CN118364829B - Multi-language offline intelligent translation system and method - Google Patents
Multi-language offline intelligent translation system and method Download PDFInfo
- Publication number
- CN118364829B CN118364829B CN202410788983.3A CN202410788983A CN118364829B CN 118364829 B CN118364829 B CN 118364829B CN 202410788983 A CN202410788983 A CN 202410788983A CN 118364829 B CN118364829 B CN 118364829B
- Authority
- CN
- China
- Prior art keywords
- translation
- text
- offline
- target language
- data packet
- 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
- G06F40/30—Semantic analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The embodiment of the application relates to the technical field of computers, and discloses a multi-language offline intelligent translation system and a multi-language offline intelligent translation method, which respond to the demands of users on the timeliness and the accuracy of text translation, and automatically selecting different translation data packages such as the first translation data package, the second translation data package and the third translation data package to perform translation processing of the text so as to obtain a translation text of the second target language. Therefore, the user can select translation timeliness according to different translation requirements, and therefore requirements of the user on translation instantaneity and translation accuracy under different states are met.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a multilingual offline intelligent translation system and method.
Background
The offline translation device is convenient to carry and is not influenced by a network, so that the offline translation device can be used at any time and any place, but in the related technology, the offline translation device adopts the offline translation data packet trained in advance to carry out accurate translation of sentences, and the offline translation data packet cannot adapt to various types of sentences, so that the translation efficiency is lower under specific conditions, and the requirements of users on translation instantaneity and translation accuracy under different states cannot be met.
Disclosure of Invention
The invention mainly aims to provide a multilingual offline intelligent translation system and method, and aims to solve the technical problem that offline translation equipment in the prior art cannot meet requirements of users on translation instantaneity and translation accuracy in different states.
In order to achieve the above object, in a first aspect, an embodiment of the present application provides a multi-language offline intelligent translation method, which is applied to an intelligent offline translation device, where the method includes:
acquiring an offline translation instruction, wherein the offline translation instruction carries first target language text information to be translated and translation aging information, and the translation aging information comprises fast translation and slow translation;
When the translation aging information is slow translation, determining the text length and the semantic association degree of the first target language text according to the first target language text information;
Selecting a target offline translation data packet to perform offline translation according to the text length and the semantic relevance of the first target language text, wherein the target offline translation data packet comprises a first offline translation data packet and a second offline translation data packet, the first offline translation data packet is used for translating simple texts, and the second offline translation data packet is used for translating complex texts;
performing offline translation in response to the second offline translation data packet to obtain an initial translation text;
And carrying out offline correction on the initial translation text by adopting a pre-trained offline correction model to obtain a second target language translation text, wherein the offline correction model is an error correction model obtained by training according to historical manual error correction big data.
In one possible implementation manner, after selecting a target offline translation data packet according to the text length and the semantic relevance of the first target language text to perform offline translation, the method further includes:
performing offline translation in response to the first offline translation data packet to obtain an initial translation text;
Performing feature recognition on the initial translation text according to a preset recognition rule to obtain a segmentation boundary, wherein the preset recognition rule is a recognition rule obtained by training in advance and is used for performing feature recognition of the corresponding recognition rule on the initial translation text;
and performing a spacer insertion operation before or after the segmentation boundary to obtain the second target language translation text.
In one possible implementation manner, after the acquiring the offline translation instruction, the method further includes:
And when the translation aging information is fast translation, selecting a third offline translation data packet for offline translation, wherein the third offline translation data packet directly translates the first target language text into the second target language translation text.
In one possible implementation manner, the determining the text length and the semantic association degree of the first target language text according to the first target language text information includes:
Calling a text length function to perform length recognition on the first target language text information to obtain the text length of the first target language text;
Inputting the first target language text information into an offline semantic analysis model to obtain the semantic relevance of the first target language text, wherein the offline semantic analysis model is a pre-trained semantic analysis algorithm model and is used for analyzing the text semantic relevance of the first target language text.
In one possible implementation manner, the selecting a target offline translation data packet according to the text length and the semantic relevance of the first target language text for offline translation includes:
determining that the text length of the first target language text is smaller than or equal to a first preset length value, and selecting the first offline translation data packet for offline translation; or determining that the text length of the first target language text is larger than a first preset length value and smaller than a second preset length value, and the semantic association is smaller than the preset association, and selecting the first offline translation data packet to perform offline translation;
determining that the text length of the first target language text is larger than a second preset length value, and selecting the second offline translation data packet for offline translation; or determining that the text length of the first target language text is larger than a first preset length value and smaller than a second preset length value, and the semantic association is larger than or equal to the preset association, and selecting the second offline translation data packet to perform offline translation;
wherein the first preset length value is smaller than the second preset length value.
In one possible implementation manner, the offline correction model includes a first error correction model, the first error correction model corrects for a first language element error, and the offline correction of the initial translation text by using a pre-trained offline correction model to obtain a second target language translation text includes:
detecting whether the primary translation text has a first language element error;
And if the initial translation text has the first language element error, carrying out first language element error correction on the initial translation text by adopting the first error correction model to obtain a second target language translation text.
In one possible implementation manner, the offline correction model includes a second error correction model, the second error correction model corrects for a second language element error, and the offline correction of the initial translation text by using a pre-trained offline correction model obtains a second target language translation text, including:
Detecting whether the primary translation text has a second language element error;
And if the initial translation text has the second language element error, carrying out second language element error correction on the initial translation text by adopting the second error correction model to obtain a second target language translation text.
In one possible implementation manner, before the selecting a target offline translation data packet according to the text length and the semantic relevance of the first target language text to perform offline translation, the method further includes:
Determining the complexity coefficient of the first target language text according to the text length and the semantic association degree of the first target language text;
if the complexity coefficient of the first target language text is greater than or equal to the complexity coefficient threshold value, popping up a prompt dialog box to prompt the user to reenter the translation sentence;
and if the complexity coefficient of the first target language text is smaller than the complexity coefficient threshold, selecting a target offline translation data packet to perform offline translation according to the text length and the semantic association degree of the first target language text.
In one possible implementation manner, the determining the complexity coefficient of the first target language text according to the text length and the semantic association degree of the first target language text includes:
And determining the complexity coefficient of the first target language text according to a pre-trained target relation mapping table, wherein the target relation mapping table is a relation mapping table of text length, semantic association degree and text complexity coefficient.
In a second aspect, an embodiment of the present application further provides a multilingual offline intelligent translation system, including:
a memory for storing program code; and
A processor for invoking the program code to perform the method according to the first aspect.
Compared with the prior art, the multi-language offline intelligent translation system and the multi-language offline intelligent translation method provided by the embodiment of the application have the advantages that when translation aging information is slow translation, the text length and the semantic association degree of a first target language text are determined according to the first target language text information; then selecting a target offline translation data packet to perform offline translation according to the text length and the semantic association degree of the first target language text; when the first offline translation data packet is adopted for offline translation to obtain an initial translation text, a pre-trained offline correction model is adopted for offline correction of the initial translation text to obtain a second target language translation text; when the first offline translation data packet is adopted for offline translation to obtain a primary translation text, spacer insertion operation is carried out on the primary translation text to obtain a translation text of a second target language, so that the translation accuracy is improved; when the translation aging information is fast translation, the third offline translation data packet is adopted to directly translate the first target language text into the second target language translation text, and text processing programs such as interval insertion and error correction after translation are not needed, so that the translation accuracy is ensured, and the translation aging is improved. Therefore, the user can select translation timeliness according to different translation requirements, and therefore requirements of the user on translation instantaneity and translation accuracy under different states are met.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the structures shown in these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of an interface of an intelligent offline translation device according to some embodiments of the present application;
FIG. 2 is a flow chart of a multi-language offline intelligent translation method according to some embodiments of the present application;
FIG. 3 is a flowchart illustrating a step S200 of a multi-language offline intelligent translation method according to some embodiments of the present application;
FIG. 4 is a schematic diagram of a hardware architecture of a multilingual offline intelligent translation system according to some embodiments of the present application.
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
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that all directional indicators (such as up, down, left, right, front, and rear … …) in the embodiments of the present invention are merely used to explain the relative positional relationship, movement, etc. between the components in a particular posture (as shown in the drawings), and if the particular posture is changed, the directional indicator is changed accordingly.
Furthermore, the description of "first," "second," etc. in this disclosure is for descriptive purposes only and is not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In addition, "and/or" throughout this document includes three schemes, taking a and/or B as an example, including a technical scheme, a technical scheme B, and a technical scheme that both a and B satisfy; in addition, the technical solutions of the embodiments may be combined with each other, but it is necessary to base that the technical solutions can be realized by those skilled in the art, and when the technical solutions are contradictory or cannot be realized, the combination of the technical solutions should be considered to be absent and not within the scope of protection claimed in the present invention.
The offline translation device is convenient to carry and is not influenced by a network, so that the offline translation device can be used at any time and any place, but in the related technology, the offline translation device adopts the offline translation data packet trained in advance to carry out accurate translation of sentences, and the offline translation data packet cannot adapt to various types of sentences, so that the translation efficiency is lower under specific conditions, and the requirements of users on translation instantaneity and translation accuracy under different states cannot be met.
Aiming at the problems, the application provides a multi-language offline intelligent translation method which is applied to intelligent offline translation equipment, wherein the intelligent offline translation equipment is provided with three translation data packets, (1) a first translation data packet is used for translating simple texts such as short sentences or phrases, and the like, and the translation data packet is provided with a spacer insertion algorithm which can be used for inserting spacers; (2) The second translation data packet is used for translating long sentences and other complex texts, and the translation data packet error correction algorithm can correct translation errors; (3) And the third translation data packet is used for translating simple texts such as short sentences or phrases and the like, and does not carry any subsequent processing algorithm.
It can be understood that, because the first translation data packet and the second translation data packet need to be subjected to corresponding text processing after translation, the obtained translation text has higher accuracy, but poorer timeliness, and the corresponding translation aging information is slow translation; and the third translation data packet does not carry out any text processing after translation, so that the translation timeliness is good, and the corresponding translation timeliness information is fast translation.
It should be noted that, as shown in fig. 1, the intelligent offline translation device of the present application includes a gear switch with a translation speed, when the gear switch is shifted to a "fast" gear, at this time, the translation aging information is indicated as fast translation, the gear switch of the intelligent offline translation device is shifted to "slow", and the display interface includes an input display area 200, a translation display area 300, and a virtual keyboard area 400, where the virtual keyboard area 400 is used for inputting characters to be translated, the input display area 200 is used for displaying the text to be translated, and the translation display area 300 is used for displaying the translated text obtained after translation.
1-3, The specific steps of the multilingual offline intelligent translation method will be primarily described below, with the understanding that, although a logical sequence is illustrated in the flowchart, in some cases, the steps illustrated or described may be performed in a different order than that illustrated herein. Referring to fig. 2, the method comprises the steps of:
S100, acquiring an offline translation instruction, wherein the offline translation instruction carries first target language text information to be translated and translation aging information, and the translation aging information comprises fast translation and slow translation;
As shown in fig. 1, when the gear switch of the intelligent offline translation device is shifted to the "fast" gear, the translation aging information is represented as fast translation at this time, and when the gear switch of the intelligent offline translation device is shifted to the "slow" gear, the translation aging information is represented as slow translation at this time.
When a user inputs a first target language text to be translated on the intelligent offline translation equipment and clicks 'translation', a processor of the intelligent offline translation equipment automatically acquires an offline translation instruction, and the offline translation instruction carries the first target language text information to be translated and translation aging information at the moment, so that corresponding aging and translation of the corresponding text are carried out according to related information.
S200, determining the text length and the semantic association degree of the first target language text according to the first target language text information when the translation aging information is slow translation;
when a user dials a gear switch of the intelligent offline translation device to a slow gear, the user is stated to take the accuracy of translation into consideration preferentially, the translation speed is not considered, the translation aging information is slow translation, and at the moment, the text length and the semantic association degree of the first target language text are required to be determined according to the first target language text information, so that a proper translation data packet is selected for translation according to the text length and the semantic association degree of the first target language text.
In one embodiment, the step S200 of determining the text length and the semantic association degree of the first target language text according to the first target language text information includes:
S210, calling a text length function to perform length recognition on the first target language text information to obtain the text length of the first target language text;
S220, inputting the first target language text information into an offline semantic analysis model to obtain the semantic relevance of the first target language text.
Specifically, a text length function may be called to perform text length recognition on the first target language text information to obtain the text length of the first target language text, semantic relevance analysis may be performed through a semantic analysis model to obtain the semantic relevance of the first target language text, for example, semantic relevance feature recognition may be performed by using a pre-trained deep learning model to perform relevance prediction to obtain the text semantic relevance of the first target language text, and the deep learning model may be obtained by performing parameter configuration after a large number of data tests.
S300, selecting a target offline translation data packet to perform offline translation according to the text length and the semantic relevance of the first target language text, wherein the target offline translation data packet comprises a first offline translation data packet and a second offline translation data packet;
Because the processor of the intelligent offline translation device processes the "comfort" range for the complexity of the text, the processor can also understand that when the complexity of the text is in a certain range, the processor can handle the text for translation, and when the complexity of the text exceeds a certain range, the processor cannot translate, so that a user is required to manually split the text to be translated, and the text to be translated is simplified.
Therefore, in one embodiment, before selecting the target offline translation data packet for offline translation according to the text length and the semantic relevance of the first target language text in step S300, the method further comprises:
Determining the complexity coefficient of the first target language text according to the text length and the semantic association degree of the first target language text;
if the complexity coefficient of the first target language text is greater than or equal to the complexity coefficient threshold value, popping up a prompt dialog box to prompt the user to reenter the translation sentence;
and if the complexity coefficient of the first target language text is smaller than the complexity coefficient threshold, selecting a target offline translation data packet to perform offline translation according to the text length and the semantic association degree of the first target language text.
Specifically, firstly, obtaining a complexity coefficient of a first target language text according to the text length and semantic relevance of the first target language text, after obtaining the complexity coefficient of the first target language text, when judging that the complexity coefficient of the first target language text is greater than or equal to a complexity coefficient threshold value, indicating that the first target language text is too complex, and exceeding the translation processing range of a processor, at the moment, popping up a prompt dialog box to prompt a user to reenter a translation sentence (reentry after simplification); when the complexity coefficient of the first target language text is judged to be smaller than the complexity coefficient threshold value, the explanation processor can process the text to be translated, and then the target offline translation data packet is selected to be translated offline according to the text length and the semantic relevance of the first target language text.
It should be noted that, when determining the complexity coefficient of the first target language text, the complexity coefficient of the first target language text may be determined according to a pre-trained target relationship mapping table, where the target relationship mapping table is a relationship mapping table of text length, semantic association degree and text complexity coefficient.
In the case that the translation aging information is slow translation and the processor can process the translation aging information, the translation aging information and the translation accuracy can be further divided according to the text length and the semantic association degree, for example, when a first translation data packet is selected, since only simple text processing is needed after translation is performed to obtain a translation text, the time spent on selecting the first translation data packet for translation is relatively less, and therefore, the timeliness is better than that of selecting a second translation data packet for translation; when the second translation data packet is selected, since text processing such as error detection and correction is required after translation is performed to obtain a translation text, the accuracy is better than that when the first translation data packet is selected for translation.
In one embodiment, the step S300 of selecting a target offline translation data packet for offline translation according to the text length and the semantic relevance of the first target language text includes:
S310, determining that the text length of the first target language text is smaller than or equal to a first preset length value, or that the text length of the first target language text is larger than the first preset length value and smaller than a second preset length value, and that the semantic association is smaller than or equal to a preset association, and selecting the first offline translation data packet to perform offline translation;
S320, determining that the text length of the first target language text is larger than a second preset length value, or that the text length of the first target language text is larger than a first preset length value and smaller than the second preset length value, and that the semantic association degree is larger than or equal to the preset association degree, and selecting the second offline translation data packet to perform offline translation, wherein the first preset length value is smaller than the second preset length value.
Specifically, when the text length of the first target language text is smaller than or equal to a first preset length value, or when the text length of the first target language text is larger than the first preset length value and smaller than a second preset length value, but the semantic association degree is smaller than the preset association degree, the complexity of the first target language text is lower, the probability of translation errors is smaller, at this time, a first offline translation data packet is adopted for offline translation, after the first offline translation data packet is adopted for offline translation to obtain an initial translation text, feature recognition can be performed on the initial translation text according to a preset recognition rule to obtain a segmentation boundary, then spacer insertion operation is performed before or after the segmentation boundary to obtain a second target language translation text, for example, according to a name recognition rule, namely different names are recognized, and spacer insertion operation is performed between names to obtain the translation text accurately; therefore, the accuracy of text translation is improved, so that the user can conveniently review, and the recognition error of the user review is reduced.
When the text length of the first target language text is greater than a second preset length value, or when the text length of the first target language text is greater than the first preset length value and less than the second preset length value, and the semantic association degree is greater than or equal to the preset association degree, the complexity of the first target language text is higher, the probability of translating mistakes is also higher, at this time, the second offline translation data packet is adopted for offline translation, and after the second offline translation data packet is adopted for offline translation to obtain the primary translation text, the translation text can be further processed to further improve the accuracy of text translation, for example, the primary translation text is automatically corrected in error, so that the accuracy of translation is further improved.
S400, performing offline translation in response to the second offline translation data packet to obtain an initial translation text;
after the system detects that the first translation text is obtained by offline translation by adopting the second offline translation data packet, further processing is needed to be carried out on the first translation text so as to improve the accuracy of translation.
S500, performing offline correction on the initial translation text by adopting a pre-trained offline correction model to obtain a second target language translation text, wherein the offline correction model is an error correction model obtained by training according to historical manual error correction big data.
In one embodiment, the offline correction model includes a first error correction model, the first error correction model corrects the first language element error, and the step S500 includes performing offline correction on the initial translation text by using a pre-trained offline correction model to obtain a second target language translation text, including:
s510, detecting whether the primary translation text has a first language element error or not;
s520, if the initial translation text has the first language element error, carrying out first language element error correction on the initial translation text by adopting the first error correction model to obtain a second target language translation text.
Specifically, the first error correction model is an error correction model obtained by training according to historical manual error correction big data, and an exemplary first language element error correction model is a "scholarly language position correction" model, for example, when aiming at English-Chinese translation, there is often a scholarly language position grammar error, the first error correction model is obtained by manually feeding back and correcting and then carrying out data training, if a processor receives a translation text of "we don't need to go to school today If the storm is coming", the translation is carried out to obtain a translation text of "we need not learn today, if storm comes," the translation text is detected and found "we need not learn today, if storm comes," the Chinese grammar is not met, and the translation is inaccurate, and at this time, the first language element error correction model is obtained by training through user historical feedback data, so that "if storm comes," we need not learn today "is obtained, thereby obtaining an accurate second target language translation text, and facilitating understanding of the translation text by a user.
In other embodiments, the second error correction model may be obtained by performing artificial intelligent machine training according to the user history feedback data, for example, the second error correction model is a "subject position correction" model, and similarly, when detecting that the subject position of the primary translated text is wrong, subject position correction may be performed through the second error correction model to obtain an accurate second target language translated text, so as to further facilitate understanding of the translated text by the user.
In another embodiment, after obtaining the offline translation instruction, the method further comprises:
And S600, when the translation aging information is fast translation, selecting a third offline translation data packet for offline translation, wherein the third offline translation data packet directly translates the first target language text into a second target language translation text.
Specifically, when the user toggles the gear switch of the intelligent offline translation device to a 'fast' gear, the user is stated to take the timeliness of translation into priority, the timeliness of translation is the fast translation, at the moment, a third offline translation data packet is selected for offline translation, the third offline translation data packet directly translates the first target language text into the second target language translation text, text processing programs such as interval insertion and error correction after translation are not needed, and the timeliness of translation is greatly improved.
Based on the above, according to the multi-language offline intelligent translation method, when translation aging information is slow translation, firstly, the text length and semantic association degree of a first target language text are determined according to the first target language text information; then selecting a target offline translation data packet to perform offline translation according to the text length and the semantic association degree of the first target language text; when the first offline translation data packet is adopted for offline translation to obtain a primary translation text, spacer insertion operation is carried out on the primary translation text to obtain a translation text of a second target language, so that the translation accuracy is improved; when the first offline translation text is obtained by offline translation by adopting the second offline translation data packet, offline correction is carried out on the first translation text by adopting a pre-trained offline correction model to obtain a second target language translation text, so that the translation accuracy is further improved; when the translation aging information is fast translation, the third offline translation data packet is adopted to directly translate the first target language text into the second target language translation text, and text processing programs such as interval insertion and error correction after translation are not needed, so that the translation accuracy is ensured, and the translation aging is improved. Therefore, the user can select translation timeliness according to different translation requirements, and therefore requirements of the user on translation instantaneity and translation accuracy under different states are met.
Referring to fig. 4, fig. 4 is a schematic hardware structure diagram of a multi-language offline intelligent translation system according to some embodiments of the present application, where the multi-language offline intelligent translation system includes a memory 110 and a processor 120, and the memory 110 is used for storing program codes, and the processor 120 is used for calling the program codes to execute the method described above.
Wherein the processor 120 is configured to provide computing and control capabilities to control the multi-language offline intelligent translation system to perform corresponding tasks, for example, to control the multi-language offline intelligent translation system to perform the multi-language offline intelligent translation method in any of the method embodiments described above, the method comprising: acquiring an offline translation instruction, wherein the offline translation instruction carries first target language text information to be translated and translation aging information, and the translation aging information comprises fast translation and slow translation; when the translation aging information is slow translation, determining the text length and the semantic association degree of the first target language text according to the first target language text information; selecting a target offline translation data packet to perform offline translation according to the text length and the semantic relevance of the first target language text, wherein the target offline translation data packet comprises a first offline translation data packet and a second offline translation data packet; performing offline translation in response to the second offline translation data packet to obtain an initial translation text; and carrying out offline correction on the initial translation text by adopting a pre-trained offline correction model to obtain a second target language translation text, wherein the offline correction model is an error correction model obtained by training according to historical manual error correction big data.
Processor 120 may be a general-purpose processor including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), a hardware chip, or any combination thereof; it may also be a digital signal processor (DIGITAL SIGNAL Processing, DSP), application SPECIFIC INTEGRATED Circuit (ASIC), programmable logic device (programmable logic device, PLD), or a combination thereof. The PLD may be a complex programmable logic device (complex programmable logic device, CPLD), a field-programmable gate array (FPGA) GATE ARRAY, generic array logic (GENERIC ARRAY logic, GAL), or any combination thereof.
The memory 110 is used as a non-transitory computer readable storage medium for storing non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the multi-language offline intelligent translation method in the embodiment of the present application. The processor 120 may implement the multilingual offline intelligent translation method in any of the method embodiments described above by running non-transitory software programs, instructions, and modules stored in the memory 110.
In particular, memory 110 may include Volatile Memory (VM), such as random access memory (random access memory, RAM); memory 110 may also include non-volatile memory (NVM), such as read-only memory (ROM), flash memory (flash memory), hard disk (HARD DISK DRIVE, HDD) or solid-state disk (solid-state drive-STATE DRIVE, SSD) or other non-transitory solid state storage device; memory 110 may also include a combination of the types of memory described above.
In summary, the multi-language offline intelligent translation system of the present application adopts the technical scheme of any one of the embodiments of the multi-language offline intelligent translation method, so that the system at least has the beneficial effects brought by the technical scheme of the embodiments, and will not be described in detail herein.
Embodiments of the present application also provide a computer readable storage medium, such as a memory, including program code executable by a processor to perform the multilingual offline intelligent translation method of the embodiments described above. For example, the computer readable storage medium may be Read-Only Memory (ROM), random-access Memory (Random Access Memory, RAM), compact disc Read-Only Memory (CDROM), magnetic tape, floppy disk, optical data storage device, and the like.
Embodiments of the present application also provide a computer program product comprising one or more program codes stored in a computer-readable storage medium. The program code is read from the computer readable storage medium by a processor of the multi-language offline intelligent translation system, and the processor executes the program code to complete the steps of the multi-language offline intelligent translation method provided in the above embodiments.
It will be appreciated by those of ordinary skill in the art that all or part of the steps of implementing the above embodiments may be implemented by hardware, or may be implemented by program code related hardware, where the program may be stored in a computer readable storage medium, where the storage medium may be a read only memory, a magnetic disk or optical disk, etc.
It should be noted that the above-described apparatus embodiments are merely illustrative, and the units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
From the above description of embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus a general purpose hardware platform, or may be implemented by hardware. Those skilled in the art will appreciate that all or part of the processes implementing the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and where the program may include processes implementing the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random-access Memory (Random Access Memory, RAM), or the like.
The foregoing description is only of the preferred embodiments of the present invention and is not intended to limit the scope of the invention, and all equivalent structural changes made by the description of the present invention and the accompanying drawings or direct/indirect application in other related technical fields are included in the scope of the invention.
Claims (10)
1. A multi-language offline intelligent translation method applied to intelligent offline translation equipment, characterized in that the method comprises the following steps:
acquiring an offline translation instruction, wherein the offline translation instruction carries first target language text information to be translated and translation aging information, and the translation aging information comprises fast translation and slow translation;
When the translation aging information is slow translation, determining the text length and the semantic association degree of the first target language text according to the first target language text information;
Selecting a target offline translation data packet to perform offline translation according to the text length and the semantic relevance of the first target language text, wherein the target offline translation data packet comprises a first offline translation data packet and a second offline translation data packet, the first offline translation data packet is used for translating simple texts, and the second offline translation data packet is used for translating complex texts;
performing offline translation in response to the second offline translation data packet to obtain an initial translation text;
And carrying out offline correction on the initial translation text by adopting a pre-trained offline correction model to obtain a second target language translation text, wherein the offline correction model is an error correction model obtained by training according to historical manual error correction big data.
2. The multi-language offline intelligent translation method according to claim 1, further comprising, after selecting a target offline translation data package for offline translation according to the text length and the semantic relevance of the first target language text:
performing offline translation in response to the first offline translation data packet to obtain an initial translation text;
Performing feature recognition on the initial translation text according to a preset recognition rule to obtain a segmentation boundary, wherein the preset recognition rule is a recognition rule obtained by training in advance and is used for performing feature recognition of the corresponding recognition rule on the initial translation text;
and performing a spacer insertion operation before or after the segmentation boundary to obtain the second target language translation text.
3. The multi-language offline intelligent translation method according to claim 1, further comprising, after said obtaining the offline translation instruction:
And when the translation aging information is fast translation, selecting a third offline translation data packet for offline translation, wherein the third offline translation data packet directly translates the first target language text into the second target language translation text.
4. The method for intelligent offline translation of multiple languages according to claim 1, wherein determining the text length and semantic relevance of the first target language text according to the first target language text information comprises:
Calling a text length function to perform length recognition on the first target language text information to obtain the text length of the first target language text;
Inputting the first target language text information into an offline semantic analysis model to obtain the semantic relevance of the first target language text, wherein the offline semantic analysis model is a pre-trained semantic analysis algorithm model and is used for analyzing the text semantic relevance of the first target language text.
5. The method for intelligent offline translation of multiple languages according to claim 1, wherein selecting a target offline translation data packet for offline translation according to the text length and semantic relevance of the first target language text comprises:
determining that the text length of the first target language text is smaller than or equal to a first preset length value, and selecting the first offline translation data packet for offline translation; or determining that the text length of the first target language text is larger than a first preset length value and smaller than a second preset length value, and the semantic association is smaller than the preset association, and selecting the first offline translation data packet to perform offline translation;
determining that the text length of the first target language text is larger than a second preset length value, and selecting the second offline translation data packet for offline translation; or determining that the text length of the first target language text is larger than a first preset length value and smaller than a second preset length value, and the semantic association is larger than or equal to the preset association, and selecting the second offline translation data packet to perform offline translation;
wherein the first preset length value is smaller than the second preset length value.
6. The multi-language offline intelligent translation method of claim 1, wherein the offline correction model comprises a first error correction model, the first error correction model corrects for a first language element error, the offline correction of the initial translation text using a pre-trained offline correction model to obtain a second target language translation text comprises:
detecting whether the primary translation text has a first language element error;
And if the initial translation text has the first language element error, carrying out first language element error correction on the initial translation text by adopting the first error correction model to obtain a second target language translation text.
7. The multi-language offline intelligent translation method of claim 1, wherein the offline correction model comprises a second error correction model, the second error correction model corrects for a second language element error, the offline correction of the initial translation text using a pre-trained offline correction model to obtain a second target language translation text comprises:
Detecting whether the primary translation text has a second language element error;
And if the initial translation text has the second language element error, carrying out second language element error correction on the initial translation text by adopting the second error correction model to obtain a second target language translation text.
8. The multi-language offline intelligent translation method according to claim 1, wherein before the selecting a target offline translation data packet for offline translation according to the text length and the semantic relevance of the first target language text, the method further comprises:
Determining the complexity coefficient of the first target language text according to the text length and the semantic association degree of the first target language text;
if the complexity coefficient of the first target language text is greater than or equal to the complexity coefficient threshold value, popping up a prompt dialog box to prompt the user to reenter the translation sentence;
and if the complexity coefficient of the first target language text is smaller than the complexity coefficient threshold, selecting a target offline translation data packet to perform offline translation according to the text length and the semantic association degree of the first target language text.
9. The method of claim 8, wherein determining the complexity factor of the first target language text based on the text length and the semantic relevance of the first target language text comprises:
And determining the complexity coefficient of the first target language text according to a pre-trained target relation mapping table, wherein the target relation mapping table is a relation mapping table of text length, semantic association degree and text complexity coefficient.
10. A multi-language offline intelligent translation system, comprising:
a memory for storing program code; and
A processor for invoking the program code to perform the method of any of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410788983.3A CN118364829B (en) | 2024-06-19 | 2024-06-19 | Multi-language offline intelligent translation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410788983.3A CN118364829B (en) | 2024-06-19 | 2024-06-19 | Multi-language offline intelligent translation system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118364829A CN118364829A (en) | 2024-07-19 |
| CN118364829B true CN118364829B (en) | 2024-08-27 |
Family
ID=91885032
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410788983.3A Active CN118364829B (en) | 2024-06-19 | 2024-06-19 | Multi-language offline intelligent translation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118364829B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118675529B (en) * | 2024-05-31 | 2025-03-11 | 深圳市新画面软件通讯有限公司 | Data processing method and system for call voice translation |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107590137A (en) * | 2017-08-07 | 2018-01-16 | 北京小米移动软件有限公司 | Interpretation method, device and computer-readable recording medium |
| CN113076760A (en) * | 2020-01-03 | 2021-07-06 | 阿里巴巴集团控股有限公司 | Translation method, commodity retrieval method, translation device, commodity retrieval device, electronic equipment and computer storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9430465B2 (en) * | 2013-05-13 | 2016-08-30 | Facebook, Inc. | Hybrid, offline/online speech translation system |
| US11036940B2 (en) * | 2018-08-30 | 2021-06-15 | Mmt Srl | Translation system and method |
| CN109783824B (en) * | 2018-12-17 | 2023-04-18 | 北京百度网讯科技有限公司 | Translation method, device and storage medium based on translation model |
-
2024
- 2024-06-19 CN CN202410788983.3A patent/CN118364829B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107590137A (en) * | 2017-08-07 | 2018-01-16 | 北京小米移动软件有限公司 | Interpretation method, device and computer-readable recording medium |
| CN113076760A (en) * | 2020-01-03 | 2021-07-06 | 阿里巴巴集团控股有限公司 | Translation method, commodity retrieval method, translation device, commodity retrieval device, electronic equipment and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118364829A (en) | 2024-07-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7267365B2 (en) | Text error correction method, apparatus, electronic equipment and storage medium | |
| US8990066B2 (en) | Resolving out-of-vocabulary words during machine translation | |
| EP0031493A1 (en) | Alpha content match prescan method and system for automatic spelling error correction | |
| US20200184953A1 (en) | Method, device, and storage medium for correcting error in speech recognition result | |
| KR101495240B1 (en) | Method and system for statistical context-sensitive spelling correction using confusion set | |
| CN108090043B (en) | Error correction report processing method and device based on artificial intelligence and readable medium | |
| CN118364829B (en) | Multi-language offline intelligent translation system and method | |
| CN111753530B (en) | Statement processing method, device, equipment and medium | |
| TWI567569B (en) | Natural language processing systems, natural language processing methods, and natural language processing programs | |
| JP6778655B2 (en) | Word concatenation discriminative model learning device, word concatenation detection device, method, and program | |
| US11954439B2 (en) | Data labeling method and device, and storage medium | |
| WO2018153316A1 (en) | Method and apparatus for obtaining text extraction model | |
| CN110929514A (en) | Text proofreading method, apparatus, computer-readable storage medium and electronic device | |
| CN112434537B (en) | Translation text consistency verification method, computing device and storage medium | |
| CN116806338A (en) | Determining and utilizing auxiliary language proficiency metrics | |
| CN116579327A (en) | Text error correction model training method, text error correction method, device and storage medium | |
| CN112784611A (en) | Data processing method, device and computer storage medium | |
| CN109614624B (en) | English sentence recognition method and electronic equipment | |
| CN110633474B (en) | Mathematical formula identification method, device, equipment and readable storage medium | |
| CN118072323A (en) | Recognition result error correction method, device, computer equipment and medium | |
| CN112380879A (en) | Intelligent translation method and device, computer equipment and storage medium | |
| CN111639160A (en) | Domain identification method, interaction method, electronic device and storage medium | |
| EP3970154A1 (en) | Correcting an examination report | |
| KR101626386B1 (en) | Method and apparatus for automatic post-editing based on factored language model | |
| CN112905025B (en) | Information processing method, electronic device and readable 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 |