CN118070818B - Translation task processing method, device, equipment and storage medium - Google Patents
Translation task processing method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN118070818B CN118070818B CN202410472685.3A CN202410472685A CN118070818B CN 118070818 B CN118070818 B CN 118070818B CN 202410472685 A CN202410472685 A CN 202410472685A CN 118070818 B CN118070818 B CN 118070818B
- Authority
- CN
- China
- Prior art keywords
- target
- segment
- translation
- language
- speech
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/232—Orthographic correction, e.g. spell checking or vowelisation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Acoustics & Sound (AREA)
- Human Computer Interaction (AREA)
- Machine Translation (AREA)
Abstract
The invention belongs to the technical field of translation task processing, and discloses a translation task processing method, a translation task processing device, translation task processing equipment and a storage medium. The method comprises the following steps: acquiring a translation task, wherein the translation task comprises a translation requirement and data to be translated; determining an initial translation result of the data to be translated according to the translation requirement; determining a target speech segment in the initial translation result; adjusting the initial translation result according to the target speech segment to obtain a target translation result; judging whether the evaluation score of the target translation result is larger than a target score; and when the evaluation score of the target translation result is larger than the target score, the target translation result is used as the final processing result of the translation task. By the method, fluency among the speech segments in the translation result can be effectively enhanced.
Description
Technical Field
The present invention relates to the field of translation task processing technologies, and in particular, to a translation task processing method, device, equipment, and storage medium.
Background
With the acceleration of globalization, multilingual translation is becoming increasingly important. However, the traditional machine translation often has the problems of unsmooth translation between the front translation and the back translation of the language segments, hard full-text translation and the like.
Disclosure of Invention
The invention mainly aims to provide a translation task processing method, device, equipment and storage medium, and aims to solve the technical problems that in the prior art, traditional machine translation is unsmooth in translation of language segments before and after and the full-text translation is hard.
In order to achieve the above object, the present invention provides a translation task processing method, which includes the following steps:
acquiring a translation task, wherein the translation task comprises a translation requirement and data to be translated;
determining an initial translation result of the data to be translated according to the translation requirement;
determining a target speech segment in the initial translation result;
adjusting the initial translation result according to the target speech segment to obtain a target translation result;
judging whether the evaluation score of the target translation result is larger than a target score;
and when the evaluation score of the target translation result is larger than the target score, the target translation result is used as the final processing result of the translation task.
Optionally, the determining an initial translation result of the data to be translated according to the translation requirement includes:
when the data to be translated is voice data, converting the data to be translated into voice text;
Determining abnormal words in the voice text;
Determining a target word according to the context information and/or pronunciation information of the abnormal word;
Replacing the abnormal word in the voice text with the target word to obtain a target voice text;
and translating the target voice text to obtain an initial translation result of the data to be translated.
Optionally, the determining an initial translation result of the data to be translated according to the translation requirement includes:
Acquiring client information related to the translation task, wherein the client information comprises client position information and previous translation information;
determining target translation information in the previous translation information according to the position information;
analyzing the data to be translated to obtain the language information of the client;
And determining an initial translation result of the data to be translated according to the translation requirement, the target translation information and the client language information.
Optionally, the determining the target speech segment in the initial translation result includes:
splitting the initial translation result into a plurality of initial speech segments, and sequentially inputting each initial speech segment into a preset neural network model to obtain the speech segment score of each initial speech segment;
Setting a first score range and a second score range, wherein the score in the first score range is larger than the score in the second score range;
When the segment score of the initial segment is in the first score range, determining the initial segment as a first target segment in the target segments;
When the segment score of the initial segment is in the second score range, determining the initial segment as a second target segment in the target segments;
And determining the target speech segments except the first target speech segment and the second target speech segment in the target speech segments as a third target speech segment.
Optionally, the adjusting the initial translation result according to the target speech segment to obtain a target translation result includes:
determining adjacent language segments of the first target language segment, and adjusting the adjacent language segments according to the first target language segment to obtain a first adjusted language segment;
When the first adjacent speech segment of the first adjustment speech segment is a third target speech segment, judging whether the second adjacent speech segment of the third target speech segment is a second target speech segment or not;
When the second adjacent speech segment of the third target speech segment is judged to be the second target speech segment, the third target speech segment is adjusted according to the second target speech segment, and a second adjustment speech segment is obtained;
When the second adjacent language segment of the third target language segment is not the second target language segment, adjusting the third target language segment according to the first adjusting language segment to obtain a third adjusting language segment;
and adjusting the initial translation result based on the first target language segment, the second target language segment, the first adjustment language segment, the second adjustment language segment and the third adjustment language segment to obtain the target translation result.
Optionally, the adjusting the initial translation result according to the target speech segment to obtain a target translation result includes:
determining a first location of the first target segment and determining a second location of the second target segment;
determining a number of third target segments between the first target segment and the second target segment according to the first position and the second position;
judging whether the number is larger than a preset number or not;
When the number is not larger than the preset number, integrating a third target language segment between the first position and the second position into an integrated language segment, and adjusting the integrated language segment according to the first target language segment and the second target language segment to obtain a first integrated language segment;
When the number is judged to be larger than the preset number, dividing a third target speech segment between the first position and the second position into a first divided speech segment and a second divided speech segment according to a preset rule, adjusting the first divided speech segment according to the first target speech segment to obtain a first target divided speech segment, and adjusting the second divided speech segment according to the two target speech segments to obtain a second target divided speech segment;
And adjusting the initial translation result based on the first target language segment, the second target language segment, the first comprehensive language segment, the first target division language segment and the second target division language segment to obtain the target translation result.
Optionally, the determining whether the evaluation score of the target translation result is greater than a target score includes:
training an initial evaluation model to obtain a preset evaluation model;
and inputting the target translation result into the preset evaluation model to obtain an evaluation score of the target translation result, and judging whether the evaluation score is larger than a target score.
In addition, in order to achieve the above object, the present invention also provides a translation task processing device, including:
The system comprises an acquisition module, a translation module and a translation module, wherein the acquisition module is used for acquiring a translation task, and the translation task comprises a translation requirement and data to be translated;
The determining module is used for determining an initial translation result of the data to be translated according to the translation requirement;
The determining module is further used for determining a target speech segment in the initial translation result;
the adjustment module is used for adjusting the initial translation result according to the target speech segment to obtain a target translation result;
the judging module is used for judging whether the evaluation score of the target translation result is larger than the target score;
And the judging module is further used for taking the target translation result as a final processing result of the translation task when judging that the evaluation score of the target translation result is larger than the target score.
In addition, to achieve the above object, the present invention also proposes a translation task processing device including: a memory, a processor, and a translation task processing program stored on the memory and executable on the processor, the translation task processing program configured to implement the steps of the translation task processing method as described above.
In addition, in order to achieve the above object, the present invention also proposes a storage medium having stored thereon a translation task processing program which, when executed by a processor, implements the steps of the translation task processing method as described above.
The translation task processing method, the device, the equipment and the storage medium provided by the invention are used for acquiring the translation task, wherein the translation task comprises translation requirements and data to be translated; determining an initial translation result of the data to be translated according to the translation requirement; determining a target speech segment in the initial translation result; adjusting the initial translation result according to the target speech segment to obtain a target translation result; judging whether the evaluation score of the target translation result is larger than a target score; and when the evaluation score of the target translation result is larger than the target score, the target translation result is used as the final processing result of the translation task. According to the method, the translation data can be initially translated according to the translation requirement, then the speech segments with higher smoothness are selected from the speech segments of the initial translation result to adjust the initial translation result, then the target translation result is obtained, finally, when the target translation result is verified to meet the smoothness requirement, the target translation result is used as the final processing result of the translation task, the initial translation result is adjusted by using the target speech segments with higher smoothness, so that the linkage of each speech segment in the target translation result is stronger, the smoothness between the speech segments can be effectively enhanced, and finally, whether the target translation result meets the smoothness requirement or not is verified to be output, and the final processing result of the translation task can be further ensured to meet the smoothness requirement.
Drawings
FIG. 1 is a schematic diagram of a translation task processing device in a hardware runtime environment according to an embodiment of the present invention;
FIG. 2 is a flowchart of a translation task processing method according to a first embodiment of the present invention;
FIG. 3 is a flowchart illustrating a translation task processing method according to a second embodiment of the present invention;
FIG. 4 is a block diagram showing a first embodiment of a translation task processing device according to the present invention.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a translation task processing device in a hardware running environment according to an embodiment of the present invention.
As shown in fig. 1, the translation task processing device may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (Wi-Fi) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) Memory or a stable Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
It will be appreciated by those skilled in the art that the structure shown in fig. 1 is not limiting of the translation task processing device and may include more or fewer components than shown, or may combine certain components, or may be a different arrangement of components.
As shown in fig. 1, an operating system, a network communication module, a user interface module, and a translation task processing program may be included in the memory 1005 as one type of storage medium.
In the translation task processing device shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the translation task processing device of the present invention may be disposed in the translation task processing device, and the translation task processing device calls the translation task processing program stored in the memory 1005 through the processor 1001 and executes the translation task processing method provided by the embodiment of the present invention.
Based on the hardware structure, the embodiment of the translation task processing method is provided.
Referring to fig. 2, fig. 2 is a flowchart of a first embodiment of a translation task processing method according to the present invention.
In this embodiment, the translation task processing method includes the following steps:
step S10: and acquiring a translation task, wherein the translation task comprises a translation requirement and data to be translated.
It should be noted that, the execution body of the embodiment may be a computing service device with functions of data processing, network communication and program running, such as a mobile phone, a tablet computer, a personal computer, or an electronic device or a translation task processing device capable of implementing the above functions. The present embodiment and the following embodiments will be described below by taking the translation task processing device as an example.
It should be noted that, the translation task may be manually entered by the user, may be entered by the user through voice, or may be manually uploaded by the user; translation requirements may include language translation types (e.g., mid-translation english, in-english, daily translation, etc.), and may also include translation accuracy; the data to be translated may be voice data or text data.
Step S20: and determining an initial translation result of the data to be translated according to the translation requirement.
It should be noted that, when the translation requirement is in english translation and the data to be translated is speech data, the data to be translated needs to be converted into speech text data first, and then the english speech text data is translated into chinese speech text data, that is, the initial translation result; when the translation requirement is in english translation and the data to be translated is voice text data, the english voice text data can be directly translated into chinese voice text data, i.e. the initial translation result.
In an embodiment, the determining the initial translation result of the data to be translated according to the translation requirement includes:
when the data to be translated is voice data, converting the data to be translated into voice text;
Determining abnormal words in the voice text;
Determining a target word according to the context information and/or pronunciation information of the abnormal word;
Replacing the abnormal word in the voice text with the target word to obtain a target voice text;
and translating the target voice text to obtain an initial translation result of the data to be translated.
It can be understood that the abnormal word refers to a word in the voice text which is not connected with the context information, for example, one of the voice text is "i love to eat the blind country", then the abnormal word in the voice text can be determined to be "blind country", the context information of the voice text can be contacted to determine that the target word is "mango", the target word can be determined to be "mango" through the pronunciation information "mangguo" of the "blind country", or the target word can be determined to be "mango" through the context information and the pronunciation information of the voice text; phonetic text can be input into the deep learning model and abnormal words of the phonetic text can be output.
In this embodiment, it is necessary to determine an abnormal word of the voice text in advance, and replace the abnormal word with a target word and then translate the target word, so that it is possible to avoid that the translation result of the data to be translated is affected due to the existence of the abnormal word in the voice text, and further effectively ensure the translation quality.
In an embodiment, the determining the initial translation result of the data to be translated according to the translation requirement includes:
Acquiring client information related to the translation task, wherein the client information comprises client position information and previous translation information;
determining target translation information in the previous translation information according to the position information;
analyzing the data to be translated to obtain the language information of the client;
And determining an initial translation result of the data to be translated according to the translation requirement, the target translation information and the client language information.
It should be noted that, the previous translation information refers to the translation information of the user before the current time; the user's translation requirements at different locations are different, such as when the user is in the uk, the user's translation requirements are in the middle translation or the english translation, and also such as when the user is in the france, the user's translation requirements are in the middle translation or the french, so that the target translation information in the previous translation information can be determined according to the location information of the user, such as when the user is in the uk, the target translation information is the translation information related to the location information (in the middle translation or the english translation) in the previous translation information.
In a specific implementation, the client mood information can be determined according to the mood words in the data to be translated, for example, when a plurality of cheerful mood words exist in the data to be translated, the client mood information can be determined to be cheerful mood; the method can also preset a preset language model, and the data to be translated can be directly input into the preset language model to determine the language information of the client.
It can be understood that when the data to be translated is translated, the translation needs to be performed according to the translation requirement, the target translation information and the customer language information, so that the statement smoothness of the initial translation result can be ensured to be higher.
In this embodiment, the translation of the data to be translated in combination with the translation requirement, the target translation information and the customer language information can effectively ensure the accuracy of the initial translation result, and also can accelerate the adjustment rate of the initial translation result.
Step S30: and determining a target speech segment in the initial translation result.
It should be noted that the target speech segment refers to a speech segment with higher fluency in the initial translation result, for example, the target speech segment may be a speech segment with fluency higher than the target fluency in the initial translation result.
Step S40: and adjusting the initial translation result according to the target speech segment to obtain a target translation result.
In a specific implementation, the surrounding speech segments of the target speech segment can be determined first, then the surrounding speech segments are adjusted according to the target speech segment, then the unadjusted non-target speech segment is continuously adjusted according to the adjusted speech segment, and finally the target translation result can be obtained.
Step S50: and judging whether the evaluation score of the target translation result is larger than a target score.
It should be noted that, the higher the fluency of the speech segment of the target translation result is, the higher the evaluation score of the target translation result is, otherwise, the lower the evaluation score of the target translation result is; a preset evaluation model can be set, and the evaluation score of the target translation result can be output by inputting the target translation result into the preset evaluation model; the target score may be set in advance.
Step S60: and when the evaluation score of the target translation result is larger than the target score, the target translation result is used as the final processing result of the translation task.
It should be noted that, when the evaluation score of the target translation result is greater than the target score, the target translation result may be used as the final processing result of the translation task; when the evaluation score of the target translation result is not greater than the target score, the target translation result needs to be continuously adjusted until the evaluation score of the target translation result is greater than the target score.
In an embodiment, the determining whether the evaluation score of the target translation result is greater than a target score includes:
training an initial evaluation model to obtain a preset evaluation model;
and inputting the target translation result into the preset evaluation model to obtain an evaluation score of the target translation result, and judging whether the evaluation score is larger than a target score.
It should be noted that, an initial evaluation model for evaluating the target translation result may be preset, a large amount of evaluation data is used to train the initial evaluation model to obtain a preset evaluation model, and the target translation result is input into the preset evaluation model to obtain the evaluation score of the target translation result.
The embodiment obtains a translation task, wherein the translation task comprises a translation requirement and data to be translated; determining an initial translation result of the data to be translated according to the translation requirement; determining a target speech segment in the initial translation result; adjusting the initial translation result according to the target speech segment to obtain a target translation result; judging whether the evaluation score of the target translation result is larger than a target score; and when the evaluation score of the target translation result is larger than the target score, the target translation result is used as the final processing result of the translation task. According to the method, the translation data can be initially translated according to the translation requirement, then the speech segments with higher smoothness are selected from the speech segments of the initial translation result to adjust the initial translation result, then the target translation result is obtained, finally, when the target translation result is verified to meet the smoothness requirement, the target translation result is used as the final processing result of the translation task, the initial translation result is adjusted by using the target speech segments with higher smoothness, so that the linkage of each speech segment in the target translation result is stronger, the smoothness between the speech segments can be effectively enhanced, and finally, whether the target translation result meets the smoothness requirement or not is verified to be output, and the final processing result of the translation task can be further ensured to meet the smoothness requirement.
Referring to fig. 3, fig. 3 is a flowchart of a translation task processing method according to a second embodiment of the present invention.
Based on the above first embodiment, the determining, by the translation task processing method of the present embodiment, the target speech segment in the initial translation result includes:
step S301: splitting the initial translation result into a plurality of initial speech segments, and sequentially inputting each initial speech segment into a preset neural network model to obtain the speech segment score of each initial speech segment.
It should be noted that, the initial translation result may be split into a plurality of initial segments according to punctuation marks in the initial translation result; presetting a neural network model as a trained neural network model; the paragraph score is used to evaluate the sentence consistency of the initial paragraph, and the higher the paragraph score, the better the sentence consistency (fluency) of the initial paragraph.
Step S302: setting a first score range and a second score range, wherein the score in the first score range is larger than the score in the second score range.
Step S303: and when the segment score of the initial segment is in the first score range, determining the initial segment as a first target segment in the target segments.
Step S304: and when the segment score of the initial segment is in the second score range, determining the initial segment as a second target segment in the target segments.
Step S305: and determining the target speech segments except the first target speech segment and the second target speech segment in the target speech segments as a third target speech segment.
It can be understood that the sentence fluency of the first target sentence is higher than that of the second target sentence; the sentence fluency of the third target language segment is low, and the third target language segment needs to be adjusted according to the first target language segment and the second target language segment, so that the sentence of the whole translation result is smoother.
In an embodiment, the adjusting the initial translation result according to the target speech segment to obtain a target translation result includes:
determining adjacent language segments of the first target language segment, and adjusting the adjacent language segments according to the first target language segment to obtain a first adjusted language segment;
When the first adjacent speech segment of the first adjustment speech segment is a third target speech segment, judging whether the second adjacent speech segment of the third target speech segment is a second target speech segment or not;
When the second adjacent speech segment of the third target speech segment is judged to be the second target speech segment, the third target speech segment is adjusted according to the second target speech segment, and a second adjustment speech segment is obtained;
When the second adjacent language segment of the third target language segment is not the second target language segment, adjusting the third target language segment according to the first adjusting language segment to obtain a third adjusting language segment;
and adjusting the initial translation result based on the first target language segment, the second target language segment, the first adjustment language segment, the second adjustment language segment and the third adjustment language segment to obtain the target translation result.
It should be noted that, the number of adjacent speech segments of the target speech segment is generally 2, and the adjacent speech segments are respectively the speech segments on the left side of the target speech segment and the speech segments on the right side of the target speech segment; the first adjacent speech segment refers to the rest adjacent speech segments except the first target speech segment in the adjacent speech segments of the first adjustment speech segment; the second adjacent segment refers to the remaining adjacent segments except for the first adjustment segment among the adjacent segments of the third target segment.
It can be understood that, because the sentence fluency of the first target speech segment is higher, the adjacent speech segment can be adjusted according to the first target speech segment to obtain the first adjusted speech segment, then the speech segment with other fluency not higher than the first adjusted speech segment is adjusted continuously, whether the adjacent speech segment of the third target speech segment is the second target speech segment can be judged, if yes, the speech segment with other fluency not higher than the first fluency is adjusted directly according to the second target speech segment, and if not, the speech segment can be adjusted continuously according to the first adjusted speech segment.
In this embodiment, the first target speech segment with higher fluency may be first adopted to adjust the third target speech segment to obtain the first adjusted speech segment, then it is determined whether the second target speech segment exists around the third target speech segment of the first adjusted speech segment, if yes, the second target speech segment is adopted to adjust the third target speech segment, if not, the first adjusted speech segment is continuously adopted to adjust the third target speech segment, so that it can be ensured that the fluency and higher speech segment is adopted to perform every time the third target speech segment is adjusted, and further sentence fluency of the target translation result is improved.
In an embodiment, the adjusting the initial translation result according to the target speech segment to obtain a target translation result includes:
determining a first location of the first target segment and determining a second location of the second target segment;
determining a number of third target segments between the first target segment and the second target segment according to the first position and the second position;
judging whether the number is larger than a preset number or not;
When the number is not larger than the preset number, integrating a third target language segment between the first position and the second position into an integrated language segment, and adjusting the integrated language segment according to the first target language segment and the second target language segment to obtain a first integrated language segment;
When the number is judged to be larger than the preset number, dividing a third target speech segment between the first position and the second position into a first divided speech segment and a second divided speech segment according to a preset rule, adjusting the first divided speech segment according to the first target speech segment to obtain a first target divided speech segment, and adjusting the second divided speech segment according to the two target speech segments to obtain a second target divided speech segment;
And adjusting the initial translation result based on the first target language segment, the second target language segment, the first comprehensive language segment, the first target division language segment and the second target division language segment to obtain the target translation result.
It should be noted that, when the number of the third target speech segments between the first position and the second position is small, the third target speech segments can be directly integrated into a comprehensive speech segment and then adjusted by the first target speech segment and the second target speech segment together; when the number of the third target speech segments between the first position and the second position is large, the third target speech segments can be divided into a first divided speech segment and a second divided speech segment according to preset rules, and the first divided speech segment is close to the first target speech segment, the second divided speech segment is closer to the second target speech segment, so that the first divided speech segment can be adjusted according to the first target speech segment, and the second divided speech segment can be adjusted according to the second target speech segment.
In this embodiment, the first position of the first target speech segment and the second position of the second target speech segment may be predetermined, and then a specific adjustment mode may be determined according to the number of speech segments between the first position and the second position, for example, when the number of speech segments is less than or equal to a preset number, all speech segments between the first position and the second position may be directly integrated into an integrated speech segment and then adjusted, for example, when the number of speech segments is greater than the preset number, the number of speech segments may be divided into two speech segments, where one speech segment is adjusted by the first target speech segment and the other speech segment is adjusted by the second target speech segment.
According to the embodiment, the initial translation result is split into a plurality of initial speech segments, and each initial speech segment is sequentially input into a preset neural network model to obtain the speech segment score of each initial speech segment; setting a first score range and a second score range, wherein the score in the first score range is larger than the score in the second score range; when the segment score of the initial segment is in the first score range, determining the initial segment as a first target segment in the target segments; when the segment score of the initial segment is in the second score range, determining the initial segment as a second target segment in the target segments; and determining the target speech segments except the first target speech segment and the second target speech segment in the target speech segments as a third target speech segment. Through the method, the speech segment score of each initial speech segment in the initial translation result can be determined according to the preset neural network model, then the target speech segments are divided according to the speech segment scores, finally the target speech segments are adjusted according to the divided target speech segments, and therefore the initial translation result can be adjusted more accurately, and the adjustment rate and the adjustment accuracy are improved.
In addition, the embodiment of the invention also provides a storage medium, wherein the storage medium stores a translation task processing program, and the translation task processing program realizes the steps of the translation task processing method when being executed by a processor.
Referring to fig. 4, fig. 4 is a block diagram showing the construction of a first embodiment of the translation task processing device according to the present invention.
As shown in fig. 4, the translation task processing device provided in the embodiment of the present invention includes:
the obtaining module 10 is configured to obtain a translation task, where the translation task includes a translation requirement and data to be translated.
And the determining module 20 is configured to determine an initial translation result of the data to be translated according to the translation requirement.
The determining module 20 is further configured to determine a target speech segment in the initial translation result.
And the adjustment module 30 is configured to adjust the initial translation result according to the target speech segment to obtain a target translation result.
A judging module 40, configured to judge whether the evaluation score of the target translation result is greater than a target score.
The judging module 40 is further configured to, when it is determined that the evaluation score of the target translation result is greater than the target score, use the target translation result as a final processing result of the translation task.
It should be understood that the foregoing is illustrative only and is not limiting, and that in specific applications, those skilled in the art may set the invention as desired, and the invention is not limited thereto.
The embodiment obtains a translation task, wherein the translation task comprises a translation requirement and data to be translated; determining an initial translation result of the data to be translated according to the translation requirement; determining a target speech segment in the initial translation result; adjusting the initial translation result according to the target speech segment to obtain a target translation result; judging whether the evaluation score of the target translation result is larger than a target score; and when the evaluation score of the target translation result is larger than the target score, the target translation result is used as the final processing result of the translation task. According to the method, the translation data can be initially translated according to the translation requirement, then the speech segments with higher smoothness are selected from the speech segments of the initial translation result to adjust the initial translation result, then the target translation result is obtained, finally, when the target translation result is verified to meet the smoothness requirement, the target translation result is used as the final processing result of the translation task, the initial translation result is adjusted by using the target speech segments with higher smoothness, so that the linkage of each speech segment in the target translation result is stronger, the smoothness between the speech segments can be effectively enhanced, and finally, whether the target translation result meets the smoothness requirement or not is verified to be output, and the final processing result of the translation task can be further ensured to meet the smoothness requirement.
In an embodiment, the determining the initial translation result of the data to be translated according to the translation requirement includes:
when the data to be translated is voice data, converting the data to be translated into voice text;
Determining abnormal words in the voice text;
Determining a target word according to the context information and/or pronunciation information of the abnormal word;
Replacing the abnormal word in the voice text with the target word to obtain a target voice text;
and translating the target voice text to obtain an initial translation result of the data to be translated.
In an embodiment, the determining the initial translation result of the data to be translated according to the translation requirement includes:
Acquiring client information related to the translation task, wherein the client information comprises client position information and previous translation information;
determining target translation information in the previous translation information according to the position information;
analyzing the data to be translated to obtain the language information of the client;
And determining an initial translation result of the data to be translated according to the translation requirement, the target translation information and the client language information.
In an embodiment, the determining the target speech segment in the initial translation result includes:
splitting the initial translation result into a plurality of initial speech segments, and sequentially inputting each initial speech segment into a preset neural network model to obtain the speech segment score of each initial speech segment;
Setting a first score range and a second score range, wherein the score in the first score range is larger than the score in the second score range;
When the segment score of the initial segment is in the first score range, determining the initial segment as a first target segment in the target segments;
When the segment score of the initial segment is in the second score range, determining the initial segment as a second target segment in the target segments;
And determining the target speech segments except the first target speech segment and the second target speech segment in the target speech segments as a third target speech segment.
In an embodiment, the adjusting the initial translation result according to the target speech segment to obtain a target translation result includes:
determining adjacent language segments of the first target language segment, and adjusting the adjacent language segments according to the first target language segment to obtain a first adjusted language segment;
When the first adjacent speech segment of the first adjustment speech segment is a third target speech segment, judging whether the second adjacent speech segment of the third target speech segment is a second target speech segment or not;
When the second adjacent speech segment of the third target speech segment is judged to be the second target speech segment, the third target speech segment is adjusted according to the second target speech segment, and a second adjustment speech segment is obtained;
When the second adjacent language segment of the third target language segment is not the second target language segment, adjusting the third target language segment according to the first adjusting language segment to obtain a third adjusting language segment;
and adjusting the initial translation result based on the first target language segment, the second target language segment, the first adjustment language segment, the second adjustment language segment and the third adjustment language segment to obtain the target translation result.
In an embodiment, the adjusting the initial translation result according to the target speech segment to obtain a target translation result includes:
determining a first location of the first target segment and determining a second location of the second target segment;
determining a number of third target segments between the first target segment and the second target segment according to the first position and the second position;
judging whether the number is larger than a preset number or not;
When the number is not larger than the preset number, integrating a third target language segment between the first position and the second position into an integrated language segment, and adjusting the integrated language segment according to the first target language segment and the second target language segment to obtain a first integrated language segment;
When the number is judged to be larger than the preset number, dividing a third target speech segment between the first position and the second position into a first divided speech segment and a second divided speech segment according to a preset rule, adjusting the first divided speech segment according to the first target speech segment to obtain a first target divided speech segment, and adjusting the second divided speech segment according to the two target speech segments to obtain a second target divided speech segment;
And adjusting the initial translation result based on the first target language segment, the second target language segment, the first comprehensive language segment, the first target division language segment and the second target division language segment to obtain the target translation result.
In an embodiment, the determining whether the evaluation score of the target translation result is greater than a target score includes:
training an initial evaluation model to obtain a preset evaluation model;
and inputting the target translation result into the preset evaluation model to obtain an evaluation score of the target translation result, and judging whether the evaluation score is larger than a target score.
It should be noted that the above-described working procedure is merely illustrative, and does not limit the scope of the present invention, and in practical application, a person skilled in the art may select part or all of them according to actual needs to achieve the purpose of the embodiment, which is not limited herein.
In addition, technical details that are not described in detail in this embodiment may refer to the translation task processing method provided in any embodiment of the present invention, which is not described herein again.
Furthermore, it should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of embodiments, it will be clear to a person skilled in the art that the above embodiment method may be implemented by means of software plus a necessary general hardware platform, but may of course also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. Read Only Memory (ROM)/RAM, magnetic disk, optical disk) and comprising several instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (7)
1. The translation task processing method is characterized by comprising the following steps of:
acquiring a translation task, wherein the translation task comprises a translation requirement and data to be translated;
determining an initial translation result of the data to be translated according to the translation requirement;
determining a target speech segment in the initial translation result;
adjusting the initial translation result according to the target speech segment to obtain a target translation result;
judging whether the evaluation score of the target translation result is larger than a target score;
When the evaluation score of the target translation result is larger than the target score, the target translation result is used as a final processing result of the translation task;
The determining the target speech segment in the initial translation result comprises the following steps:
splitting the initial translation result into a plurality of initial speech segments, and sequentially inputting each initial speech segment into a preset neural network model to obtain the speech segment score of each initial speech segment;
Setting a first score range and a second score range, wherein the score in the first score range is larger than the score in the second score range;
When the segment score of the initial segment is in the first score range, determining the initial segment as a first target segment in the target segments;
When the segment score of the initial segment is in the second score range, determining the initial segment as a second target segment in the target segments;
determining a target speech segment except the first target speech segment and the second target speech segment in the target speech segments as a third target speech segment;
the step of adjusting the initial translation result according to the target speech segment to obtain a target translation result comprises the following steps:
Determining adjacent language segments of the first target language segment, and adjusting the adjacent language segments according to the first target language segment to obtain a first adjusted language segment; when the first adjacent speech segment of the first adjustment speech segment is a third target speech segment, judging whether the second adjacent speech segment of the third target speech segment is a second target speech segment or not; when the second adjacent speech segment of the third target speech segment is judged to be the second target speech segment, the third target speech segment is adjusted according to the second target speech segment, and a second adjustment speech segment is obtained; when the second adjacent language segment of the third target language segment is not the second target language segment, adjusting the third target language segment according to the first adjusting language segment to obtain a third adjusting language segment; adjusting the initial translation result based on the first target language segment, the second target language segment, the first adjustment language segment, the second adjustment language segment and the third adjustment language segment to obtain the target translation result;
Or, determining a first location of the first target segment and determining a second location of the second target segment; determining a number of third target segments between the first target segment and the second target segment according to the first position and the second position; judging whether the number is larger than a preset number or not; when the number is not larger than the preset number, integrating a third target language segment between the first position and the second position into an integrated language segment, and adjusting the integrated language segment according to the first target language segment and the second target language segment to obtain a first integrated language segment; when the number is judged to be larger than the preset number, dividing a third target speech segment between the first position and the second position into a first divided speech segment and a second divided speech segment according to a preset rule, adjusting the first divided speech segment according to the first target speech segment to obtain a first target divided speech segment, and adjusting the second divided speech segment according to the two target speech segments to obtain a second target divided speech segment; and adjusting the initial translation result based on the first target language segment, the second target language segment, the first comprehensive language segment, the first target division language segment and the second target division language segment to obtain the target translation result.
2. The method of claim 1, wherein the determining an initial translation result of the data to be translated based on the translation requirement comprises:
when the data to be translated is voice data, converting the data to be translated into voice text;
Determining abnormal words in the voice text;
Determining a target word according to the context information and/or pronunciation information of the abnormal word;
Replacing the abnormal word in the voice text with the target word to obtain a target voice text;
and translating the target voice text to obtain an initial translation result of the data to be translated.
3. The method of claim 1, wherein the determining an initial translation result of the data to be translated based on the translation requirement comprises:
Acquiring client information related to the translation task, wherein the client information comprises client position information and previous translation information;
determining target translation information in the previous translation information according to the position information;
analyzing the data to be translated to obtain the language information of the client;
And determining an initial translation result of the data to be translated according to the translation requirement, the target translation information and the client language information.
4. The method of claim 1, wherein the determining whether the evaluation score of the target translation result is greater than a target score comprises:
training an initial evaluation model to obtain a preset evaluation model;
and inputting the target translation result into the preset evaluation model to obtain an evaluation score of the target translation result, and judging whether the evaluation score is larger than a target score.
5. A translation task processing device, characterized in that the translation task processing device includes:
The system comprises an acquisition module, a translation module and a translation module, wherein the acquisition module is used for acquiring a translation task, and the translation task comprises a translation requirement and data to be translated;
The determining module is used for determining an initial translation result of the data to be translated according to the translation requirement;
The determining module is further used for determining a target speech segment in the initial translation result;
the adjustment module is used for adjusting the initial translation result according to the target speech segment to obtain a target translation result;
the judging module is used for judging whether the evaluation score of the target translation result is larger than the target score;
The judging module is further configured to, when it is determined that the evaluation score of the target translation result is greater than the target score, use the target translation result as a final processing result of the translation task;
The determining module is further configured to:
splitting the initial translation result into a plurality of initial speech segments, and sequentially inputting each initial speech segment into a preset neural network model to obtain the speech segment score of each initial speech segment;
Setting a first score range and a second score range, wherein the score in the first score range is larger than the score in the second score range;
When the segment score of the initial segment is in the first score range, determining the initial segment as a first target segment in the target segments;
When the segment score of the initial segment is in the second score range, determining the initial segment as a second target segment in the target segments;
determining a target speech segment except the first target speech segment and the second target speech segment in the target speech segments as a third target speech segment;
The determining module is further configured to:
Determining adjacent language segments of the first target language segment, and adjusting the adjacent language segments according to the first target language segment to obtain a first adjusted language segment; when the first adjacent speech segment of the first adjustment speech segment is a third target speech segment, judging whether the second adjacent speech segment of the third target speech segment is a second target speech segment or not; when the second adjacent speech segment of the third target speech segment is judged to be the second target speech segment, the third target speech segment is adjusted according to the second target speech segment, and a second adjustment speech segment is obtained; when the second adjacent language segment of the third target language segment is not the second target language segment, adjusting the third target language segment according to the first adjusting language segment to obtain a third adjusting language segment; adjusting the initial translation result based on the first target language segment, the second target language segment, the first adjustment language segment, the second adjustment language segment and the third adjustment language segment to obtain the target translation result;
Or, determining a first location of the first target segment and determining a second location of the second target segment; determining a number of third target segments between the first target segment and the second target segment according to the first position and the second position; judging whether the number is larger than a preset number or not; when the number is not larger than the preset number, integrating a third target language segment between the first position and the second position into an integrated language segment, and adjusting the integrated language segment according to the first target language segment and the second target language segment to obtain a first integrated language segment; when the number is judged to be larger than the preset number, dividing a third target speech segment between the first position and the second position into a first divided speech segment and a second divided speech segment according to a preset rule, adjusting the first divided speech segment according to the first target speech segment to obtain a first target divided speech segment, and adjusting the second divided speech segment according to the two target speech segments to obtain a second target divided speech segment; and adjusting the initial translation result based on the first target language segment, the second target language segment, the first comprehensive language segment, the first target division language segment and the second target division language segment to obtain the target translation result.
6. A translation task processing device, the device comprising: a memory, a processor and a translation task processing program stored on the memory and executable on the processor, the translation task processing program being configured to implement the steps of the translation task processing method according to any one of claims 1 to 4.
7. A storage medium having stored thereon a translation task processing program which, when executed by a processor, implements the steps of the translation task processing method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410472685.3A CN118070818B (en) | 2024-04-19 | 2024-04-19 | Translation task processing method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410472685.3A CN118070818B (en) | 2024-04-19 | 2024-04-19 | Translation task processing method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118070818A CN118070818A (en) | 2024-05-24 |
| CN118070818B true CN118070818B (en) | 2024-08-02 |
Family
ID=91102495
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410472685.3A Active CN118070818B (en) | 2024-04-19 | 2024-04-19 | Translation task processing method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118070818B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116468052A (en) * | 2023-04-03 | 2023-07-21 | 网易（杭州）网络有限公司 | Translation task processing method and device, storage medium and electronic device |
| CN117592536A (en) * | 2023-11-28 | 2024-02-23 | 中银金融科技有限公司 | Translation model training method, device, electronic equipment and storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016071439A (en) * | 2014-09-26 | 2016-05-09 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカＰａｎａｓｏｎｉｃ Ｉｎｔｅｌｌｅｃｔｕａｌ Ｐｒｏｐｅｒｔｙ Ｃｏｒｐｏｒａｔｉｏｎ ｏｆ Ａｍｅｒｉｃａ | Translation method and translation system |
| CN111027331B (en) * | 2019-12-05 | 2022-04-05 | 百度在线网络技术（北京）有限公司 | Method and apparatus for evaluating translation quality |
-
2024
- 2024-04-19 CN CN202410472685.3A patent/CN118070818B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116468052A (en) * | 2023-04-03 | 2023-07-21 | 网易（杭州）网络有限公司 | Translation task processing method and device, storage medium and electronic device |
| CN117592536A (en) * | 2023-11-28 | 2024-02-23 | 中银金融科技有限公司 | Translation model training method, device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118070818A (en) | 2024-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5462001B2 (en) | Contextual input method | |
| US11669698B2 (en) | Method and system for automatic formality classification | |
| US8612206B2 (en) | Transliterating semitic languages including diacritics | |
| US20140122056A1 (en) | Chatbot system and method with enhanced user communication | |
| CN112084769B (en) | Dependency syntax model optimization method, apparatus, device and readable storage medium | |
| US20170185590A1 (en) | Method and system for automatic formality transformation | |
| KR20110081194A (en) | System for extracting terms from documents with text segments | |
| O'Brien | Machine translation and cognition | |
| JP7537555B2 (en) | Scoring support device, scoring support method and program | |
| CN111177307A (en) | Test scheme and system based on semantic understanding similarity threshold configuration | |
| CN113312463B (en) | Intelligent evaluation method and device for voice questions and answers, computer equipment and storage medium | |
| CN111444729A (en) | Information processing method, device, equipment and readable storage medium | |
| US11256409B2 (en) | Systems, methods, devices, and computer readable media for facilitating distributed processing of documents | |
| KR20220084915A (en) | System for providing cloud based grammar checker service | |
| Walker | 20 Variation analysis | |
| KR20140066921A (en) | Apparatus and method for evaluating machine translation | |
| CN111737961A (en) | Method and device for generating story, computer equipment and medium | |
| TW201544976A (en) | Natural language processing system, natural language processing method, and natural language processing program | |
| CN118070818B (en) | Translation task processing method, device, equipment and storage medium | |
| CN115101151A (en) | Character testing method and device based on man-machine conversation and electronic equipment | |
| JP2021022292A (en) | Information processor, program, and information processing method | |
| US8504580B2 (en) | Systems and methods for creating an artificial intelligence | |
| CN112417849A (en) | English mail text data processing method, device, equipment and storage medium | |
| CN119398063B (en) | Text translation methods, devices, electronic equipment, and storage media | |
| CN112084316A (en) | Training method and device of emotion recognition model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |