information between the above-mentioned reference translation.
24. device according to any one described computer-aided translation among the claim 20-23, comprise also and preserve unit (storing unit) that the sentence and the alignment information between them of the reference translation of second languages after being used for the user confirmed and above-mentioned first languages to be translated are saved in above-mentioned bilingual example sentence storehouse.
25. according to the device of any one described computer-aided translation among the claim 20-24, wherein, above-mentioned translation unit is configured to search the translation of partly each of above-mentioned at least one difference from above-mentioned bilingual example sentence storehouse.
26. the device of a computer-aided translation comprises:
Search unit (searching unit), be used to the sentence of first languages to be translated, in the bilingual example sentence storehouse of having carried out alignment, search the example sentence of the first similar languages, wherein, above-mentioned bilingual example sentence storehouse of having carried out alignment comprise many to corresponding first languages and second languages example sentence and the alignment information between the every pair of example sentence;
Judging unit (determining unit), whether the similarity of sentence that is used to judge the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated is less than a predetermined threshold;
Assembled unit (combining unit), when being used for similarity when the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated less than above-mentioned predetermined threshold, utilize the right combination of a plurality of example sentences in the above-mentioned bilingual example sentence storehouse of having carried out alignment to obtain reference translation, the example sentence of first languages of each that wherein above-mentioned a plurality of example sentences are right comprises at least one fragment of the sentence of above-mentioned first languages to be translated; And
Tip element (indicating unit) is used to utilize the right alignment information of above-mentioned a plurality of example sentence, points out each fragment in the sentence of above-mentioned first languages to be translated and the corresponding relation of the corresponding fragment in the above-mentioned reference translation.
27. the device of computer-aided translation according to claim 26, wherein, above-mentioned Tip element comprises display unit (displaying unit), is used for showing accordingly respectively each fragment and its fragment in above-mentioned reference translation of the sentence of above-mentioned first languages to be translated.
28. according to the device of claim 26 or 27 described computer-aided translation, wherein, the combinations thereof configuration of cells is to utilize dynamic programming algorithm to obtain the right combination of above-mentioned a plurality of example sentence.
29. the device of computer-aided translation according to claim 28, wherein, the combinations thereof unit also comprises cutting unit (dividing unit), and the sentence of above-mentioned first languages to be translated is divided into a plurality of fragments; And
The above-mentioned unit of searching further is configured to, a plurality of fragments of cutting apart for each is above-mentioned, and it is right to search the example sentence that comprises this fragment in above-mentioned bilingual example sentence storehouse, to obtain the right combination of above-mentioned a plurality of example sentence.
30. the device of computer-aided translation according to claim 29, wherein, above-mentioned cutting unit further is configured to, if do not comprise that the example sentence of above-mentioned fragment is right in above-mentioned bilingual example sentence storehouse, above-mentioned fragment is divided into a plurality of fragments once more; And
The above-mentioned unit of searching further is configured to, a plurality of fragments of cutting apart once more for each is above-mentioned, and it is right to search the example sentence that comprises this fragment in above-mentioned bilingual example sentence storehouse.
31. according to the device of claim 29 or 30 described computer-aided translation, wherein, the combinations thereof unit also comprises computing unit (calculating unit), calculates each weight of above-mentioned a plurality of fragments; And
The aforementioned calculation unit further is configured to, and according to the weight of aforementioned calculation, calculates the score of the right combination of above-mentioned a plurality of example sentence;
Wherein, the right combination of example sentence that score is the highest is as the right combination of above-mentioned a plurality of example sentences.
32. the device of computer-aided translation according to claim 31, wherein, the aforementioned calculation unit calculates above-mentioned weight based on the length of each fragment, the number of times that occurs or the right alignment information of above-mentioned a plurality of example sentence in described bilingual example sentence storehouse.
33. the device of computer-aided translation according to claim 26 also comprises:
Translation unit (translating unit), be used for similarity when the sentence of the example sentence of above-mentioned similar first languages and above-mentioned first languages to be translated greater than above-mentioned predetermined threshold, and when the sentence of comparing above-mentioned first languages to be translated with the example sentence of above-mentioned similar first languages has at least one same section and at least one difference part, for each of above-mentioned at least one difference part is prepared corresponding translation;
The combinations thereof unit further is configured to, and the translation combination of the example sentence of second languages that will be corresponding with the example sentence of above-mentioned similar first languages and each of above-mentioned at least one difference part generates reference translation; And
Above-mentioned Tip element further is configured to, utilize the alignment information of the example sentence of the example sentence of above-mentioned similar first languages and second languages corresponding with it, each that point out above-mentioned at least one same section in the sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and in above-mentioned reference translation the corresponding relation of corresponding part.
34. the device of computer-aided translation according to claim 33, wherein, above-mentioned Tip element comprises display unit (displaying unit), each that is used for showing accordingly respectively above-mentioned at least one same sections of sentence of above-mentioned first languages to be translated and/or above-mentioned at least one difference part with and part in above-mentioned reference translation.
35. according to the device of claim 33 or 34 described computer-aided translation, wherein, above-mentioned translation unit is configured to search the translation of above-mentioned difference part from above-mentioned bilingual example sentence storehouse.
36. according to the device of any one described computer-aided translation among the claim 26-35, also comprise edit cell (editing unit), be used to allow the user to edit the reference translation that generates by the combinations thereof unit.
37. device according to any one described computer-aided translation among the claim 26-36, also comprise setup unit (setting unit), be used to allow the user to set the sentence of above-mentioned first languages to be translated and the alignment information between the above-mentioned reference translation.
38. device according to any one described computer-aided translation among the claim 26-37, comprise also and preserve unit (storing unit) that the sentence and the alignment information between them of the reference translation of second languages after being used for the user confirmed and above-mentioned first languages to be translated are saved in above-mentioned bilingual example sentence storehouse.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007100891631A CN101271451A (en) | 2007-03-20 | 2007-03-20 | Computer aided translation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007100891631A CN101271451A (en) | 2007-03-20 | 2007-03-20 | Computer aided translation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101271451A true CN101271451A (en) | 2008-09-24 |
Family
ID=40005430
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2007100891631A Pending CN101271451A (en) | 2007-03-20 | 2007-03-20 | Computer aided translation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101271451A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102193914A (en) * | 2011-05-26 | 2011-09-21 | 中国科学院计算技术研究所 | Computer aided translation method and system |
| CN102236643A (en) * | 2010-04-29 | 2011-11-09 | 英业达股份有限公司 | Processing system and method capable of simultaneously displaying multilingual words and example sentences |
| CN102541821A (en) * | 2010-12-30 | 2012-07-04 | 微软公司 | Displaying method and system with notes for interactive multiple languages |
| CN102929865A (en) * | 2012-10-12 | 2013-02-13 | 广西大学 | PDA (Personal Digital Assistant) translation system for inter-translating Chinese and languages of ASEAN (the Association of Southeast Asian Nations) countries |
| CN103034627A (en) * | 2011-10-09 | 2013-04-10 | 北京百度网讯科技有限公司 | Method and device for calculating sentence similarity and method and device for machine translation |
| CN103530284A (en) * | 2013-09-22 | 2014-01-22 | 中国专利信息中心 | Short sentence segmenting device, machine translation system and corresponding segmenting method and translation method |
| CN104750820A (en) * | 2015-04-24 | 2015-07-01 | 中译语通科技（北京）有限公司 | Filtering method and device for corpuses |
| CN104750687A (en) * | 2013-12-25 | 2015-07-01 | 株式会社东芝 | Method for improving bilingual corpus, device for improving bilingual corpus, machine translation method and machine translation device |
| CN104933038A (en) * | 2014-03-20 | 2015-09-23 | 株式会社东芝 | Machine translation method and machine translation device |
| CN105183723A (en) * | 2015-09-17 | 2015-12-23 | 成都优译信息技术有限公司 | Associating method for translation software and language material searching |
| CN102541821B (en) * | 2010-12-30 | 2016-12-14 | 微软技术许可有限责任公司 | The band note display method of interactive multiple languages and system |
| CN107818086A (en) * | 2016-09-13 | 2018-03-20 | 株式会社东芝 | Machine translation method and device |
| CN108351871A (en) * | 2015-09-22 | 2018-07-31 | 脸谱公司 | General translator |
-
2007
- 2007-03-20 CN CNA2007100891631A patent/CN101271451A/en active Pending
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102236643A (en) * | 2010-04-29 | 2011-11-09 | 英业达股份有限公司 | Processing system and method capable of simultaneously displaying multilingual words and example sentences |
| CN102541821A (en) * | 2010-12-30 | 2012-07-04 | 微软公司 | Displaying method and system with notes for interactive multiple languages |
| CN102541821B (en) * | 2010-12-30 | 2016-12-14 | 微软技术许可有限责任公司 | The band note display method of interactive multiple languages and system |
| CN102193914A (en) * | 2011-05-26 | 2011-09-21 | 中国科学院计算技术研究所 | Computer aided translation method and system |
| CN103034627B (en) * | 2011-10-09 | 2016-05-25 | 北京百度网讯科技有限公司 | Calculate the method and apparatus of sentence similarity and the method and apparatus of machine translation |
| CN103034627A (en) * | 2011-10-09 | 2013-04-10 | 北京百度网讯科技有限公司 | Method and device for calculating sentence similarity and method and device for machine translation |
| CN102929865A (en) * | 2012-10-12 | 2013-02-13 | 广西大学 | PDA (Personal Digital Assistant) translation system for inter-translating Chinese and languages of ASEAN (the Association of Southeast Asian Nations) countries |
| CN103530284A (en) * | 2013-09-22 | 2014-01-22 | 中国专利信息中心 | Short sentence segmenting device, machine translation system and corresponding segmenting method and translation method |
| CN103530284B (en) * | 2013-09-22 | 2016-07-06 | 中国专利信息中心 | Short sentence cutting device, machine translation system and corresponding cutting method and interpretation method |
| CN104750687A (en) * | 2013-12-25 | 2015-07-01 | 株式会社东芝 | Method for improving bilingual corpus, device for improving bilingual corpus, machine translation method and machine translation device |
| CN104750687B (en) * | 2013-12-25 | 2018-03-20 | 株式会社东芝 | Improve method and device, machine translation method and the device of bilingualism corpora |
| US10061768B2 (en) | 2013-12-25 | 2018-08-28 | Kabushiki Kaisha Toshiba | Method and apparatus for improving a bilingual corpus, machine translation method and apparatus |
| CN104933038A (en) * | 2014-03-20 | 2015-09-23 | 株式会社东芝 | Machine translation method and machine translation device |
| CN104750820A (en) * | 2015-04-24 | 2015-07-01 | 中译语通科技（北京）有限公司 | Filtering method and device for corpuses |
| CN105183723A (en) * | 2015-09-17 | 2015-12-23 | 成都优译信息技术有限公司 | Associating method for translation software and language material searching |
| CN108351871A (en) * | 2015-09-22 | 2018-07-31 | 脸谱公司 | General translator |
| US10346537B2 (en) | 2015-09-22 | 2019-07-09 | Facebook, Inc. | Universal translation |
| CN108351871B (en) * | 2015-09-22 | 2019-09-27 | 脸谱公司 | General translator |
| CN107818086A (en) * | 2016-09-13 | 2018-03-20 | 株式会社东芝 | Machine translation method and device |
| CN107818086B (en) * | 2016-09-13 | 2021-08-10 | 株式会社东芝 | Machine translation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101271451A (en) | Computer aided translation method and device | |
| CN101271452B (en) | Method and device for generating version and machine translation | |
| Hasler et al. | Source sentence simplification for statistical machine translation | |
| KR101266361B1 (en) | Automatic translation system based on structured translation memory and automatic translating method using the same | |
| JP2006012168A (en) | Method for improving coverage and quality in translation memory system | |
| US20130262077A1 (en) | Machine translation device, machine translation method, and recording medium storing machine translation program | |
| US20100228538A1 (en) | Computational linguistic systems and methods | |
| US20040243403A1 (en) | Document relationship inspection apparatus, translation process apparatus, document relationship inspection method, translation process method, and document relationship inspection program | |
| Melby et al. | Translation memory | |
| JP4113235B2 (en) | Translation support device | |
| Dandapat et al. | Using example-based MT to support statistical MT when translating homogeneous data in a resource-poor setting | |
| CN111680523B (en) | Man-machine collaborative translation system and method based on context semantic comparison | |
| Tambouratzis et al. | Language-independent hybrid MT with PRESEMT | |
| JPH10312382A (en) | Similar example translation system | |
| JP2018072979A (en) | Parallel translation sentence extraction device, parallel translation sentence extraction method and program | |
| Muntés Mulero et al. | Context-aware machine translation for software localization | |
| JP2009289219A (en) | Translation-memory translation device and translation program | |
| CN107967303B (en) | Corpus display method and apparatus | |
| JPH04160473A (en) | Method and device for example reuse type translation | |
| JP5207016B2 (en) | Machine translation evaluation apparatus and method | |
| JPH10283368A (en) | Information processor and method therefor | |
| JP5909123B2 (en) | Machine translation apparatus, machine translation method and program | |
| JP2018055328A (en) | Parallel translation sentence extracting device, parallel translation sentence extracting method and program | |
| JP3999771B2 (en) | Translation support program, translation support apparatus, and translation support method | |
| JPH08329059A (en) | General purpose reference device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20080924 |