CN110175337A - A kind of textual presentation method and device - Google Patents

A kind of textual presentation method and device Download PDF

Info

Publication number
CN110175337A
CN110175337A CN201910457348.6A CN201910457348A CN110175337A CN 110175337 A CN110175337 A CN 110175337A CN 201910457348 A CN201910457348 A CN 201910457348A CN 110175337 A CN110175337 A CN 110175337A
Authority
CN
China
Prior art keywords
text
unit
result
cypher
machine
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201910457348.6A
Other languages
Chinese (zh)
Other versions
CN110175337B (en
Inventor
王影
乔玉平
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
iFlytek Co Ltd
Original Assignee
iFlytek Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by iFlytek Co Ltd filed Critical iFlytek Co Ltd
Priority to CN201910457348.6A priority Critical patent/CN110175337B/en
Publication of CN110175337A publication Critical patent/CN110175337A/en
Application granted granted Critical
Publication of CN110175337B publication Critical patent/CN110175337B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/40—Processing or translation of natural language
    • G06F40/58—Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/26—Speech to text systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Human Computer Interaction (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Machine Translation (AREA)

Abstract

This application discloses a kind of textual presentation method and devices, this method comprises: after showing the machine translation result of target voice, translator is obtained in real time to the translation result of voice unit in target voice, as current cypher text, it matches and is obtained with the current cypher text in semantically immediate cypher text from machine translation result, as matching cypher text, and the exhibition method of matching cypher text is changed, so that translator has translated content and untranslated content carries out difference displaying.Wherein, translator's each voice unit in special translating purpose voice in real time, the change of mode can be so shown one by one to machine translation result corresponding to each voice unit according to the translation progress of translator, convenient for translator according to the machine translation result of current presentation, accurately know its own translation progress, enable translator to be absorbed in untranslated content, is conducive to the subsequent translation to untranslated content of translator.

Description

A kind of textual presentation method and device
Technical field
This application involves field of computer technology more particularly to a kind of textual presentation method and devices.
Background technique
With the propulsion of economic development and globalization, demand of the various countries to interpretation occupation is increasing, and interpretation market is also fluffy The exhibition of breaking out.Currently, a part accompanies interpretation and simple alternative interpretation scene can be competent at by machine translation, however it is some important , sensibility, accurate meeting occasion, such as the occasion answered reporters' questions still be unable to do without meeting alternative interpretation.
Meeting alternative interpretation be one worker's psychological quality, language grounding in basic skills, Listening Ability of Ethnic, memory are required it is very high Work.Operating mode is generally speaker and gives a lecture several minutes, and alternative interpretation interpreter remembers speaker by the notes of oneself and says Then content, logic and the intention of words are come out with another kind language expression.Alternative interpretation interpreter needs to complete the work of three steps Make: understanding and analysis source language, records the note and the externally output speech content of speaker (namely translate), and preceding two parts It is completed at the same time, workload is very big.
Although passing through speech recognition under the relevant technologies auxiliary, it is possible to reduce the notes of interpreter are born, and are turned over by machine It translates, reference Help can be provided for interpreter, promote translation skill.But when how effectively to show that it is translated to interpreter Required information is a urgent problem to be solved.
Summary of the invention
The main purpose of the embodiment of the present application is to provide a kind of textual presentation method and device, by the machine processing of voice As a result, the translation progress according to translator carries out tracking displaying, so that supplementary translation personnel are quickly concerned about machine processing As a result the required content in.
The embodiment of the present application provides a kind of textual presentation method, comprising:
Show the machine translation result of target voice；
Translator is obtained in real time to the translation result of the voice unit in the target voice, as current translation text This；
Obtain matching cypher text, it is described matching cypher text be from the machine translation result matching obtain with institute Current cypher text is stated in semantically immediate cypher text；
Change the exhibition method of the matching cypher text.
Optionally, the acquisition matches cypher text, comprising:
By each text unit in the first text section in the current cypher text, with the machine translation result Carry out Semantic Similarity Measurement；Wherein, first text section includes the translator in the machine translation result Untranslated full text unit or first text untranslated including the translator in the machine translation result N number of text unit that unit starts；
Using the corresponding text unit of maximum semantic similarity being calculated as matching cypher text.
Optionally, the method also includes:
After getting the current cypher text, next voice unit that the translator is translated is predicted；
The cypher text of next voice unit is obtained from the machine translation result, and/or, it is turned over from the machine Translate the cypher text obtained in result with next voice unit in semantically similar voice unit；
The cypher text that will acquire carries out special displaying.
Optionally, the method also includes:
In the machine translation result, by the matching cypher text, as text to be replaced；
If the text to be replaced is from the current cypher text in the semantically different or described text to be replaced and institute State current cypher text it is semantically identical and in expression way it is different, then in the machine translation result, using described Current cypher text replaces the text to be replaced.
Optionally, the machine translation is the result is that obtained using the translation model constructed in advance, then the method is also wrapped It includes:
After the translator completes the translation to the target voice, using the translator to the target language The human translation of sound makes the translation model have the translator's as a result, to translation model progress parameter update Translate style.
The embodiment of the present application also provides a kind of textual presentation methods, comprising:
Show the machine recognition result of target voice；
Translator is obtained in real time to the translation result of the voice unit in the target voice, as current translation text This；
Obtain match cognization text, the match cognization text be from the machine recognition result matching obtain with institute Current cypher text is stated in semantically immediate identification text；
Change the exhibition method of the match cognization text.
Optionally, the acquisition match cognization text, comprising:
The current cypher text is converted according to the languages type of the target voice, obtains currently identifying text This；
By each text unit in the current identification text, with the second text section in the machine recognition result Carry out Semantic Similarity Measurement；Wherein, second text section includes the translator in the machine recognition result Untranslated full text unit or first text untranslated including the translator in the machine recognition result N number of text unit that unit starts；
Using the corresponding text unit of maximum semantic similarity being calculated as match cognization text.
Optionally, the method also includes:
After getting the current cypher text, next voice unit that the translator is translated is predicted；
The identification text of next voice unit is obtained from the machine recognition result, and/or, know from the machine The identification text with next voice unit in semantically similar voice unit is obtained in other result；
The identification text that will acquire carries out special displaying.
The embodiment of the present application also provides a kind of textual presentation methods, comprising:
Show the machine processing of target voice as a result, the machine processing result includes the machine translation of the target voice As a result with machine recognition result；
Translator is obtained in real time to the translation result of the voice unit in the target voice, as current translation text This；
Obtain matching treatment text, the matching treatment text be from the machine processing result matching obtain with institute Current cypher text is stated in semantically immediate cypher text and the identification text of the cypher text；
Change the exhibition method of the matching treatment text.
The embodiment of the present application also provides a kind of textual presentation devices, comprising:
Machine translation result display unit, for showing the machine translation result of target voice；
Current cypher text acquiring unit, for obtaining translator in real time to the voice unit in the target voice Translation result, as current cypher text；
Cypher text acquiring unit is matched, for obtaining matching cypher text, the matching cypher text is from the machine In device translation result matching obtain with the current cypher text in semantically immediate cypher text；
First exhibition method changing unit, for changing the exhibition method of the matching cypher text.
Present invention also provides a kind of textual presentation devices, comprising:
Machine recognition result display unit, for showing the machine recognition result of target voice；
Current cypher text acquiring unit, for obtaining translator in real time to the voice unit in the target voice Translation result, as current cypher text；
Match cognization text acquiring unit, for obtaining match cognization text, the match cognization text is from the machine In device recognition result matching obtain with the current cypher text in semantically immediate identification text；
Second exhibition method changing unit, for changing the exhibition method of the match cognization text.
The embodiment of the present application also provides a kind of textual presentation devices, comprising:
Machine processing result display unit, for showing the machine processing of target voice as a result, the machine processing result Machine translation result and machine recognition result including the target voice；
Current cypher text acquiring unit, for obtaining translator in real time to the voice unit in the target voice Translation result, as current cypher text；
Matching treatment text acquiring unit, for obtaining matching treatment text, the matching treatment text is from the machine What matching obtained in device processing result is literary in semantically immediate cypher text and the translation with the current cypher text This identification text；
Third exhibition method changing unit, for changing the exhibition method of the matching treatment text.
Based on the above-mentioned technical proposal, the application has the advantages that
In textual presentation method and apparatus provided by the present application, after showing the machine translation result of target voice, Translator is first obtained in real time to the translation result of the voice unit in the target voice, as current cypher text, then from this Matching is obtained with the current cypher text in semantically immediate cypher text, as matching translation text in machine translation result This, and the exhibition method of the matching cypher text is changed, so that the translated content of translator and untranslated content carry out Difference is shown.Wherein, translator in real time translates each voice unit in target voice, so that in machine translation It as a result, can be according to the translation progress of translator to machine translation result corresponding to each voice unit during displaying It is shown the change of mode one by one, in this way, translator can be according to the machine translation of current presentation as a result, accurately knowing The translation progress of its own, so that translator can be absorbed in untranslated content, after being so conducive to translator The continuous translation to untranslated content.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is the application Some embodiments for those of ordinary skill in the art without creative efforts, can also basis These attached drawings obtain other attached drawings.
Fig. 1 is the flow chart for the textual presentation method that the application embodiment of the method one provides；
Fig. 2 is the flow chart for the textual presentation method that the application embodiment of the method three provides；
Fig. 3 is the flow chart for the textual presentation method that the application embodiment of the method four provides；
Fig. 4 is the structural schematic diagram for the textual presentation device that the application Installation practice one provides；
Fig. 5 is the structural schematic diagram for the textual presentation device that the application Installation practice two provides；
Fig. 6 is the structural schematic diagram for the textual presentation device that the application Installation practice three provides.
Specific embodiment
To keep the purposes, technical schemes and advantages of the embodiment of the present application clearer, below in conjunction with the embodiment of the present application In attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is Some embodiments of the present application, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art Every other embodiment obtained without making creative work, shall fall in the protection scope of this application.
Embodiment of the method one
Referring to Fig. 1, which is the flow chart for the textual presentation method that the application embodiment of the method one provides.
Textual presentation method provided by the embodiments of the present application, the methods of exhibiting of the translation result of voice mainly to be translated, The method comprising the steps of S101-S104:
S101: the machine translation result of target voice is shown.
The embodiment of the present application does not limit the source of target voice, for example, target voice can be in meeting alternative interpretation field Received voice to be translated under scape is also possible to received voice to be translated under the simultaneous interpretation scene in convention, also Or received voice to be translated under other translation scenes.For example, target voice is that every wheel is handed under meeting alternative interpretation scene For the voice to be translated in translation.
In addition, the present embodiment does not limit the voice translation method of target voice yet, it can be using existing or future appearance Any one voice translation method translates target voice, to obtain the machine translation result of target voice.
It should be noted that the target voice in the present embodiment can be translated by translator, wherein based on translation The difference of scene, the translator can be the interpreter in meeting alternative interpretation scene, be also possible in simultaneous interpretation scene Interpreter can also be the interpreter in other scenes.
In addition, the embodiment of the present application does not limit the methods of exhibiting of machine translation result yet, can using it is existing or future go out Existing any methods of exhibiting is shown machine translation result.
S102: translator is obtained in real time to the translation result of the voice unit in target voice, as current translation text This.
Voice unit can be the dividing unit referred in target voice.For the ease of explanation and understanding voice unit, below It will be illustrated in conjunction with two examples.
As the first example, each voice unit in target voice can be each of target voice sentence, this When, S102 is specifically as follows: obtaining translator in real time to the translation result of each sentence in target voice, will currently obtain Sentence as current cypher text.
As the second example, each voice unit can refer to each combination of sentences, and each combination of sentences can be with Including at least two sentences (for example, each combination of sentences may include 3 sentences), at this point, S102 is specifically as follows: in real time Translator is obtained to the translation result of each combination of sentences in target voice, using each combination of sentences currently obtained as Current cypher text.
It should be noted that above-mentioned two kinds of embodiments that voice unit is described with two examples, however, in the application Voice unit can also be not specifically limited in this embodiment using other embodiments, the application.
Based on the related content of upper speech units, step S102 is explained and illustrated below in conjunction with example.
As an example, when target voice includes 10 voice units, and translator is to the 2nd in target voice When voice unit is translated, then S102 is specifically as follows: obtaining translator in real time to the 2nd voice list in target voice The translation result of member, and using the translation result of the 2nd voice unit in target voice as current cypher text, to be based on The current cypher text executes subsequent step.
S103: obtain matching cypher text, the matching cypher text be from machine translation result matching obtain with work as Preceding cypher text is in semantically immediate cypher text.
The embodiment of the present application does not limit used text matching technique when acquisition matching cypher text, can use any Kind of text matching technique is matched to current cypher text in semantically immediate cypher text from machine translation result, from And obtain the corresponding matching cypher text of current cypher text.
It is understood that the corresponding expression way of different language has difference, so that when stating identical content, it is different The corresponding statement word order of languages also can be different.However, some machine translation softwares are machine when translating to target voice It is translated according to the statement word order of target voice to tool formula, can't consider the otherness of the statement word order of different language, such as This result in serve as interpreter personnel according to be different from the target voice statement word order target voice is translated when, translator The statement word order of given cypher text will differ from the statement word order of machine translation result, thus, in order to accurately It determines the content in machine translation result corresponding to current cypher text given by translator, will can currently translate The text all or part of the content untranslated with the translator in machine translation result matches, and matching degree is highest Content so can be improved the matching accuracy of matching cypher text as matching cypher text.Based on above-mentioned thinking, the application Embodiment additionally provides a kind of embodiment of S103, which can specifically include step S1031-S1032:
S1031: by each text unit in the first text section in current cypher text, with machine translation result into Row Semantic Similarity Measurement.
Wherein, the first text section refers to when matching to current cypher text, can be into machine translation result It composes a piece of writing this matched section；Moreover, the interval range in the first text section can be preset.For the ease of explanation and understanding One text section is illustrated below in conjunction with two kinds of embodiments.
As the first embodiment, the first text section may include that the translator in machine translation result is untranslated Full text unit.
Wherein, the text unit in machine translation result is corresponding with the voice unit in target voice, specifically: in machine In device translation result, each text unit refers to translation result corresponding to each voice unit.For example, in machine translation result The 1st text unit can be used to indicate that the translation result of the 1st voice unit in target voice.
For the ease of the first embodiment in the first text of explanation and understanding section, said below in conjunction with example It is bright.
It is assumed that target voice includes the 1st voice unit to the 10th voice unit；Machine translation result includes the 1st Text unit is to the 10th text unit；1st text unit to the 10th text unit is the 1st voice unit respectively to The translation result of 10 voice units.
As an example, based on above-mentioned it is assumed that working as translation result conduct of the translator that will acquire to the 2nd voice unit After current cypher text；If translator first completes before translating the 2nd voice unit and turns over to the 1st voice unit It translates, at this point, the 2nd voice unit to the 10th voice unit is that translator does not complete translation, then the first text section can To include the 2nd text unit to the 10th text unit；If translator does not complete before translating the 2nd voice unit Translation (that is, first having translated the 2nd voice unit) to the 1st voice unit, at this point, the 1st voice unit, the 2nd voice Unit to the 10th voice unit is that translator does not complete translation, then the first text section may include the 1st text list Member, the 2nd text unit to the 10th text unit.
The above are the first embodiments in the first text section.
Change in addition, though translator may will do it statement word order during translating to target voice Become, still, the statement changed cypher text of word order still can be located at around original table predicate sequence, in this way, in order to improve The matching efficiency of cypher text is matched, the translator that the first text section can be made to only include in machine translation result is untranslated Partial content, thus, present invention also provides second of embodiments in the first text section, in this embodiment, One text section may include N number of text that first untranslated text unit of the translator in machine translation result starts Unit.
Wherein, N is preset, it is particularly possible to be set according to the demand of user itself, N may be greater than or wait In 1 any integer.
For the ease of second of embodiment in the first text of explanation and understanding section, said below in conjunction with example It is bright.
It is assumed that target voice includes the 1st voice unit to the 10th voice unit；Machine translation result includes the 1st Text unit is to the 10th text unit；1st text unit to the 10th text unit is the 1st voice unit respectively to The translation result of 10 voice units.
As an example, based on it is above-mentioned it is assumed that when preset N be 3 when, then in the translator that will acquire to the 2nd language After the translation result of sound unit is as current cypher text；If translator has completed before translating the 2nd voice unit Translation to the 1st voice unit, at this point, the 2nd voice unit to the 10th voice unit is that translator does not complete and turns over It translates, then the first text section includes the 2nd text unit, the 3rd text unit and the 4th text unit, wherein the 2nd Text unit is first untranslated text unit of the translator in machine translation result, and the 3rd text unit is located at the 2nd After a text unit and adjacent with the 2nd text unit, the 4th text unit is located at after the 3rd text unit and with the 3 A text unit is adjacent；If translator does not complete the translation to the 1st voice unit before translating the 2nd voice unit (that is, first having translated the 2nd voice unit), at this point, the 1st voice unit, the 2nd voice unit to the 10th voice unit are Translator does not complete translation, then the first text section may include the 1st text unit, the 2nd text unit and the 3rd Text unit；Wherein, the 1st text unit is first untranslated text unit of translator in machine translation result, 2nd text unit is after the 1st text unit and adjacent with the 1st text unit, and the 3rd text unit is located at the 2nd It is after a text unit and adjacent with the 2nd text unit.
The above are second of embodiments in the first text section.
The above are the related contents in the first text section.
Semantic similarity is used to indicate similarity degree between two texts, specifically: currently cypher text is turned over machine The semantic similarity between each text unit in the first text section in result is translated, is respectively used to indicate currently to translate text Originally the similarity degree between each text unit in the first text section in machine translation result.In addition, two texts Between semantic similarity it is higher, then it represents that the similarity degree between two texts is bigger.
In addition, the embodiment of the present application does not limit the semantic similarity calculation method between two texts, can use existing Or the following any semantic similarity calculation method occurred, to the first text in current cypher text and machine translation result Each text unit in section carries out Semantic Similarity Measurement, alternatively, in current cypher text and machine translation result The corresponding combine text of each text unit in first text section carries out Semantic Similarity Measurement, to respectively obtain first The semantic similarity of each text unit and current cypher text in text section；Wherein, said combination text may include The combine text that the adjacent text unit of corresponding text unit and the correspondence text unit is formed, and adjacent text unit can be with It is adjacent including at least one text unit adjacent on the left of corresponding with this text unit and/or on the right side of the corresponding text unit At least one text unit.
For the ease of explanation and understanding semantic similarity calculation method, will be with a kind of semantic similarity calculation method below Example is illustrated.
As an implementation, semantic similarity calculation method can use cosine similarity and be calculated, moreover, remaining The calculating of string similarity can use formula (1) realization, and current cypher text is defined as by formula (1) here for ease of description Each text unit in the first text section in machine translation result is defined as the second text by the first text, in this way, can With calculated using formula (1) each text unit in the first text section in current cypher text and machine translation result it Between semantic similarity.
In formula (1), cosine similarity of the cos (θ) between the first text and the second text is (it is, semantic similar Degree)；A is the text vector of the first text；B is the text vector of the second text；| | A | | it is the text vector of the first text Mould；| | B | | for the mould of the text vector of the second text；AiFor i-th of element of the text vector of the first text；BiFor the second text I-th of element of this text vector；N is included in the text vector (or, text vector of the second text) of the first text Element number.
Alternatively, cosine similarity of the cos (θ) between the first text and the second text is (it is, semantic phase in formula (1) Like degree)；A is the text vector of the first text；B is the text vector of the corresponding combine text of the second text；| | A | | for the first text The mould of this text vector；| | B | | for the mould of the text vector of the corresponding combine text of the second text；AiFor the text of the first text I-th of element of this vector；BiFor i-th of element of the text vector of the corresponding combine text of the second text；N is the first text Text vector (or, text vector of the corresponding combine text of the second text) in included element number.
It should be noted that the cosine that the value of cosine similarity is respectively less than between the 1, and first text and the second text is similar Degree is closer to 1, then it represents that more similar between the first text and the second text.
The calculation method of cosine similarity based on above-mentioned introduction below will for the ease of explanation and understanding step S1031 It is introduced in conjunction with example.
As an example, when including the 2nd text unit, the 3rd text in the first text section in machine translation result When unit and the 4th text unit, then S1031 is specifically as follows: utilizing formula (1), calculates separately current cypher text and the 2nd The semantic similarity cos (θ) of a text unit2, current cypher text and the 3rd text unit semantic similarity cos (θ)3、 And the semantic similarity cos (θ) of current cypher text and the 4th text unit4。
It should be noted that above-mentioned example is for including the first text section of 4 text units to step S1031 It is illustrated.However, in this application, the first text section in machine translation result may include at least one text list Member, moreover, can be using semanteme phase provided by above-mentioned example when the first text section includes at least one text unit Like the calculation method of degree, for the sake of brevity, details are not described herein.
The above are the specific embodiments of S1031.
S1032: using the corresponding text unit of maximum semantic similarity being calculated as matching cypher text.
The embodiment of the present application does not limit the screening technique of maximum semantic similarity.
S1032 can use numerous embodiments, for the ease of explanation and understanding, below will in one embodiment for It is illustrated.
As an implementation, S1032 is specifically as follows: each semantic similarity that step S1031 is obtained carries out Sequence, and maximum semantic similarity is selected from the semantic similarity after sequence, so as to by the corresponding text of maximum semantic similarity This unit (or the corresponding combine text of text unit) is as matching cypher text.
It should be noted that in the above-described embodiment, it can be by each semantic similarity according to sequence from small to large It is ranked up, at this point, the maximum semantic similarity of similarity conduct that selected and sorted is last；It can also be similar by each semanteme Degree is ranked up according to sequence from big to small, at this point, the maximum semantic similarity of similarity conduct that selected and sorted is most preceding.
The above are the specific embodiments of S103.
S104: the exhibition method of change matching cypher text.
In the embodiment of the present application, matching cypher text can use numerous variations exhibition method, for example, can change Displaying color with cypher text can also change the displaying font style of matching cypher text (for example, font type, font One of background color, font size etc. are a variety of), or matching cypher text increases label symbol, can also be right Cypher text is matched using other change exhibition methods, the application is not specifically limited in this embodiment.
For the ease of explanation and understanding S104, will be illustrated by taking the three of S104 kinds of embodiments as an example below.
As the first embodiment, S104 is specifically as follows: the displaying color for matching cypher text is changed to preset Color.Wherein, pre-set color can be preset, it is particularly possible to be set according to the demand of user, for example, pre-set color can Think light gray.
As second of embodiment, S104 is specifically as follows: the displaying font style for matching cypher text is changed to Default font style.Wherein, default font style can be preset, it is particularly possible to be set according to the demand of user, example Such as, default font style can be with are as follows: font type is regular script, and font size is No. four, and font color is light green.
As the third embodiment, S104 is specifically as follows: label is added on the predeterminated position of matching cypher text Symbol.Wherein, predeterminated position can be preset, it is particularly possible to be set according to the demand of user, for example, predeterminated position can To be the top (or lower section or front) for matching the initial position of cypher text, it is also possible to match the interposition of cypher text The top (or lower section or front) set can also be that the other positions of matching cypher text, the application do not do specific limit to this It is fixed.
In addition, label symbol can be any symbol, the application is not specifically limited in this embodiment；Moreover, label symbol can To preset, it is particularly possible to be set according to the demand of user.For example, label symbol can be " → ", be also possible to " # " is also possible to " & ", is also possible to " * " etc..
The above are the textual presentation methods for being used to show cypher text that embodiment of the method one provides, and are showing target language After the machine translation result of sound, translator is first obtained in real time to the translation result of the voice unit in the target voice, is made For current cypher text, then from the machine translation result matching obtains turning over the current cypher text semantically immediate Translation sheet as matching cypher text, and changes the exhibition method of the matching cypher text, so that in translator is translated Hold and carries out difference displaying with untranslated content.Wherein, translator in real time to each voice unit in target voice into Row translation, so that during the displaying of machine translation result, it can be according to the translation progress of translator to each voice list Machine translation result corresponding to member is shown the change of mode one by one, in this way, translator can be according to current presentation Machine translation as a result, accurately know the translation progress of its own so that translator can be absorbed in it is untranslated interior Hold, is so conducive to the subsequent translation to untranslated content of translator.
Embodiment of the method two
It should be noted that when the personnel that serve as interpreter translate target voice, it may be according to different from target voice Used statement word order carries out the statement of cypher text, and still, machine translation result but may be to be made according to target voice Statement word order is shown, and so causes translator that can not rapidly find next voice from machine translation result Cypher text corresponding to unit.Thus, it is wanted in order to enable translator rapidly can find it from machine translation result Cypher text corresponding to next voice unit of translation, can be with the lower voice unit of look-ahead, and shows to translator Cypher text corresponding to next voice unit.
According to the demand, textual presentation method (step that the embodiment of the present application is provided based on above method embodiment one S101-S104), it can further include step S105-S107:
S105: after getting current cypher text, next voice unit that translator is translated is predicted.
Next voice unit refers to after translator provides current cypher text next translator will turn over The voice unit translated.Specifically, it can be matched in the content of text and machine translation result in conjunction with current cypher text The content of text of cypher text and its surrounding text predicts next voice unit.
It is understood that in target voice, voice unit corresponding to current cypher text and next voice unit Between the sequence of positions do not fixed；And the reason of leading to this phenomenon, is, translator translates to target voice When, the translation sequence between different phonetic unit is changed, so that the human translation text of target voice more meets translation text The statement of this correspondence languages is accustomed to.
S106: according to machine translation as a result, determining the corresponding cypher text of next voice unit.
In S106, the corresponding cypher text of next voice unit can be determined using various ways.It below will be with It is illustrated for three kinds of embodiments of S106.
As the first embodiment of S106, S106 is specifically as follows: next voice is obtained from machine translation result The cypher text of unit.
Due to machine translation the result is that being translated to target voice, thus, include in machine translation result The machine translation result of each voice unit (namely each text unit), moreover, the machine translation of each voice unit As a result with each voice unit have one-to-one relationship, thus, when determined in step s105 next voice unit it It afterwards, can be according to the corresponding relationship between the machine translation result and each voice unit of each voice unit, from machine translation As a result the cypher text of next voice unit is obtained in.
In order to facilitate understanding with the first embodiment of explanation S106, it is illustrated below in conjunction with example.
Assuming that target voice includes the 1st voice unit to the 3rd voice unit；Machine translation result includes the 1st text This unit is to the 3rd text unit；1st text unit is the machine translation of the 1st voice unit as a result, the 2nd text list Member is the machine translation of the 2nd voice unit as a result, and the 3rd text unit is the machine translation result of the 3rd voice unit.
As an example, based on above-mentioned it is assumed that ought predict next voice unit in step s105 is the 3rd voice unit When, S106 is specifically as follows: obtaining the 3rd text unit in machine translation result, and by the 3rd in the machine translation result Cypher text of a text unit as next voice unit.
It should be noted that above-mentioned example is to include the target voice of 3 voice units, and including 3 text units Machine translation result for the first embodiment of S106 is illustrated.However, in this application, target voice can To include at least one voice unit, and machine translation result also may include at least one text unit, moreover, working as target language Sound includes at least one voice unit, and when machine translation result includes at least one text unit, can be shown using above-mentioned The method that example provides obtains the cypher text of next voice unit, and for the sake of brevity, details are not described herein.
The above are the first embodiments of S106.
In addition, since the cypher text corresponding to semantically similar two voice units should also be similar, because And the embodiment of the present application also provides second of embodiments of S106, in this embodiment, S106 is specifically as follows: from The cypher text with next voice unit in semantically similar voice unit is obtained in machine translation result, and the translation is literary This is as the corresponding cypher text of next voice unit.
For the ease of second of embodiment of explanation and understanding S106, it is specifically described below in conjunction with example.
Assuming that target voice includes the 1st voice unit to the 3rd voice unit；Machine translation result includes the 1st text This unit is to the 3rd text unit；1st text unit is the machine translation of the 1st voice unit as a result, the 2nd text list Member is the machine translation of the 2nd voice unit as a result, and the 3rd text unit is the machine translation result of the 3rd voice unit； 3rd voice unit and the 1st voice unit are in semantically similar two voice units.
As an example, based on above-mentioned it is assumed that ought predict next voice unit in step s105 is the 3rd voice unit When, S106 can specifically include S106a-S106c:
S106a: determining and the 3rd voice unit is in semantically similar voice unit, that is, the 1st voice unit.
S106b: the cypher text of the 1st voice unit is obtained from machine translation result.
It should be noted that S106b can use any implementation corresponding to the first embodiment of above-mentioned S106 " next voice unit " in the first embodiment of above-mentioned S106 need to only be replaced with the " the 1st in step S106b by mode Voice unit ".
S106c: the cypher text by the cypher text of the 1st voice unit, as the 3rd voice unit.
It should be noted that above-mentioned example is to include the target voice of 3 voice units, and including 3 text units Machine translation result for second of embodiment of S106 be illustrated.However, in this application, target voice can To include at least one voice unit, and machine translation result also may include at least one text unit, moreover, working as target language Sound includes at least one voice unit, and when machine translation result includes at least one text unit, can be shown using above-mentioned The method that example provides obtains the cypher text of next voice unit, and for the sake of brevity, details are not described herein.
The above are second of embodiments of S106.
In addition, can be combined above-mentioned to further increase the accuracy of cypher text corresponding to next voice unit Two kinds of embodiments of S106 carry out the acquisition of cypher text, thus, the embodiment of the present application also provides the third realities of S106 Mode is applied, which can specifically include step S1061-S1063:
S1061: obtaining the cypher text of next voice unit from machine translation result, as the first cypher text.
It should be noted that " obtaining the translation text of next voice unit from machine translation result in step S1061 This " any embodiment corresponding to the first embodiment of above-mentioned S106 can be used.
S1062: the translation with next voice unit in semantically similar voice unit is obtained from machine translation result Text, as the second cypher text.
It should be noted that " obtaining with next voice unit from machine translation result semantically in step S1062 The cypher text of similar voice unit " can use any embodiment party corresponding to second of embodiment of above-mentioned S106 Formula is realized.
S1063: the first cypher text and the second cypher text are regard as the corresponding cypher text of next voice unit.
It should be noted that not having fixed to execute sequence between step S1061 and step S1062.For example, can be successively Step S1061 and step S1062 is executed, step S1062 and step S1061 also can be executed sequentially, may also be performed simultaneously step Rapid S1061 and step S1062, the application are not specifically limited in this embodiment.
The above are the third embodiments of S106.
S107: the cypher text that will acquire carries out special displaying.
The corresponding cypher text of next voice unit can be rapidly known for the ease of translator, it can be by step The cypher text that S106 is obtained carries out special displaying, can highlight in machine translation result.
Wherein, special displaying refers to that exhibition method used in the corresponding cypher text of next voice unit is different from machine Exhibition method used in other texts in translation result, so that translator can quickly be concerned about next voice unit Corresponding cypher text.
In the present embodiment, the corresponding cypher text of next voice unit can be shown come special using multiple means.For It is convenient for explanation and understanding, is illustrated below with reference to two special displaying examples.
As the first example, carried out in the case where not changing the display position of the corresponding cypher text of next voice unit Special displaying, at this point, the cypher text can be using the special exhibition method of default or using the preset spy of translator Different exhibition method.For example, can use the mode being highlighted shows the corresponding cypher text of next voice unit to protrude；It can Project is arranged by least one of font color, font type, font background color, font size etc. font, to dash forward The corresponding cypher text of next voice unit is shown out.
As the second example, change the display position of the corresponding cypher text of next voice unit, so that next voice list The corresponding cypher text of member can be shown in predeterminable area.Wherein, predeterminable area is one different from machine translation result Display area another region, which can preset, for example, the predeterminable area can be preset window, this When, it can use the preset window and prominent displaying carried out to the corresponding cypher text of next voice unit, wherein preset window can To preset, it is particularly possible to be set according to user demand, for example, preset window can be a suspended frame.
For the ease of explanation and understanding S107, it is illustrated below in conjunction with three kinds of embodiments.
As the first embodiment, when obtaining cypher text using the first embodiment of S106, then S107 It is specifically as follows: the cypher text is highlighted.
As second of embodiment, when obtaining cypher text using second of embodiment of S106, then S107 It is specifically as follows: if the cypher text is highlighted by the cypher text in the current presentation page；And/or pass through Suspended frame shows the cypher text.
As the third embodiment, when obtaining cypher text using the third embodiment of S106, then S107 It is specifically as follows: the first cypher text is highlighted, and the second cypher text is shown by suspended frame.
It should be noted that above-mentioned three kinds of embodiments are to be with two kinds of embodiments being highlighted or suspended frame is shown What example was illustrated the display mode of the corresponding cypher text of next voice unit.However, in this application, to next voice The display mode of the corresponding cypher text of unit can also be not specifically limited herein using other display modes, the application.
It should be noted that the execution time of step S105-S107 is not fixed, and step S105-S107 " need to only obtained To current cypher text " it executes later.For example, step S105-S107 can be executed and be held after executing step S102 It is executed before row S103, can also execute, can also be held after executing S104 after executing S103 and before executing S104 Row.
The above are a kind of tools for the textual presentation method for showing cypher text that the application embodiment of the method two provides Body embodiment can predict next voice unit that translator is translated, and under Special display in this embodiment The corresponding cypher text of one voice unit.In this way, when the personnel that serve as interpreter complete the translation of current speech unit, under need to being only absorbed in The translation of one voice unit, without spending time and efforts to find next voice unit pair from machine translation result The machine translation answered is as a result, be so conducive to the subsequent translation to untranslated content of translator.
In addition, machine translation result is only through what mechanically special translating purpose voice obtained, it is so easy to cause machine The partial content semanteme statement of translation result is incorrect, or the partial content of machine translation result is caused not meet actual translations When language expression scene or statement habit, thus, in order to optimize machine translation as a result, machine translation result can had found Partial content there are the artificial cypher texts in part that when above problem, personnel are provided with the help of an interpreter to replace machine translation result Partial content, so ensure that can be with reference to the machine translation knot for correcting for the above problem during translator's subsequent translation Fruit is conducive to the subsequent translation to untranslated content of translator.
According to the demand, textual presentation method (step that the embodiment of the present application is provided based on above method embodiment one S101-S104) or on the basis of the step S105-S107 of the offer of this method embodiment two, it can further include step S108-S109:
S108: in machine translation result, cypher text will be matched, as text to be replaced.
S109: it if it is determined that there are flaws for text to be replaced, then in machine translation result, is replaced using current cypher text Text to be replaced.
Text to be replaced can refer in text to be replaced there are flaw effectively to be translated in the presence of can influence translator The problem of；Moreover, flaw present in text to be replaced can be it is any can influence the problem of translator translates, The application is not specifically limited in this embodiment.For example, when text to be replaced from current cypher text when semantically different, then can be true There are flaws for fixed text to be replaced；When text to be replaced and current cypher text it is semantically identical and in expression way not Together, then it can determine that there are flaws for text to be replaced.
For the ease of explanation and understanding S109, it is illustrated below in conjunction with two kinds of embodiments of S109.
As the first embodiment, S109 is specifically as follows: if text to be replaced and current cypher text are semantically Difference replaces text to be replaced using current cypher text then in machine translation result.
Due to target voice carry out machine translation during, may be because many factors be (for example, meaning transference is wrong Accidentally, the factors such as the correct meaning of a word can not be obtained) cause machine translator that can not correctly translate certain voice units, so that these voices The translation result of unit goes wrong semantically, so as to cause the translation result of these voice units and turning over for translator Content is translated semantically different, at this point, in order to optimize machine translation as a result, can with the translation progress of translator, In machine translation result, the translation result of these voice units is substituted one by one using the translation content of translator.
For the ease of the first embodiment of explanation and understanding S109, it is illustrated below in conjunction with example.
It is assumed that after translator gives the translation result of the 2nd voice unit of target voice, by the 2nd language The translation result of sound unit has determined that the 2nd text unit in machine translation result is current as current cypher text The matching cypher text of cypher text；And current cypher text and matching cypher text are different semantically.
As an example, based on it is above-mentioned it is assumed that when in step S108 using the 2nd text unit as when text to be replaced, Then S109 is specifically as follows: determine the 2nd text unit from current cypher text after semantically different, in machine translation As a result in, the 2nd text unit is replaced using current cypher text.
It should be noted that above-mentioned example is by taking the replacement process of the 2nd text unit in machine translation result as an example The first embodiment of S109 is illustrated.However, machine translation result may include at least one in this application Text unit, moreover, there are the text units of matter of semantics can use above-mentioned example institute by any in machine translation result The replacement process of offer, for the sake of brevity, details are not described herein.
It, can be according to turning over using replacement process provided by the embodiment the above are the first embodiment of S109 The translation progress for translating personnel turns over replacing with one by one in machine translation result there are at least one text unit of matter of semantics The translation result of voice unit corresponding to the text unit that the personnel of translating provide so, it is possible to overcome in machine translation result Existing matter of semantics, to be conducive to the subsequent translation to untranslated content of translator.
As second of embodiment, S109 is specifically as follows: if text to be replaced and current cypher text are semantically It is identical and different in expression way, then in machine translation result, text to be replaced is replaced using current cypher text.
It, may be directly according to the expression side of target voice since machine translator is during translating target voice Formula is translated, so that the expression way of the partial content of machine translation result may not met needed for actual translations Expression way so will affect the understanding of translator, thus, it can be fast during subsequent translation for the ease of translator Machine translation is understood fastly as a result, can replace one by one with the translation progress of translator using the translation content of translator It has changed and there are problems that the partial content of expression way in machine translation result.
For the ease of second of embodiment of explanation and understanding S109, it is illustrated below in conjunction with example.
It is assumed that after translator gives the translation result of the 2nd voice unit of target voice, by the 2nd language The translation result of sound unit has determined that the 2nd text unit in machine translation result is current as current cypher text The matching cypher text of cypher text；And current cypher text and matching cypher text are semantically identical and in expression way It is different.
As an example, based on it is above-mentioned it is assumed that when in step S108 using the 2nd text unit as when text to be replaced, Then S109 is specifically as follows: determining the 2nd text unit and current cypher text semantically identical and in expression way After difference, in machine translation result, the 2nd text unit is replaced using current cypher text.
It should be noted that above-mentioned example is by taking the replacement process of the 2nd text unit in machine translation result as an example Second of embodiment of S109 is illustrated.However, machine translation result may include at least one in this application Text unit, moreover, any text unit that there are problems that expression way in machine translation result can be shown using above-mentioned Replacement process provided by example, for the sake of brevity, details are not described herein.
It, can be according to turning over using replacement process provided by the embodiment the above are second of embodiment of S109 The translation progress for translating personnel will have that at least one text unit of expression way is replaced one by one in machine translation result The translation result of voice unit corresponding to the text unit provided for translator so, it is possible to overcome machine translation knot Expression way problem in the presence of fruit, to be conducive to the subsequent translation to untranslated content of translator.
It should be noted that the execution time of step S108-S109 is not fixed, and step S108-S109 " need to only obtained To matching cypher text " it executes later, it is, step S108-S109 only need to execution be after step S103 execution Can, the application is not specifically limited the execution time of step S108-S109.
In addition, for second of embodiment of above-mentioned S106, that is, obtained and next voice list from machine translation result Member is for the semantically process of the cypher text of similar voice unit, due to similar semantically with next voice unit The cypher text of voice unit has overcome the matter of semantics and expression way problem in the presence of machine translation result, so The determining corresponding cypher text of next voice unit can more meet the translation result of translator, so can be improved translation Translation accuracy and translation efficiency of the personnel to next voice unit.
Further, since different translators has different translation styles, thus, in order to enable machine translation result energy The translation style for enough meeting each translator, can use translator translated the cypher text of completion to target voice into Used translation model is trained when row machine translation, and the machine translation result which is provided more accords with The translation style for closing translator, to be conducive to the subsequent translation to untranslated content of translator.
According to the demand, textual presentation method (step that the embodiment of the present application is provided based on above method embodiment one S101-S104) or this method embodiment two provide step S105-S107 or this method embodiment two provide On the basis of S108-S109, step S110 can further include:
S110: after the personnel that serve as interpreter complete to the translation of target voice, translator's manually turning over to target voice is utilized It translates as a result, making translation model that there is the translation style of translator translation model progress parameter update.
Since the translation style of translator can be embodied by the completed translation content of translator, thus, it is turning over It translates after personnel's translation gives the human translation result of target voice in current translation cycle, can use the human translation knot Fruit is trained translation model, to update the parameter of the translation model, so that the updated translation model of parameter more meets The translation style of translator is turned over so in next translation cycle using the machine that the updated translation model of parameter obtains The translation style that result more meets translator is translated, so that translator carries out to the target voice in next translation cycle Content needed for more translate can be got when translation from the machine translation result, reduces the burden of translator, mentions High translation efficiency of the translator to the target voice in next translation cycle, so that it is subsequent to not to be conducive to translator The translation of the content of translation.
In addition, the application does not limit the embodiment of translation model, translation model can be appointing for existing or future appearance A kind of model that can be translated according to voice.
In addition, the parameter for not limiting translation model in this application updates interval, specifically: the parameter of translation model can To be updated after each translation cycle terminates, at least one translation cycle can also be spaced and once updated.
For the ease of explanation and understanding S110, will be illustrated by taking two kinds of embodiments as an example below.
Example one, S110 is specifically as follows: the personnel that serve as interpreter complete turning over to the target voice in the first translation cycle After translating, using translator to the human translation of the target voice in the first translation cycle as a result, joining to translation model Number updates, and is translated using the updated translation model of parameter to the target voice in the second translation cycle, to obtain The machine translation result of target voice in second translation cycle.
Example two, S110 is specifically as follows: the personnel that serve as interpreter have been continuously finished to the target voice in the first translation cycle After translation with the target voice in the second translation cycle, using translator to the target voice in the first translation cycle Human translation result and to the human translation of the target voice in the second translation cycle as a result, carrying out parameter more to translation model Newly, and using parameter updated translation model the target voice in third translation cycle is translated, to obtain third The machine translation result of target voice in translation cycle.
The above are another of the textual presentation method for showing cypher text of the offer of the application embodiment of the method two Specific embodiment can utilize translator after translator completes to the translation of target voice in this embodiment Human translation to the target voice makes the translation model have the interpreter as a result, to translation model progress parameter update The translation style of member.In this way, since the updated translation model of parameter more meets the translation style of translator, thus next In translation cycle, the translation wind of translator is more met using the machine translation result that the updated translation model of parameter obtains Lattice enable translator when translating to the target voice in next translation cycle from the machine translation result Content needed for getting more translate, reduces the burden of translator, improves translator to next translation cycle The translation efficiency of interior target voice, to be conducive to the subsequent translation to untranslated content of translator.
Embodiment of the method three
Based on (the i.e. translation text of any textual presentation method provided by above method embodiment one to embodiment of the method two This methods of exhibiting), required information when in order to more effectively show that it is translated to translator can also be by target language The identification text and cypher text of sound are shown simultaneously, so that translator can be quick to target voice according to displaying content It is translated on ground.Thus, this method embodiment three additionally provides a kind of textual presentation method, explains below in conjunction with attached drawing And explanation.
Referring to fig. 2, which is the flow chart for the textual presentation method that the application embodiment of the method three provides, including step S201-204:
S201: showing the machine processing of target voice as a result, the machine processing result includes the machine translation of target voice As a result with machine recognition result.
The embodiment of the present application does not limit the audio recognition method of target voice, can use any one audio recognition method Target voice is identified, to obtain the machine recognition result of target voice.
In the embodiment of the present application, " the machine translation result for showing target voice " is implemented with the above method in step S201 Step S101 in example one is identical, and related content refers to the related introduction of S101, and details are not described herein.
It should be noted that in step s 201, machine translation result and machine recognition result can use identical text This exhibition method, can also use different textual presentation modes, and the application is not specifically limited in this embodiment.In addition, the application The distribution relation between the display location of machine translation result and the display location of machine recognition result is not limited, for example, machine The display location of translation result and the display location of machine recognition result can be adjacent, can also be non-conterminous, moreover, when machine turns over When the display location of the display location and machine recognition result of translating result is adjacent, the display location of machine translation result can be located at The top of the display location of machine recognition result can also be located at the lower section of the display location of machine recognition result, for example, can be with Machine recognition result is subjected to corresponding displaying with its machine translation result as unit of sentence.
S202: translator is obtained in real time to the translation result of the voice unit in target voice, as current translation text This.
In the present embodiment, this step S202 is identical as the step S102 in above method embodiment one, and related content is asked Referring to the related introduction of S102, details are not described herein.
S203: obtain matching treatment text, the matching treatment text be from machine processing result matching obtain with work as Preceding cypher text is in semantically immediate cypher text and the identification text of the cypher text.
Since in machine processing result, the translation result of each voice unit and its recognition result have one-to-one correspondence Relationship, thus, in order to improve the matching efficiency of current cypher text Yu matching treatment text, can obtain matching treatment text During this, first carries out and the cypher text of each voice unit (or identification text) is matched in machine processing result Process, determine the voice list further according to the corresponding relationship of translation result and recognition result included in machine processing result The identification text (or cypher text) of member.Thus, present invention also provides two kinds of embodiments of S203, below will be successively right Two kinds of embodiments of S203 are explained and illustrated.
For ease of description, the cypher text and identification text in matching treatment text are respectively defined as matching and are turned over here Translation sheet and match cognization text.
As the first embodiment, S203 can specifically include step S203a1-S203a4:
S203a1: by each text unit in the first text section in current cypher text, with machine translation result Carry out Semantic Similarity Measurement.
Wherein, the first text section include the untranslated full text unit of translator in machine translation result or The N number of text unit started including first untranslated text unit of the translator in machine translation result.
S203a2: the corresponding text unit of maximum semantic similarity that will be calculated, as matching cypher text.
It should be noted that the step of S203a1-S203a2 and phase the step of S1031-S1032 in embodiment of the method one Together, the introduction of S1031-S1032 is referred to, details are not described herein.
S203a3: in machine recognition result, will match the corresponding recognition result of cypher text, as match cognization text This.
Since in machine processing result, the translation result and recognition result of each voice unit have one-to-one close System, thus, it, can be according to the translation knot of each voice unit after the corresponding matching cypher text of current cypher text has been determined Corresponding relationship between fruit and recognition result determines identification knot corresponding to the matching cypher text in machine recognition result Fruit, and using the recognition result as match cognization text.
S203a4: by matching cypher text and match cognization text, as matching treatment text.
As second of embodiment of S203, S203 can specifically include step S203b1-S203b5:
S203b1: current cypher text is converted according to the languages type of target voice, obtains currently identifying text.
The present embodiment does not limit the languages conversion method of current cypher text, can use any languages conversion method will Current cypher text is converted according to the languages type of target voice, obtains currently identifying text.
S203b2: will currently identify text, with each text unit in the second text section in machine recognition result Carry out Semantic Similarity Measurement.
Wherein, the second text section include the untranslated full text unit of translator in machine recognition result or The N number of text unit started including first untranslated text unit of the translator in machine recognition result.
It should be noted that the second text section is similar with the first text section in embodiment of the method one, difference only exists Correspond to machine recognition in the second text section as a result, the first text section corresponds to machine translation as a result, the second text section Division mode can refer to the first text section related introduction, details are not described herein.
S203b3: the corresponding text unit of maximum semantic similarity that will be calculated, as match cognization text.
In the present embodiment, any for the step S1032 that step S203b3 can be provided using above method embodiment one " matching cypher text " in step S1032 need to only be replaced with " match cognization text " by embodiment.
S203b4: in machine translation result, by the corresponding translation result of match cognization text, as matching translation text This.
Since in machine processing result, the recognition result and translation result of each voice unit have one-to-one close System, thus, it, can be according to the identification knot of each voice unit after the corresponding match cognization text of current identification text has been determined Corresponding relationship between fruit and translation result determines translation knot corresponding to the match cognization text in machine translation result Fruit, and using the translation result as matching cypher text.
S203b5: by matching cypher text and match cognization text, as matching treatment text.
S204: the exhibition method of change matching treatment text.
In the present embodiment, the execution movement of " exhibition method of change matching cypher text " can use in step S204 Any embodiment of step S104 in above method embodiment one, related content refer to the related of above-mentioned steps S104 and are situated between It continues, details are not described herein.
It should be noted that the matching cypher text and match cognization text that include in matching treatment text, the change of the two More exhibition method can it is identical, can also be different, the application is not specifically limited in this embodiment.
The above are the textual presentation methods that the application embodiment of the method three provides, and translator is in real time in target voice Each voice unit translated, can be according to the translation of translator so that during the displaying of machine translation result Progress is shown mode to machine translation result corresponding to each voice unit and machine recognition result one by one respectively Change, in this way, translator can be according to the machine translation result and machine recognition of current presentation as a result, accurately knowing it It is subsequent to not to be conducive to translator so that translator can be absorbed in untranslated content for the translation progress of itself The translation of the content of translation.
Further, in needed for being quickly concerned about in machine processing result in order to preferably supplementary translation personnel Hold, can also include step S205-S207 other than including above-mentioned steps S201-S204:
S205: after getting current cypher text, next voice unit that translator is translated is predicted.
This step S205 is identical as the S105 that above method embodiment two provides, and related content refers to the phase of step S105 It closes and introduces, details are not described herein.
S206: according to machine processing as a result, determining the corresponding identification text of next voice unit and/or cypher text.
Numerous embodiments can be used in S206, will be illustrated by taking three kinds of embodiments as an example below.
As the first embodiment, S206 is specifically as follows: according to machine translation as a result, determining next voice unit pair The cypher text answered.That is, the cypher text of next voice unit is obtained from machine translation result, and/or, from machine translation knot The cypher text with next voice unit in semantically similar voice unit is obtained in fruit.It should be noted that this first Embodiment is identical as the S106 that above method embodiment two provides, and related content refers to the related introduction of step S106, This is repeated no more.
As second of embodiment, S206 is specifically as follows: according to machine recognition as a result, determining next voice unit pair The identification text answered.That is, the identification text of next voice unit is obtained from machine recognition result, and/or, from machine recognition knot The identification text with next voice unit in semantically similar voice unit is obtained in fruit.It should be noted that this second Any embodiment for the step S106 that embodiment can be provided using above method embodiment two, only need to will be in step S106 " machine translation result " replace with " machine recognition result " and " cypher text " replaced with to " identification text ".
As the third embodiment, S206 is specifically as follows: according to machine processing as a result, determining next voice unit pair The cypher text and identification text answered.That is, obtaining the cypher text and slave of next voice unit from machine translation result The identification text of next voice unit is obtained in device recognition result；Alternatively, being obtained and next voice list from machine translation result Member is obtained with next voice unit in the semantically cypher text of similar voice unit and from machine recognition result in language The identification text of similar voice unit in justice；Alternatively, obtained from machine translation result the cypher text of next voice unit, And the identification text of next voice unit is obtained from machine recognition result, and, from machine translation result obtain with it is next Voice unit obtains and next voice list in the semantically cypher text of similar voice unit and from machine recognition result Identification text of the member in semantically similar voice unit.
S207: the identification text and/or cypher text that will acquire carry out special displaying.
In S207, needs " cypher text that will acquire carries out special displaying " and/or " the identification text that will acquire carries out Special displaying ".About " cypher text that will acquire carries out special displaying ", with the S107 phase in above method embodiment two Together, for details, reference can be made to the related introduction of S107, details are not described herein；About " the identification text that will acquire carries out special displaying ", Can using with the similar exhibition method of the S107 in above method embodiment two, " cypher text " in S107 need to be replaced For " identification text ".
It should be understood that the execution time of step S205-S207 is not fixed, and step S205-S207 " need to only obtained To current cypher text " it executes later.For example, step S205-S207 can be executed and be held after executing step S202 It is executed before row S203, can also execute, can also be held after executing S204 after executing S203 and before executing S204 Row.
As it can be seen that next voice unit that the application can be translated by predicting translator, and special displaying is next The corresponding cypher text of voice unit and/or identification text.In this way, when the personnel that serve as interpreter complete the translation of current speech unit, The translation of next voice unit need to be only absorbed in, without spending time and efforts from machine translation result and/or machine The corresponding machine translation result of next voice unit and/or machine recognition are found in device recognition result as a result, so being conducive to turn over Translate the subsequent translation to untranslated content of personnel.
Further, in needed for being quickly concerned about in machine processing result in order to preferably supplementary translation personnel Hold, can also include step S208-S209 other than including above-mentioned steps S201-S204 or step S201-S207:
S208: in machine translation result, using matching cypher text as text to be replaced.
S209a: if text to be replaced is from current cypher text in text semantically different or to be replaced and current translation Text it is semantically identical and in expression way it is different, then in machine translation result, using the replacement of current cypher text to Replace text；
S209b: it if the corresponding identification text of text to be replaced is different semantically from current cypher text, will currently turn over Translation sheet is converted according to the languages type of target voice, obtains currently identifying text, and in machine recognition result, is utilized Current identification text replaces the corresponding identification text of text to be replaced.
It should be noted that S209a and/or S209b can be executed.Wherein, the implementation and the above method of S 209a The implementation of S109 in embodiment two is identical, and for details, reference can be made to the related introductions of S209, and details are not described herein.
It is understood that since translator is when translating target voice, it can be with Primary Reference machine translation As a result, can also with Primary Reference machine recognition as a result, can be combined with reference to machine translation result and machine recognition as a result, because And reference habit when can be translated according to translator, to the mistake in machine translation result or machine recognition result It is corrected accordingly, correction object can be arranged in translator in advance.Based on aforementioned need, this application provides S209 (S209a-S209b) three kinds of embodiments:
It is then only executed when personnel's Primary Reference machine translation result of serving as interpreter is translated as the first embodiment S209a.In a kind of application scenarios, if above-mentioned S207 only needs to be shown cypher text, can only it execute S209a。
It is then only executed when personnel's Primary Reference machine recognition result of serving as interpreter is translated as second of embodiment S209b.In a kind of application scenarios, if above-mentioned S207 only needs to be shown identification text, can only it execute S209b。
As the third embodiment, the personnel that serve as interpreter combine to be turned over reference to machine translation result and machine recognition result When translating, then S209a and S209b are executed.In a kind of application scenarios, if above-mentioned S207 is needed to identification text and cypher text It is shown, then executes S209a and S209b.
Further, in needed for being quickly concerned about in machine processing result in order to preferably supplementary translation personnel Hold, in addition to including above-mentioned steps S201-S204 or step S201-S207 or step S201-S209 or step S201-S204 Can also include step S210 other than S208-S209:
S210: after the personnel that serve as interpreter complete to the translation of target voice, translator's manually turning over to target voice is utilized It translates as a result, making translation model that there is the translation style of translator translation model progress parameter update.
It should be noted that S210 is identical as the S110 that above method embodiment two provides, related content refers to step The related introduction of S110, details are not described herein.
Embodiment of the method four
In addition, preferring only show and know to translator with reference to the needs of identification text to meet translator Other text, it should be noted that and upper for showing the textual presentation method for identifying text provided by this method example IV It states provided by embodiment of the method one to three for showing that the textual presentation method of cypher text is similar.
Referring to Fig. 3, which is the flow chart for the textual presentation method that the application embodiment of the method four provides.
Textual presentation method provided by the embodiments of the present application, the methods of exhibiting of the identification text of voice mainly to be translated, Including step S301-S304:
S301: the machine recognition result of target voice is shown.
In the embodiment of the present application, any for the step S101 that step S301 can be provided using above method embodiment one " machine translation result " in the specific embodiment of step S101 need to only be replaced with " machine recognition result " by embodiment.
S302: translator is obtained in real time to the translation result of the voice unit in target voice, as current translation text This.
In the present embodiment, this step S302 is identical as the step S102 in above method embodiment one, and related content is asked Referring to the related introduction of S102, details are not described herein.
S303: obtain match cognization text, match cognization text be from machine recognition result matching obtain with it is current Cypher text is in semantically immediate identification text.
In the present embodiment, step S303 can be using any implementation of the step S103 in above method embodiment one " matching cypher text " in the specific embodiment of step S103 need to only be replaced with " match cognization text ", by " machine by mode Device translation result " replaces with " machine recognition result ".
For the ease of explanation and understanding S303, below will in one embodiment for be illustrated.
As an implementation, S303 can specifically include S3031-S3033:
S3031: current cypher text is converted according to the languages type of target voice, obtains currently identifying text.
S3032: will currently identify text, with each text unit in the second text section in machine recognition result into Row Semantic Similarity Measurement.
Wherein, the second text section include the untranslated full text unit of translator in machine recognition result, Or the N number of text unit started including first untranslated text unit of the translator in machine recognition result.
S3033: using the corresponding text unit of maximum semantic similarity being calculated as match cognization text.
It should be noted that above-mentioned steps S3031-S3033 is identical as the S203b1-S203b3 in embodiment of the method three, Related content refers to the related introduction of S203b1-S203b3, and details are not described herein.
S304: the exhibition method of change match cognization text.
In the present embodiment, any implementation for the step S104 that step S304 can be provided using above method embodiment one " matching cypher text " in the specific embodiment of step S104 need to only be replaced with " match cognization text " by mode.
The above are the textual presentation methods for being used to show identification text that embodiment of the method four provides, and are showing target language After the machine recognition result of sound, translator is first obtained in real time to the translation result of the voice unit in the target voice, is made For current cypher text, then from the machine recognition result matching obtains and current identification text corresponding to the current cypher text This as match cognization text, and changes the exhibition method of the match cognization text, makes in semantically immediate identification text It obtains the translated content of translator and untranslated content carries out difference displaying.Wherein, translator is in real time to target language Each voice unit in sound is translated, so that during the displaying of machine recognition result, it can be according to translator's Translation progress is shown the change of mode to machine recognition result corresponding to each voice unit one by one, in this way, interpreter Member can be according to the machine recognition of current presentation as a result, the translation progress of its own is accurately known, so that translator Untranslated content can be absorbed in, be so conducive to the subsequent translation to untranslated content of translator.
Further, in needed for being quickly concerned about in machine processing result in order to preferably supplementary translation personnel Hold, can also include step S305-S307 other than including above-mentioned steps S301-S304:
S305: after getting current cypher text, next voice unit that translator is translated is predicted.
This step S305 is identical as the step S105 in above method embodiment two, and related content refers to the correlation of S105 It introduces, details are not described herein.
S306: according to machine recognition as a result, determining the corresponding identification text of next voice unit.
Specifically, S306 includes: the identification text that next voice unit is obtained from machine recognition result, and/or, slave The identification text with next voice unit in semantically similar voice unit is obtained in device recognition result.
In the present embodiment, any embodiment party of the step S106 that can be provided using above method embodiment two of S306 " machine translation result " in the specific embodiment of step S106 need to only be replaced with " machine recognition result ", " will translated by formula Text " replaces with " identification text ".
S307: the identification text that will acquire carries out special displaying.
In the present embodiment, any embodiment party for the step S107 that S307 can be provided using above method embodiment two " cypher text " in the specific embodiment of step S107 need to only be replaced with " identification text " by formula.
It should be noted that the execution time of step S305-S307 is not fixed, and step S305-S307 " need to only obtained To current cypher text " it executes later.For example, step S305-S307 can be executed and be held after executing step S302 It is executed before row S303, can also execute, can also be held after executing S304 after executing S303 and before executing S304 Row.
As it can be seen that next voice unit that the application can be translated by predicting translator, and special displaying is next The corresponding identification text of voice unit.In this way, need to only be absorbed in next when the personnel that serve as interpreter complete the translation of current speech unit The translation of voice unit, it is corresponding without spending time and efforts to find next voice unit from machine recognition result Machine recognition as a result, being so conducive to the subsequent translation to untranslated content of translator.
Installation practice one
A kind of textual presentation device will be introduced in the present embodiment, and related content refers to above method embodiment.Ginseng See Fig. 4, which is the structural schematic diagram for the textual presentation device that the application Installation practice one provides.
Textual presentation device 400 provided by the embodiments of the present application, comprising:
Machine translation result display unit 401, for showing the machine translation result of target voice；
Current cypher text acquiring unit 402, for obtaining translator in real time to the voice list in the target voice The translation result of member, as current cypher text；
Cypher text acquiring unit 403 is matched, for obtaining matching cypher text, the matching cypher text is from described In machine translation result matching obtain with the current cypher text in semantically immediate cypher text；
First exhibition method changing unit 404, for changing the exhibition method of the matching cypher text.
As an implementation, the matching cypher text acquiring unit 403, comprising:
First Semantic Similarity Measurement subelement, being used for will be in the current cypher text, with the machine translation result The first text section in each text unit carry out Semantic Similarity Measurement；Wherein, first text section includes institute State the untranslated full text unit of the translator in machine translation result or including in the machine translation result N number of text unit that first untranslated text unit of the translator starts；
Matching cypher text determines subelement, for making the corresponding text unit of the maximum semantic similarity being calculated To match cypher text.
As an implementation, described device 400 further include:
Next voice unit predicting unit, for after getting the current cypher text, predicting the translator The next voice unit translated；
Next cypher text determination unit, for obtaining turning over for next voice unit from the machine translation result Translation sheet, and/or, it obtains with next voice unit from the machine translation result in semantically similar voice unit Cypher text；
Next cypher text display unit, the cypher text for will acquire carry out special displaying.
As an implementation, described device 400 further include:
Text determination unit to be replaced, in the machine translation result, by the matching cypher text, as to Replace text；
Text replacement unit to be replaced, if for the text to be replaced and the current cypher text semantically not With or the text to be replaced and the current cypher text it is semantically identical and in expression way it is different, then in institute It states in machine translation result, replaces the text to be replaced using the current cypher text.
As an implementation, described device 400 further include:
Translation model updating unit, if for the machine translation the result is that being obtained using the translation model constructed in advance , then after the translator completes the translation to the target voice, using the translator to the target voice Human translation as a result, carry out parameter update to the translation model, make the translation model that there is the translator to turn over Translate style.
Installation practice two
A kind of textual presentation device will be introduced in the present embodiment, and related content refers to above method embodiment.Ginseng See Fig. 5, which is the structural schematic diagram for the textual presentation device that the application Installation practice two provides.
Textual presentation device 500 provided by the embodiments of the present application, comprising:
Machine recognition result display unit 501, for showing the machine recognition result of target voice；
Current cypher text acquiring unit 502, for obtaining translator in real time to the voice list in the target voice The translation result of member, as current cypher text；
Match cognization text acquiring unit 503, for obtaining match cognization text, the match cognization text is from described In machine recognition result matching obtain with the current cypher text in semantically immediate identification text；
Second exhibition method changing unit 504, for changing the exhibition method of the match cognization text.
As an implementation, the match cognization text acquiring unit 503, comprising:
Current identification text obtains subelement, for the languages class by the current cypher text according to the target voice Type is converted, and obtains currently identifying text；
Second Semantic Similarity Measurement subelement is used for by the current identification text, in the machine recognition result The second text section in each text unit carry out Semantic Similarity Measurement；Wherein, second text section includes institute State the untranslated full text unit of the translator in machine recognition result or including in the machine recognition result N number of text unit that first untranslated text unit of the translator starts；
Match cognization text determines subelement, for making the corresponding text unit of the maximum semantic similarity being calculated For match cognization text.
As an implementation, described device 500 further include:
Next voice unit predicting unit predicts that the translator carries out after getting the current cypher text Next voice unit of translation；
Next identification text determination unit obtains the identification text of next voice unit from the machine recognition result This, and/or, the knowledge with next voice unit in semantically similar voice unit is obtained from the machine recognition result Other text；
Next identification textual presentation unit, the identification text that will acquire carry out special displaying.
Installation practice three
A kind of textual presentation device will be introduced in the present embodiment, and related content refers to above method embodiment.Ginseng See Fig. 6, which is the structural schematic diagram for the textual presentation device that the application Installation practice three provides.
Textual presentation device 600 provided by the embodiments of the present application, comprising:
Machine processing result display unit 601, for showing the machine processing of target voice as a result, the machine processing knot Fruit includes the machine translation result and machine recognition result of the target voice；
Current cypher text acquiring unit 602, for obtaining translator in real time to the voice list in the target voice The translation result of member, as current cypher text；
Matching treatment text acquiring unit 603, for obtaining matching treatment text, the matching treatment text is from described In machine processing result matching obtain with the current cypher text in semantically immediate cypher text and the translation The identification text of text；
Third exhibition method changing unit 604, for changing the exhibition method of the matching treatment text.
As an implementation, in order to be conducive to the subsequent translation to untranslated content of translator, the matching Handle text acquiring unit 603, comprising:
Third Semantic Similarity Measurement subelement, being used for will be in the current cypher text, with the machine translation result The first text section in each text unit carry out Semantic Similarity Measurement；Wherein, first text section includes institute State the untranslated full text unit of the translator in machine translation result or including in the machine translation result N number of text unit that first untranslated text unit of the translator starts；
First matching treatment text determines subelement, for by the corresponding text list of the maximum semantic similarity being calculated Member and the corresponding identification text of text unit, as matching treatment text.
As an implementation, the matching treatment text acquiring unit 603, comprising:
Current identification text obtains subelement, for the languages class by the current cypher text according to the target voice Type is converted, and obtains currently identifying text；
4th Semantic Similarity Measurement subelement is used for by the current identification text, in the machine recognition result The second text section in each text unit carry out Semantic Similarity Measurement；Wherein, second text section includes institute State the untranslated full text unit of the translator in machine recognition result or including in the machine recognition result N number of text unit that first untranslated text unit of the translator starts；
Second matching treatment text determines subelement, for by the corresponding text list of the maximum semantic similarity being calculated Member and the corresponding cypher text of text unit, as matching treatment text.
As an implementation, described device 600 further include:
Next voice unit predicting unit, for after getting the current cypher text, predicting the translator The next voice unit translated；
Next processing text acquiring unit, for obtaining the knowledge of next voice unit from the machine processing result Other text and/or cypher text；And/or it obtains with next voice unit from the machine processing result semantically The identification text and/or cypher text of similar voice unit；
Next processing textual presentation unit, identification text and/or cypher text for will acquire carry out special displaying.
As an implementation, described device 600 further include:
Text determination unit to be replaced is used in the machine translation result, the determining and current cypher text pair The cypher text answered, as text to be replaced；
Text replacement unit to be replaced, if the text to be replaced and the current cypher text semantically different or Text to be replaced described in person and the current cypher text it is semantically identical and in expression way it is different, then in the machine In translation result, the text to be replaced is replaced using the current cypher text；And/or
Text replacement unit is identified, if the corresponding identification text of the text to be replaced and the current cypher text are in language It is different in justice, then the current cypher text is converted according to the languages type of the target voice, is currently identified Text, and in the machine recognition result, the corresponding identification of the text to be replaced is replaced using the current identification text Text.
As an implementation, described device 600 further include:
Translation model updating unit, if for the machine translation the result is that being obtained using the translation model constructed in advance , then after the translator completes the translation to the target voice, using the translator to the target voice Human translation as a result, carry out parameter update to the translation model, make the translation model that there is the translator to turn over Translate style.
Further, the embodiment of the present application also provides a kind of textual presentation equipment, comprising: processor, memory, system Bus；
The processor and the memory are connected by the system bus；
The memory includes instruction, described instruction for storing one or more programs, one or more of programs The processor is set to execute any reality for the textual presentation method that above method embodiment provides when being executed by the processor Apply mode.
Further, described computer-readable to deposit the embodiment of the present application also provides a kind of computer readable storage medium Instruction is stored in storage media, when described instruction is run on the terminal device, so that the terminal device executes the above method Any embodiment for the textual presentation method that embodiment provides.
Further, the embodiment of the present application also provides a kind of computer program product, the computer program product exists When being run on terminal device, so that the terminal device executes any reality for the textual presentation method that above method embodiment provides Apply mode.
As seen through the above description of the embodiments, those skilled in the art can be understood that above-mentioned implementation All or part of the steps in example method can be realized by means of software and necessary general hardware platform.Based on such Understand, substantially the part that contributes to existing technology can be in the form of software products in other words for the technical solution of the application It embodies, which can store in storage medium, such as ROM/RAM, magnetic disk, CD, including several Instruction is used so that a computer equipment (can be the network communications such as personal computer, server, or Media Gateway Equipment, etc.) execute method described in certain parts of each embodiment of the application or embodiment.
It should be noted that each embodiment in this specification is described in a progressive manner, each embodiment emphasis is said Bright is the difference from other embodiments, and the same or similar parts in each embodiment may refer to each other.For reality For applying device disclosed in example, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place Referring to method part illustration.
It should also be noted that, herein, relational terms such as first and second and the like are used merely to one Entity or operation are distinguished with another entity or operation, without necessarily requiring or implying between these entities or operation There are any actual relationship or orders.Moreover, the terms "include", "comprise" or its any other variant are intended to contain Lid non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that There is also other identical elements in process, method, article or equipment including the element.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein General Principle can be realized in other embodiments without departing from the spirit or scope of the application.Therefore, the application It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one The widest scope of cause.

Claims (12)

1. a kind of textual presentation method characterized by comprising
Show the machine translation result of target voice；
Translator is obtained in real time to the translation result of the voice unit in the target voice, as current cypher text；
Obtain matching cypher text, it is described matching cypher text be from the machine translation result matching obtain work as with described Preceding cypher text is in semantically immediate cypher text；
Change the exhibition method of the matching cypher text.
2. the method according to claim 1, wherein the acquisition matches cypher text, comprising:
Each text unit in the first text section in the current cypher text, with the machine translation result is carried out Semantic Similarity Measurement；Wherein, first text section includes that the translator in the machine translation result does not turn over The full text unit or first text unit untranslated including the translator in the machine translation result translated The N number of text unit started；
Using the corresponding text unit of maximum semantic similarity being calculated as matching cypher text.
3. the method according to claim 1, wherein the method also includes:
After getting the current cypher text, next voice unit that the translator is translated is predicted；
The cypher text of next voice unit is obtained from the machine translation result, and/or, from the machine translation knot The cypher text with next voice unit in semantically similar voice unit is obtained in fruit；
The cypher text that will acquire carries out special displaying.
4. method according to any one of claims 1 to 3, which is characterized in that the method also includes:
In the machine translation result, using the matching cypher text as text to be replaced；
If the text to be replaced is worked as in the semantically different or described text to be replaced with described from the current cypher text Preceding cypher text it is semantically identical and in expression way it is different, then in the machine translation result, using described current Cypher text replaces the text to be replaced.
5. method according to any one of claims 1 to 3, which is characterized in that the machine translation is the result is that using preparatory What the translation model of building obtained, then the method also includes:
After the translator completes the translation to the target voice, using the translator to the target voice Human translation makes the translation model have the translation of the translator as a result, to translation model progress parameter update Style.
6. a kind of textual presentation method characterized by comprising
Show the machine recognition result of target voice；
Translator is obtained in real time to the translation result of the voice unit in the target voice, as current cypher text；
Obtain match cognization text, the match cognization text be from the machine recognition result matching obtain work as with described Preceding cypher text is in semantically immediate identification text；
Change the exhibition method of the match cognization text.
7. according to the method described in claim 6, it is characterized in that, the acquisition match cognization text, comprising:
The current cypher text is converted according to the languages type of the target voice, obtains currently identifying text；
Each text unit in the current identification text, with the second text section in the machine recognition result is carried out Semantic Similarity Measurement；Wherein, second text section includes that the translator in the machine recognition result does not turn over The full text unit or first text unit untranslated including the translator in the machine recognition result translated The N number of text unit started；
Using the corresponding text unit of maximum semantic similarity being calculated as match cognization text.
8. method according to claim 6 or 7, which is characterized in that the method also includes:
After getting the current cypher text, next voice unit that the translator is translated is predicted；
The identification text of next voice unit is obtained from the machine recognition result, and/or, from the machine recognition knot The identification text with next voice unit in semantically similar voice unit is obtained in fruit；
The identification text that will acquire carries out special displaying.
9. a kind of textual presentation method characterized by comprising
Show the machine processing of target voice as a result, the machine processing result includes the machine translation result of the target voice With machine recognition result；
Translator is obtained in real time to the translation result of the voice unit in the target voice, as current cypher text；
Obtain matching treatment text, the matching treatment text be from the machine processing result matching obtain work as with described Preceding cypher text is in semantically immediate cypher text and the identification text of the cypher text；
Change the exhibition method of the matching treatment text.
10. a kind of textual presentation device characterized by comprising
Machine translation result display unit, for showing the machine translation result of target voice；
Current cypher text acquiring unit, for obtaining translation of the translator to the voice unit in the target voice in real time As a result, as current cypher text；
Cypher text acquiring unit is matched, for obtaining matching cypher text, the matching cypher text is turned over from the machine Translate in result that matching obtains with the current cypher text in semantically immediate cypher text；
First exhibition method changing unit, for changing the exhibition method of the matching cypher text.
11. a kind of textual presentation device characterized by comprising
Machine recognition result display unit, for showing the machine recognition result of target voice；
Current cypher text acquiring unit, for obtaining translation of the translator to the voice unit in the target voice in real time As a result, as current cypher text；
Match cognization text acquiring unit, for obtaining match cognization text, the match cognization text is known from the machine In other result matching obtain with the current cypher text in semantically immediate identification text；
Second exhibition method changing unit, for changing the exhibition method of the match cognization text.
12. a kind of textual presentation device characterized by comprising
Machine processing result display unit, for showing the machine processing of target voice as a result, the machine processing result includes The machine translation result and machine recognition result of the target voice；
Current cypher text acquiring unit, for obtaining translation of the translator to the voice unit in the target voice in real time As a result, as current cypher text；
Matching treatment text acquiring unit, for obtaining matching treatment text, the matching treatment text is from the machine In reason result matching obtain with the current cypher text in semantically immediate cypher text and the cypher text Identify text；
Third exhibition method changing unit, for changing the exhibition method of the matching treatment text.
CN201910457348.6A 2019-05-29 2019-05-29 Text display method and device Active CN110175337B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910457348.6A CN110175337B (en) 2019-05-29 2019-05-29 Text display method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910457348.6A CN110175337B (en) 2019-05-29 2019-05-29 Text display method and device

Publications (2)

Publication Number Publication Date
CN110175337A true CN110175337A (en) 2019-08-27
CN110175337B CN110175337B (en) 2023-06-23

Family

ID=67695912

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910457348.6A Active CN110175337B (en) 2019-05-29 2019-05-29 Text display method and device

Country Status (1)

Country Link
CN (1) CN110175337B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111291575A (en) * 2020-02-28 2020-06-16 北京字节跳动网络技术有限公司 Text processing method and device, electronic equipment and storage medium
CN112818709A (en) * 2021-01-19 2021-05-18 传神语联网网络科技股份有限公司 Speech translation system and method for multi-user voice conference recording marks
CN116522960A (en) * 2023-05-08 2023-08-01 深圳市凝趣科技有限公司 Multi-language interactive real-time translation terminal and method supporting multiple platforms

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104050160A (en) * 2014-03-12 2014-09-17 北京紫冬锐意语音科技有限公司 Machine and human translation combined spoken language translation method and device
US20170262434A1 (en) * 2016-03-14 2017-09-14 Kabushiki Kaisha Toshiba Machine translation apparatus and machine translation method
WO2018000160A1 (en) * 2016-06-27 2018-01-04 李仁涛 Communication type speech translation device
CN107885737A (en) * 2017-12-27 2018-04-06 传神语联网网络科技股份有限公司 A kind of human-computer interaction interpretation method and system
CN108874787A (en) * 2018-06-12 2018-11-23 深圳市合言信息科技有限公司 A method of analysis speech intention simultaneously carries out depth translation explanation

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104050160A (en) * 2014-03-12 2014-09-17 北京紫冬锐意语音科技有限公司 Machine and human translation combined spoken language translation method and device
US20170262434A1 (en) * 2016-03-14 2017-09-14 Kabushiki Kaisha Toshiba Machine translation apparatus and machine translation method
WO2018000160A1 (en) * 2016-06-27 2018-01-04 李仁涛 Communication type speech translation device
CN107885737A (en) * 2017-12-27 2018-04-06 传神语联网网络科技股份有限公司 A kind of human-computer interaction interpretation method and system
CN108874787A (en) * 2018-06-12 2018-11-23 深圳市合言信息科技有限公司 A method of analysis speech intention simultaneously carries out depth translation explanation

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
张亚鹏等: "利用句法信息改进交互式机器翻译", 《中文信息学报》 *
张剑等: "机器翻译评测的新进展", 《中文信息学报》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111291575A (en) * 2020-02-28 2020-06-16 北京字节跳动网络技术有限公司 Text processing method and device, electronic equipment and storage medium
CN112818709A (en) * 2021-01-19 2021-05-18 传神语联网网络科技股份有限公司 Speech translation system and method for multi-user voice conference recording marks
CN112818709B (en) * 2021-01-19 2024-02-27 传神语联网网络科技股份有限公司 Speech translation system and method for recording marks of multi-user speech conferences
CN116522960A (en) * 2023-05-08 2023-08-01 深圳市凝趣科技有限公司 Multi-language interactive real-time translation terminal and method supporting multiple platforms
CN116522960B (en) * 2023-05-08 2023-10-20 深圳市凝趣科技有限公司 Multi-language interactive real-time translation terminal and method supporting multiple platforms

Also Published As

Publication number Publication date
CN110175337B (en) 2023-06-23

Similar Documents

Publication Publication Date Title
US10157171B2 (en) Annotation assisting apparatus and computer program therefor
CN104050160B (en) Interpreter's method and apparatus that a kind of machine is blended with human translation
US9922351B2 (en) Location-based adaptation of financial management system
CN106710592A (en) Speech recognition error correction method and speech recognition error correction device used for intelligent hardware equipment
CN110175337A (en) A kind of textual presentation method and device
CN111310440A (en) Text error correction method, device and system
Lefevre et al. Cross-lingual spoken language understanding from unaligned data using discriminative classification models and machine translation.
CN105573994B (en) Statictic machine translation system based on syntax skeleton
Barberis et al. Language resources for computer assisted translation from italian to italian sign language of deaf people
CN110263149A (en) A kind of textual presentation method and device
Popović Evaluating conjunction disambiguation on English-to-German and French-to-German WMT 2019 translation hypotheses
JP2016164707A (en) Automatic translation device and translation model learning device
Hashimoto et al. Impacts of machine translation and speech synthesis on speech-to-speech translation
CN106250367B (en) Method based on the improved Nivre algorithm building interdependent treebank of Vietnamese
CN112836525A (en) Human-computer interaction based machine translation system and automatic optimization method thereof
CN101930430A (en) Language text processing device and language learning device
TWI376656B (en) Foreign-language learning method utilizing an original language to review corresponding foreign languages and foreign-language learning database system thereof
CN113420124B (en) Method for resolving conflict under multiple conditions of voice retrieval
KR20120045906A (en) Apparatus and method for correcting error of corpus
CN110851572A (en) Session labeling method and device, storage medium and electronic equipment
Nasr et al. Automatically enriching spoken corpora with syntactic information for linguistic studies.
Miyata et al. Evaluating the suitability of human-oriented text simplification for machine translation
Hou et al. Design and Implementation of Interactive English Translation System in Internet of Things Auxiliary Information Processing
CN108766059A (en) A kind of cloud service English teaching equipment and teaching method
CN110473551A (en) A kind of audio recognition method, device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant