Language translation method and device for business data, computer equipment, readable storage medium and program product

Abstract

The invention relates to a language translation method and device for business data, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps of obtaining to-be-translated business data and a target language; obtaining an initial translation prompt word corresponding to the target language, and inputting the business data and the initial translation prompt word into a translation model for language translation to obtain initial translation data corresponding to the business data; obtaining an inverse translation cue word, and inputting the business data, the initial translation data and the inverse translation cue word into a translation model for translation analysis to obtain translation suggestion data; and obtaining an improved translation cue word, and inputting the business data, the initial translation data, the translation suggestion data and the improved translation cue word into a translation model for improved translation to obtain target translation data corresponding to the business data. By adopting the method, the language translation accuracy of the business data can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of language translation, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for translating business data into different languages. Background Art

[0002] With the development of the internet, international business has also grown. Existing ERP (Enterprise Resource Planning) systems are real-time management information systems that integrate business data from departments located in different countries, enabling cross-border business processing. However, to overcome language barriers during cross-border business processing, business data in minority languages ​​needs to be translated. For example, when Chinese employees conduct cross-border business with Vietnamese employees, business data transmitted by the Chinese employees needs to be translated into Vietnamese, which the Vietnamese employees can understand, before the corresponding business processing can be carried out.

[0003] Traditional language translation involves loading a translation package commonly used for a minority language. However, using translation packages for translation is not suitable for complex contexts and may result in inaccurate translation of specialized vocabulary, leading to low language translation accuracy for business data. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, apparatus, computer device, computer-readable storage medium and computer program product for translating business data into different languages, which can improve the accuracy of the translation of business data into different languages, in order to address the above technical problems.

[0005] In a first aspect, the present application provides a method for translating business data into different languages, comprising:

[0006] Obtain the business data and target language to be translated;

[0007] Obtain the initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into the translation model for language translation, and obtain the initial translation data corresponding to the business data;

[0008] Obtaining reflection translation prompts, inputting business data, initial translation data, and reflection translation prompts into the translation model for translation analysis to obtain translation suggestion data;

[0009] Obtain improved translation prompt words, input the business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0010] In a second aspect, the present application further provides a language translation device for business data, comprising:

[0011] Data acquisition module, used to obtain business data to be translated and the target language;

[0012] The initial translation module is used to obtain the initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into the translation model for language translation, and obtain the initial translation data corresponding to the business data;

[0013] The translation suggestion module is used to obtain reflection translation prompt words, input business data, initial translation data and reflection translation prompt words into the translation model for translation analysis, and obtain translation suggestion data;

[0014] The improved translation module is used to obtain improved translation prompt words, input business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0015] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0016] Obtain the business data and target language to be translated;

[0017] Obtain the initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into the translation model for language translation, and obtain the initial translation data corresponding to the business data;

[0018] Obtaining reflection translation prompts, inputting business data, initial translation data, and reflection translation prompts into the translation model for translation analysis to obtain translation suggestion data;

[0019] Obtain improved translation prompt words, input the business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0020] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0021] Obtain the business data and target language to be translated;

[0022] Obtain the initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into the translation model for language translation, and obtain the initial translation data corresponding to the business data;

[0023] Obtaining reflection translation prompts, inputting business data, initial translation data, and reflection translation prompts into the translation model for translation analysis to obtain translation suggestion data;

[0024] Obtain improved translation prompt words, input the business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0025] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0026] Obtain the business data and target language to be translated;

[0027] Obtain the initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into the translation model for language translation, and obtain the initial translation data corresponding to the business data;

[0028] Obtaining reflection translation prompts, inputting business data, initial translation data, and reflection translation prompts into the translation model for translation analysis to obtain translation suggestion data;

[0029] Obtain improved translation prompt words, input the business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0030] The above-mentioned business data language translation method, apparatus, computer device, computer-readable storage medium, and computer program product perform language translation by inputting the business data to be translated and initial translation prompt words into a translation model to obtain initial translation data, then obtaining reflective translation prompt words, and inputting the business data, initial translation data, and reflective translation prompt words into the translation model for translation analysis. This allows the translation model to perform translation accuracy improvement analysis on the business data and initial translation data based on the reflective translation prompt words, and output translation suggestion data that improves translation errors in the initial translation data. The initial translation data, translation suggestion data, and improved translation prompt words are then input into the translation model for improved translation. This allows the translation model to improve the translation method of translating the business data into the initial translation data based on the improved translation prompt words and according to the translation suggestion data. In other words, this optimizes the translation method of the translation model for the business data, achieving effective improvement and personalized translation of the business data by the translation model, and improving the accuracy of the target translation data. In summary, by reflecting on the translation prompt words, the translation model can self-check the initial translation data and output translation suggestion data. By improving the translation prompt words and translation suggestion data, the translation model can be optimized, ensuring that the translation model accurately translates the business data, thereby improving the translation accuracy of the business data. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0032] Figure 1 A diagram illustrating an application environment of a method for translating business data into different languages ​​according to an embodiment;

[0033] Figure 2 A flowchart of a method for translating business data into different languages ​​in one embodiment is shown;

[0034] Figure 3 A flowchart of language translation steps for business data in one embodiment;

[0035] Figure 4 A schematic diagram of a business scenario for approving an application in one embodiment;

[0036] Figure 5 A schematic diagram of abnormal translation feedback in one embodiment;

[0037] Figure 6 A schematic diagram of the translation process of business data in one embodiment;

[0038] Figure 7 A schematic diagram of editing translation prompt words in one embodiment;

[0039] Figure 8 is a schematic diagram of a translation process based on reflection in one embodiment;

[0040] Figure 9 A structural block diagram of a device for translating business data into different languages ​​according to an embodiment;

[0041] Figure 10 is a diagram of the internal structure of a computer device in one embodiment;

[0042] Figure 11 FIG. 4 is a diagram showing the internal structure of a computer device in another embodiment. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0044] The language translation method of the business data provided in the embodiment of the present application can be applied to Figure 1In the application environment shown, the terminal 102 communicates with the server 104 via a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on a cloud or other network server. The terminal 102 obtains the business data to be translated and the target language; the terminal 102 obtains the initial translation prompt words corresponding to the target language, inputs the business data and the initial translation prompt words into the translation model for language translation, and obtains the initial translation data corresponding to the business data; the terminal 102 obtains the reflective translation prompt words, inputs the business data, initial translation data, and reflective translation prompt words into the translation model for translation analysis, and obtains translation suggestion data; the terminal 102 obtains the improved translation prompt words, inputs the business data, initial translation data, translation suggestion data, and improved translation prompt words into the translation model for improved translation, and obtains the target translation data corresponding to the business data. The translation model can be deployed on the server 104, and the terminal inputs and reads data from the translation model deployed on the server 104 via the network. The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablet computers, etc. The server 104 may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0045] In an exemplary embodiment, Figure 2 As shown, a method for translating business data into different languages ​​is provided. Figure 1 The following steps are used as an example to illustrate the terminal in the figure:

[0046] Step 202: Obtain the business data to be translated and the target language.

[0047] Step 204: Obtain initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into a translation model for language translation, and obtain initial translation data corresponding to the business data.

[0048] The target language refers to the language to be translated. Translation prompts are textual information used to guide the translation model in outputting the desired translation content. Initial translation prompts are pre-set translation prompts used for the initial translation of business data. The translation model is an artificial intelligence model used for language translation. Initial translation data refers to the translation data generated after the translation model translates the business data to be translated according to the initial translation prompts.

[0049] Exemplarily, the terminal acquires business data to be translated, which may be uploaded by a user or input in real time via an interactive interface. The business data may be text data, table data, image data, etc. generated during business processing. In response to the user's request for translation of the business data, the terminal acquires the target language, which may be the target language included in the translation request or based on configuration requirements. The target languages ​​may be one or more.

[0050] When the terminal detects that the business data is being translated for the first time, it obtains an initial translation prompt for the target language. This can be done by obtaining an initial prompt template and generating an initial translation prompt based on the target language and the initial prompt template. For example, if the target language is English, the initial translation prompt might be "Translate business data into English." The terminal then invokes a translation model, inputs the business data and the initial translation prompt into the translation model, and performs language translation to obtain the initial translation data corresponding to the business data.

[0051] In one exemplary embodiment, when the business data includes image data, image source information corresponding to the image data is obtained, and based on the image source information, whether to translate the image data is determined. For example, if the image source information determines that the image data is image data entered for storage via a specified path, the field content in the image data is translated. If the image source information determines that the image data is image data entered during business communication, such as image data sent by a user to another user via the business system's communication function, the image data is not processed temporarily, thereby conserving computing resources for the translation model.

[0052] Furthermore, when it is detected that the image data is image data input during business communication, the image data also carries a translation control and is displayed on the terminal page of other users, so that the terminal of other users responds to the triggering operation of the translation control and calls the translation model to translate the field content in the image data, which can improve the translation flexibility of the image data.

[0053] In one exemplary embodiment, when translating image data, textual information within the image data can be identified and used as business data. The business data and initial translation prompts are then input into a translation model to obtain initial translation data corresponding to the image data. Alternatively, the image data can be used as business data, and image translation prompts are obtained when the business data is image data. The image translation prompts include recognition and translation requirement information, such as "identify text within the image data and translate the identified text into English." The image data and image translation prompts are then input into a translation model to obtain initial translation data corresponding to the image data.

[0054] Step 206 , obtaining reflection translation prompt words, inputting the business data, initial translation data and reflection translation prompt words into the translation model for translation analysis, and obtaining translation suggestion data.

[0055] Reflective translation prompts instruct the translation model to perform a translation accuracy analysis on the initial translation data. This analysis involves the translation model detecting translation deviations in the initial translation data using a pre-trained knowledge base, such as incorrect and ambiguous translations. Translation suggestion data refers to the translation model generating suggestions for improvement based on the translation deviations between the business data and the initial translation data, using the reflective translation prompts.

[0056] Exemplarily, the terminal obtains the initial translation data returned by the translation model and obtains a reflective translation prompt. This may involve obtaining a business scenario corresponding to the business data and generating a reflective translation prompt based on the business scenario. It is understood that the reflective translation prompt includes the business scenario corresponding to the business data and is used to instruct the translation model to call a knowledge base corresponding to the business scenario for translation. The knowledge base corresponding to the business scenario called by the translation model may be the knowledge base used during pre-training of the translation model or the knowledge base used during the current translation. For example, if the business data is a sales text and the business scenario is a sales scenario, the reflective translation prompt corresponding to the sales scenario may be, for example, "Based on the context of the sales scenario, conduct a scenario applicability analysis of the sales text and the translated version, identify translation errors, and provide improvement suggestions for the translated version." The terminal then inputs the business data, initial translation data, and reflective translation prompt into the translation model, causing the translation model to query the knowledge base corresponding to the business scenario according to the instructions of the reflective translation prompt. Based on the queried knowledge base, the model analyzes translation deviations of the initial translation data relative to the business data, such as incorrect translations and ambiguous translations, and generates translation improvement suggestions based on the translation deviations, thereby obtaining translation suggestion data.

[0057] Step 208: Obtain improved translation prompt words, input the business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0058] The improved translation prompt is a prompt that instructs the translation model to improve the translation of the initial translation data according to the translation suggestion data. The target translation data is the translation data obtained after the translation model improves the initial translation data according to the improved translation prompt.

[0059] For example, after obtaining the translation suggestion data, the translation model obtains an improved translation prompt and may also obtain format requirement text entered by the user, such as "For data of field types in a table, display the original text field and the translated text field in array form [","]". The terminal then adds the format requirement text to the improved translation prompt, and the improved translation prompt may be, for example, "Refer to the improvement suggestions in the translation suggestion data, perform an improved translation on the business data and the initial translation data, and output the translation data according to the translation format of the format requirement text". The terminal inputs the business data, initial translation data, translation suggestion data, and improved translation prompt into the translation model for improved translation, thereby obtaining target translation data corresponding to the business data.

[0060] In the above-mentioned business data language translation method, the business data to be translated and the initial translation prompt words are input into the translation model for language translation to obtain initial translation data. Then, the reflective translation prompt words are obtained. The business data, initial translation data, and reflective translation prompt words are input into the translation model for translation analysis. This allows the translation model to perform an improved translation analysis on the business data and initial translation data based on the reflective translation prompt words, and outputs translation suggestion data that improves translation errors in the initial translation data. The initial translation data, translation suggestion data, and improved translation prompt words are then input into the translation model for improved translation. This allows the translation model to improve the translation method of the business data into the initial translation data based on the improved translation prompt words according to the translation suggestion data. That is, the translation method of the translation model for the business data is optimized, achieving effective improvement and personalized translation of the business data by the translation model, and improving the accuracy of the target translation data. In summary, through the reflective translation prompt words, the translation model self-checks the initial translation data and outputs translation suggestion data. Moreover, through the improved translation prompt words and translation suggestion data, the translation model is optimized, ensuring the accurate translation of the business data by the translation model, thereby improving the translation accuracy of the business data.

[0061] In an exemplary embodiment, Figure 3 As shown, step 202, obtaining reflection translation prompt words, includes:

[0062] Step 302: Obtain the business scenario corresponding to the business data;

[0063] Step 304: Determine business translation requirement information based on the business scenario;

[0064] Step 306: Generate reflection translation prompt words based on the business scenario and business translation requirement information.

[0065] The business scenario refers to the business processing scenario when business data is generated, such as business data entry scenarios, business communication, etc. Business translation requirement information refers to reference information used to generate prompt words that meet the business scenario.

[0066] Exemplarily, after obtaining the initial translation data, the terminal identifies the business scenario corresponding to the business data. Business scenarios include data entry scenarios and business communication scenarios. The data entry scenario refers to a scenario in which business data for storage is uploaded through a designated path. Examples of business data in the data entry scenario include purchase orders, bills of materials, transaction records, and the like. The business communication scenario refers to a scenario in which communication is required during business processing. Examples of business data in the business communication scenario include text such as question and answer statements, approval applications, and approval opinions. The terminal can determine the business scenario based on the data source information of the business data. For example, the business scenario for business data obtained from a designated entry page is a data entry scenario, and the business scenario for text data obtained from user communication pages (such as instant messaging pages, approval application pages, etc.) is a business communication scenario.

[0067] The terminal determines the translation type corresponding to the business data based on the business scenario, obtains the corresponding improved keywords for the translation type, and generates business translation demand information based on the improved keywords. Generally, for business data in a data entry scenario, the determined translation type is a professional translation type, and the corresponding improved keywords for the professional translation type include "professional field", "expert", "professional terminology", etc. For another example, for business data in a business communication scenario, the determined translation type is a popular translation type, and the corresponding improved keywords for the popular translation type include "concise", "friendly attitude", "tone", etc.

[0068] After determining the translation requirements, the terminal generates reflective translation prompts based on the business scenario and the business translation requirements. For example, in the data entry scenario, the reflective translation prompt for the bill of materials for electrical equipment is "Carefully read the original and translated text of the bill of materials. From the perspective of an electrical expert, use professional terms in the electrical field to provide professional evaluation and improvement suggestions for the translation." The translation requirements corresponding to the data entry scenario instruct the translation model to call the knowledge base corresponding to the business scenario. Based on the translation method corresponding to the professional translation type and the invoked knowledge base, it provides improvement suggestions for professional terminology in the initial translation data. For another example, in the business communication scenario, the reflective translation prompt for an approval application is "Carefully read the original and translated text of the approval application. Provide improvement suggestions for the translation using a concise and friendly communication tone." The translation requirements corresponding to the business communication scenario instructs the translation model to call the knowledge base corresponding to the business scenario. Based on the translation method corresponding to the popular translation type and the invoked knowledge base, it provides improvement suggestions for politeness in the initial translation data, such as expanding or rewriting politeness in the initial translation data.

[0069] In an exemplary embodiment, Figure 4 As shown in the figure, a business scenario diagram of application approval is provided. Figure 4 In the case of a Chinese employee communicating with a Vietnamese leader about a business approval process through a business system, the language information that needs to be translated can be obtained based on the language currently used by the Chinese employee in the system and the language of the country to be communicated. For example, if the Vietnamese leader's business system language is Vietnamese and the procurement supplier is Chinese, the translation direction for this translation is Chinese -> Vietnamese, and the target language is Vietnamese. Figure 4 For example, a Chinese employee inputs an approval application in Chinese, "Please review this project as soon as possible." The business system calls the translation model to output the approval application content in Vietnamese, such as "Hello, the XXX project requires your review. Please review it as soon as possible. Thank you for your cooperation" (the approval application content after the translation model is improved) and displays it.

[0070] In this embodiment, by determining the business translation requirement information according to the business scenario of the business data, it is ensured that the reflective translation prompt words are prompt words that meet the translation requirements, thereby ensuring the translation accuracy of the translation model for the business data.

[0071] In an exemplary embodiment, step 304, determining business translation requirement information based on the business scenario, includes:

[0072] When the business scenario is a data entry scenario, the business translation requirement information is determined based on the business type and data entry scenario corresponding to the business data;

[0073] When the business scenario is a business communication scenario, the identity information of the target user to whom the business data belongs is obtained, and the business translation requirement information is determined based on the identity information and the business communication scenario.

[0074] Exemplarily, the business data may be business data uploaded by the target user through a specified path, or business data sent by the target user to other users through the communication function of the business system. When the terminal identifies that the business scenario of the business data is a data entry scenario, it determines that the translation type corresponding to the business data is a professional translation type, and obtains improved keywords corresponding to the professional translation type. Among them, the professional translation type is used to instruct the translation model to improve the translation of business data in the data entry scenario to meet the translation expectations of professional terminology translation. For example, the components in the purchase order of the component procurement business are translated into professional terms, such as [resistance] and other professional terms.

[0075] The terminal then obtains the business type corresponding to the business data. The business type is used to determine the business domain. For example, when entering a purchase order into a cross-border business, the terminal can determine that the purchase order's business type is procurement based on the purchase order name and the business module it belongs to. After determining the procurement domain, the terminal can instruct the translation model to utilize the procurement domain knowledge base for translation. Based on the business type corresponding to the business data and the refined keywords for the specialized translation type corresponding to the data entry scenario, the terminal generates business translation requirements information for the data entry scenario.

[0076] When the terminal identifies the business data as a business communication scenario, it determines the translation type corresponding to this scenario as a colloquial translation. This colloquial translation type instructs the translation model to improve the translation of text data in this business communication scenario to meet the translation expectations of politeness. The terminal then obtains the identity information of the target user to whom the business data belongs, such as the target user's rank (employee, leader, etc.). Based on the business communication scenario and the target user's identity information, the terminal determines the target user's corresponding rank relationship. For example, if the target user is an employee and is applying for approval, the corresponding rank relationship is a superior-superior relationship. Alternatively, if the target user is a leader and is providing approval feedback, the corresponding rank relationship is a subordinate-subordinate relationship. Based on the colloquial translation type and the target user's corresponding rank relationship, the terminal determines improvement keywords and generates business translation requirements information for the business communication scenario based on the determined improvement keywords. For example, if a subordinate employee requests approval from a superior, "Please review this project as soon as possible," the improved translation could be, "Hello, Leader X, project XXX requires your review. Please review it as soon as possible. Thank you for your cooperation."

[0077] In this embodiment, by determining corresponding business translation requirement information for different business scenarios, the translation flexibility and accuracy of business data in different business scenarios are guaranteed.

[0078] In an exemplary embodiment, in step 208, after the business data, initial translation data, translation suggestion data, and improved translation prompt words are input into the translation model for improved translation and target translation data corresponding to the business data is obtained, the language translation method for business data further includes:

[0079] Obtaining abnormal translation feedback corresponding to the target translation data, and generating updated translation prompt words based on the abnormal translation feedback;

[0080] The updated translation prompt word is used as the initial translation prompt word, and the process returns to the step of inputting the business data and the initial translation prompt word into the translation model for language translation, until the updated target translation data corresponding to the business data is obtained.

[0081] Abnormal translation feedback refers to feedback regarding translation deviations in the target translation data, such as incorrect, inaccurate, or untranslated translations of certain words. Update translation prompts instruct the translation model to retranslate business data based on abnormal translation feedback.

[0082] For example, the terminal receives target translation data corresponding to the business data from the translation model and displays the target translation data on the current page. Figure 5 The abnormal translation feedback diagram shown in FIG shows the problematic entry selected by the user on the current page, and displays the feedback information input box, as well as Figure 5 The input box corresponding to "My suggested terms" obtains the reference terms entered by the user through the feedback information input box.

[0083] The terminal receives abnormal translation feedback based on the problematic terms selected by the user and the reference terms entered by the user. It then generates an updated translation prompt based on the abnormal translation feedback. The updated translation prompt is then used as the initial translation prompt, and the process returns to the step of inputting the business data and the initial translation prompt into the translation model for language translation, until the updated target translation data corresponding to the business data is obtained.

[0084] In this embodiment, by obtaining abnormal translation feedback from users on target translation data and generating updated translation prompt words, the translation accuracy of the translation model can be guaranteed.

[0085] In an exemplary embodiment, Figure 6 As shown, a schematic diagram of the translation process of business data is provided. Users can implement language translation of business data through the business system. The business system can be an ERP (Enterprise Resource Planning) system, where the ERP system provides system languages ​​of corresponding languages ​​to users of different nationalities. For example, the system language of the ERP system for Chinese employees is Chinese, and the system language of the ERP system for Vietnamese employees is Vietnamese. The ERP system is connected to the AI ​​translation platform. The AI ​​translation platform can be a service platform built based on the K8S+docker cloud native architecture. The AI ​​translation platform provides a standard API interface based on token authentication to realize data interaction between the ERP end and the AI ​​translation platform. The translation model (such as Figure 6Specifically, a user enters business data into the ERP system. The ERP system determines the language of the input business data based on the system language, extracts the business data to be translated, determines the business scenario and target language corresponding to the business data, and then screens the data for sensitive information. For example, it filters out sensitive words and replaces them with unified identifiers in the content to be translated. The ERP system then calls the local knowledge base to preprocess the data to be translated. This can involve checking whether there is existing local translation data for the data to be translated in the local knowledge base. If so, it retrieves the existing local translation data. For data not found in the knowledge base, a translation model is used to translate the data. The ERP system performs interface parameter verification and interface traffic limit checks on the translation model (which can be a big data model). If the interface parameter verification passes and the interface traffic limit is not exceeded, the interface of the translation model to be called is determined. The AI ​​translation platform then performs a secondary screening of the content for sensitive information. The AI ​​translation platform calls translation prompts in different languages ​​according to the translation language. Different big models can be used for different language prompts. The AI ​​translation platform also inputs business scenario information into the prompts to generate translation prompts that match the business scenario, including initial translation prompts, reflective translation prompts, and improved translation prompts.

[0086] The AI ​​translation platform calls the interface of the translation model, inputs the content to be translated and the initial translation prompt words into the translation model, and the translation model acts as an intelligent agent to execute reflection-based translation steps, including the intelligent agent performing an initial translation of the content to be translated according to the initial translation prompt words to obtain initial translation data, then obtaining reflective translation prompt words, inputting business data, initial translation data and reflective translation prompt words into the intelligent agent for translation analysis to obtain translation suggestion data, obtaining improved translation prompt words, and inputting business data, initial translation data, translation suggestion data and improved translation prompt words into the intelligent agent for improved translation to obtain target translation data.

[0087] The agent sends the output target translation data to the AI ​​translation platform for translation result verification, which can include formatting and sensitive information verification. Once verification passes, the translation (i.e., the target translation data) is displayed in the ERP system. The ERP system receives feedback on abnormal translations entered by the user and returns it to the ERP system to perform interface parameter verification and interface traffic limit detection on the translation model. It then generates updated translation prompts, which are then used to optimize the anomaly until the updated translation data is obtained.

[0088] In an exemplary embodiment, Figure 7 As shown in the figure, a schematic diagram of editing translation prompt words is provided. The business system can provide prompt word editing function. The user sends a prompt word editing request to the business system, and the business system displays the prompt word editing box on the current page, such as Figure 7The prompt word encoding content described in the [ ] includes editable roles, skills, and restrictions. Users can then add custom variables to the edited prompt word, such as business scenarios and restrictions, to obtain the edited translated prompt word. The business system uses the edited translated prompt word as the initial translation prompt word and proceeds to the step of obtaining the initial translation prompt word corresponding to the target language until the target translation data corresponding to the business data is obtained.

[0089] In an exemplary embodiment, step 204, the business data and the initial translation prompt word are input into the translation model for language translation to obtain the initial translation data corresponding to the business data, including:

[0090] Inputting the business data and the initial translation prompt words into multiple translation models for language translation, and obtaining candidate translation data corresponding to each translation model;

[0091] Determine a target translation model from among the translation models based on the candidate translation data;

[0092] The candidate translation data corresponding to each target translation model are fused to obtain fused translation data; the fused translation data is used to input the target translation model for translation analysis.

[0093] The candidate translation data refers to the translation data obtained by translating the business data according to the initial translation prompt words by each translation model.

[0094] For example, multiple translation models may be included. The terminal calls the interface of each translation model, inputs the business data and initial translation prompt words into each translation model for language translation, and receives candidate translation data returned by each translation model. The terminal then determines a target translation model from each translation model based on the candidate translation data.

[0095] In determining the target translation model, the terminal may calculate the semantic similarity between each candidate translation data set. Based on the semantic similarity between each candidate translation data set, the terminal may determine, from among the various translation models, the translation model to which the candidate translation data set with a semantic similarity greater than a preset threshold belongs as the target translation model. This may be accomplished by using a large language model (LLM) to calculate the semantic similarity between each candidate translation data set. The terminal determines the translation model to which the candidate translation data set with a similarity greater than a similarity threshold belongs as the target translation model, and determines the translation model to which the candidate translation data set with a similarity less than the similarity threshold belongs as the abnormal translation model. It is understood that the number of target translation models being greater than the number of abnormal translation models indicates that the number of similar candidate translation data sets among the multiple candidate translation data sets is greater than the number of dissimilar candidate translation data sets. The similar candidate translation data sets may be represented as the correct translation results of the business data using the corresponding target translation model, while the dissimilar candidate translation data sets may be represented as the abnormal translation results of the corresponding dissimilar translation model. For example, if there are five candidate translation data, and based on the semantic similarity between the candidate translation data, three of them are determined to be semantically similar, while the remaining two have significantly different semantics from the three candidate translation data, indicating that the translation results of these three candidate translation data for the business data are reliable, then the translation model to which these three candidate translation data belong is determined as the target translation model. The terminal can also cluster the candidate translation data and, based on the clustering results, determine the translation model to which the largest number of candidate translation data belonging to the same cluster belong as the target translation model.

[0096] The terminal fuses the candidate translation data corresponding to each target translation model. This can be done by inputting the candidate translation data corresponding to each target translation model into a large language model (LLM) for semantic fusion, resulting in fused translation data. The business data, fused translation data, and reflective translation prompts are input into each target translation model for translation analysis, resulting in translation suggestion data corresponding to each target translation model. Improved translation prompts are then obtained, and the business data, fused translation data, translation suggestion data, and improved translation prompts are input into the corresponding target translation model for improved translation, resulting in improved translation data corresponding to each target translation model. The improved translation data corresponding to each target translation model is then fused to obtain target translation data corresponding to the business data.

[0097] In this embodiment, when there are multiple translation models, a target translation model is determined based on the semantic similarity or clustering results between the candidate translation data output by each translation model. The target translation model is used to perform translation analysis and improve the translation of the business data, thereby ensuring the accuracy of the translation data.

[0098] In an exemplary embodiment, the method for translating business data into different languages ​​further includes:

[0099] determining an abnormal translation model from each translation model based on each candidate translation data;

[0100] generating an abnormal translation prompt word corresponding to the abnormal translation model based on the translation difference between the candidate translation data corresponding to the abnormal translation model and the candidate translation data corresponding to the target translation model;

[0101] The business data and the abnormal translation prompt words are input into the abnormal translation model for language translation to obtain updated translation data; the updated translation data is used to input into the abnormal translation model for translation analysis.

[0102] Exemplarily, the terminal inputs the business data and initial translation prompt words into multiple translation models for language translation, obtaining candidate translation data corresponding to each translation model. After determining the target translation model among the translation models, the terminal may determine the remaining translation models as abnormal translation models. Alternatively, based on the semantic similarity between the candidate translation data, the terminal may determine the translation model to which the candidate translation data with a semantic similarity less than a similarity threshold belongs as the abnormal translation model. Alternatively, based on the clustering results of the translation models, the terminal may determine the translation model to which the candidate translation data belonging to the same category and with the smallest number belongs as the abnormal translation model. Based on the translation differences between the candidate translation data corresponding to the abnormal translation model and the candidate translation data corresponding to the target translation model, the terminal generates abnormal translation prompt words corresponding to the abnormal translation model.

[0103] The business data and abnormal translation prompt words are then input into the abnormal translation model for language translation, resulting in updated translation data output by the abnormal translation model. When the semantic similarity between the updated translation data and the fused translation data exceeds a similarity threshold, a reflective translation prompt word is obtained, and the business data, updated translation data, and reflective translation prompt words are input into the abnormal translation model for translation analysis, resulting in translation suggestion data corresponding to the abnormal translation model. Improved translation prompt words are then obtained, and the business data, updated translation data, translation suggestion data, and improved translation prompt words are input into the corresponding abnormal translation model for improved translation, resulting in improved translation data corresponding to the abnormal translation model. The improved translation data corresponding to the abnormal translation model is then fused with the improved translation data corresponding to each target translation model to obtain target translation data corresponding to the business data.

[0104] In this embodiment, an abnormal translation model is determined based on the semantic similarity between candidate translation data output by each translation model. The business data and abnormal translation prompt words are input into the abnormal translation model for language translation to obtain updated translation data. When the updated translation data is similar to the fused translation data, the abnormal translation model is determined to correct the candidate translation data with the abnormal translation, thereby ensuring the accuracy of the updated translation data. The improved translation data corresponding to the abnormal translation model is then fused with the improved translation data corresponding to each target translation model to obtain target translation data corresponding to the business data. This ensures the diversity of the improved translation data and improves the accuracy of the target translation data.

[0105] In an exemplary embodiment, the method for translating business data into different languages ​​further includes:

[0106] Query the translation results corresponding to the business data from the language translation package corresponding to the target language;

[0107] If the query fails, the process proceeds to the step of obtaining the initial translation prompt word corresponding to the target language until the target translation data is obtained;

[0108] Update the language translation package corresponding to the target language based on the target translation data.

[0109] The language translation package refers to the translated data of various languages ​​pre-stored locally in the terminal.

[0110] Exemplarily, after obtaining the business data and the target language, the terminal searches the language translation package (i.e., the local knowledge base) for the target language for the translation results corresponding to the business data. If no translation data corresponding to the business data is found in the language translation package, the query is determined to have failed, and the process proceeds to obtain the initial translation prompt words corresponding to the target language until the target translation data is obtained. The target translation data is then format-checked. If the format is verified, the target translation data is stored in the language translation package corresponding to the target language.

[0111] In an exemplary embodiment, after obtaining the business data, the terminal searches the local knowledge base for local translation data corresponding to the business data. The terminal may search for local translation data corresponding to existing business data in the local knowledge base, and then use the translation model to translate the target business data for the target business data that does not exist in the local knowledge base. The target business data is used as the business data, and the step of obtaining the initial translation prompt words corresponding to the target language is executed until the model translation data corresponding to the target business data is obtained. Then, the target translation data corresponding to the business data is obtained based on the model translation data corresponding to the target business data and the local translation data corresponding to the existing business data, and the local knowledge base is updated based on the model translation data corresponding to the target business data.

[0112] In this embodiment, by updating the language translation package corresponding to the target language according to the target translation data, the data volume of the language translation package is increased, so that the translation efficiency is improved when the translation result is subsequently searched according to the language translation package.

[0113] In an exemplary embodiment, Figure 8 As shown, a schematic diagram of a reflection-based translation process is provided. The ERP system presents a current page. The user opens the form metadata on this page, which is the business data entered by the user. The user then clicks the translation menu and selects at least one target language, triggering the metadata translation application to extract metadata terms, i.e., the data to be translated from the business data. The application then searches the local term library (i.e., local knowledge base) based on the data to be translated, finding translation data for metadata already in the local term library and identifying metadata to be translated for translation data not already in the local term library. The metadata translation application then invokes a translation agent (i.e., translation model), which executes the reflection-based translation steps, including initial translation of the metadata to be translated, reflection suggestions (translation analysis based on reflection translation prompts), and improved translation, resulting in the target translation data. The translation agent sends the target translation data to the metadata translation application for result display. The user proofreads the target translation data for translation terms. The metadata translation application converts the proofread target translation data into metadata of at least one target language (i.e., multilingual metadata) and displays it. The user adjusts the format of the multilingual metadata, and the metadata translation application saves the format-adjusted metadata, which may be stored in a local knowledge base.

[0114] Among them, when the metadata translation application calls the translation agent, a JSON object that conforms to the business usage KEY can be created through the structure recognition optimization component:

[0115] {

[0116] “errCode”: “0”, (translation error ID)

[0117] "status": "true", (whether the interface call is successful)

[0118] “data”:

[0119] {

[0120] "llmValue": "['Chỉnh sửa','Ý kiến','Xóa'] "

[0121] }

[0122] }.

[0123] After users adjust the format of multilingual metadata, they can use the text preprocessing component to convert a text that does not conform to the JSON format into a text that strictly conforms to the JSON text format (i.e., convert the non-standard format):

[0124] {

[0125] “input”: “['edit','view','delete']” (entry data to be translated)

[0126] “promptNumber”: “prompt-2405147584XX, (selected translation model information)

[0127] "stream": false, (the construction type of the output result)

[0128] "business":"", (entry business scenario information)

[0129] "varParams": "['view','edit']" (error limiting array, prompting that translation into this information is not allowed),

[0130] "IslimitPolicy": true, (whether to provide error limits)

[0131] } (Identify and correct structural defects).

[0132] In this embodiment, by identifying the translation language and business scenario on the ERP side and transmitting the translation language and business scenario to the AI ​​translation platform, translation prompt words that are consistent with the business scenario are generated to constrain the large model to return accurate translation results that are consistent with the business scenario, thereby improving the translation accuracy of business data.

[0133] In an exemplary embodiment, when the target language corresponding to the business data is not a designated language, the designated language can be a commonly used language, such as English. The designated language is used as the target language, and the process proceeds to the step of obtaining the initial translation prompt word corresponding to the target language until intermediate translation data in the designated language corresponding to the business data is obtained. The intermediate translation data is then used as the business data, and the process returns to the step of obtaining the initial translation prompt word corresponding to the target language until target translation data in the target language corresponding to the business data is obtained. For example, for translation from Chinese to a minority language, or between minority languages, English is set as the designated language and the minority language as the target language. The business data is first translated into English, and then translated into the minority language from English. Compared to direct translation into the minority language, using a translation model to analyze and improve the translation based on a knowledge base of the commonly used designated language can improve the accuracy of the translation results, thereby improving the translation accuracy of the minority language.

[0134] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0135] Based on the same inventive concept, embodiments of the present application also provide a business data language translation device for implementing the aforementioned business data language translation method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more of the following embodiments of the business data language translation device can be found in the aforementioned limitations of the business data language translation method and will not be further elaborated here.

[0136] In an exemplary embodiment, Figure 9 As shown, a language translation device 900 for business data is provided, comprising: a data acquisition module 902, an initial translation module 904, a translation suggestion module 906, and an improved translation module 908, wherein:

[0137] Data acquisition module 902, used to obtain business data to be translated and the target language;

[0138] Initial translation module 904, used to obtain initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into the translation model for language translation, and obtain initial translation data corresponding to the business data;

[0139] The translation suggestion module 906 is used to obtain reflection translation prompt words, input the business data, initial translation data and reflection translation prompt words into the translation model for translation analysis, and obtain translation suggestion data;

[0140] The improved translation module 908 is used to obtain improved translation prompt words, input the business data, initial translation data, translation suggestion data and improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

[0141] In an exemplary embodiment, the translation suggestion module 906 is further configured to obtain a business scenario corresponding to the business data; determine business translation requirement information based on the business scenario; and generate reflection translation prompt words based on the business scenario and the business translation requirement information.

[0142] In an exemplary embodiment, the translation suggestion module 906 is also used to determine business translation requirement information based on the business type corresponding to the business data and the data entry scenario when the business scenario is a data entry scenario; when the business scenario is a business communication scenario, obtain the identity information of the target user to whom the business data belongs, and determine the business translation requirement information based on the identity information and the business communication scenario.

[0143] In an exemplary embodiment, the language translation device 900 for business data is further configured to obtain abnormal translation feedback corresponding to the target translation data, generate updated translation prompt words based on the abnormal translation feedback, use the updated translation prompt words as initial translation prompt words, and return to the step of inputting the business data and the initial translation prompt words into the translation model for language translation until updated target translation data corresponding to the business data is obtained.

[0144] In an exemplary embodiment, the initial translation module 904 is further configured to input the business data and the initial translation prompt words into multiple translation models for language translation to obtain candidate translation data corresponding to each translation model; determine a target translation model from each translation model based on each candidate translation data; fuse the candidate translation data corresponding to each target translation model to obtain fused translation data; and input the fused translation data into the target translation model for translation analysis.

[0145] In an exemplary embodiment, the initial translation module 904 is further configured to determine an abnormal translation model from various translation models based on various candidate translation data; generate abnormal translation prompt words corresponding to the abnormal translation model based on translation differences between the candidate translation data corresponding to the abnormal translation model and the candidate translation data corresponding to the target translation model; input the business data and the abnormal translation prompt words into the abnormal translation model for language translation to obtain updated translation data; and input the updated translation data into the abnormal translation model for translation analysis.

[0146] In an exemplary embodiment, the language translation device 900 for business data is further configured to query a language translation package corresponding to the target language for a translation result corresponding to the business data; when the query fails, the step of obtaining an initial translation prompt word corresponding to the target language is executed until the target translation data is obtained; and the language translation package corresponding to the target language is updated based on the target translation data.

[0147] Each module in the aforementioned business data language translation device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0148] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 10 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store business data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a language translation method for business data is implemented.

[0149] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 11As shown. The computer device includes a processor, memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals via wired or wireless means, and the wireless means may be implemented via Wi-Fi, a mobile cellular network, near-field communication (NFC), or other technologies. When executed by the processor, the computer program implements a method for translating business data between languages. The display unit of the computer device is used to form a visually visible image and may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0150] Those skilled in the art will understand that Figure 10-11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0151] In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0152] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0153] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0154] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0155] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0156] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0157] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for translating business data into different languages, characterized in that: The method comprises: Obtain the business data and target language to be translated; Obtaining initial translation prompt words corresponding to the target language, inputting the business data and the initial translation prompt words into a translation model for language translation, and obtaining initial translation data corresponding to the business data; Obtaining a reflective translation prompt word, inputting the business data, the initial translation data, and the reflective translation prompt word into the translation model for translation analysis to obtain translation suggestion data; Obtaining improved translation prompt words, inputting the business data, the initial translation data, the translation suggestion data and the improved translation prompt words into the translation model for improved translation, and obtaining target translation data corresponding to the business data.

2. The method according to claim 1, characterized in that The step of obtaining the reflection translation prompt words includes: Obtaining the business scenario corresponding to the business data; Determining business translation demand information based on the business scenario; Reflective translation prompt words are generated based on the business scenario and the business translation requirement information.

3. The method according to claim 2, characterized in that The determining of business translation requirement information based on the business scenario includes: When the business scenario is a data entry scenario, determining business translation requirement information based on the business type corresponding to the business data and the data entry scenario; When the business scenario is a business communication scenario, identity information of a target user to which the business data belongs is obtained, and business translation requirement information is determined based on the identity information and the business communication scenario.

4. The method according to claim 1, wherein After inputting the business data, the initial translation data, the translation suggestion data, and the improved translation prompt words into the translation model for improved translation to obtain target translation data corresponding to the business data, the method further includes: Obtaining abnormal translation feedback corresponding to the target translation data, and generating updated translation prompt words based on the abnormal translation feedback; The updated translation prompt word is used as the initial translation prompt word, and the process returns to the step of inputting the business data and the initial translation prompt word into the translation model for language translation, until the updated target translation data corresponding to the business data is obtained.

5. The method according to claim 1, wherein The step of inputting the business data and the initial translation prompt word into a translation model for language translation to obtain initial translation data corresponding to the business data includes: Inputting the business data and the initial translation prompt words into multiple translation models for language translation to obtain candidate translation data corresponding to each translation model; Determining a target translation model from the translation models based on the candidate translation data; The candidate translation data corresponding to each target translation model are fused to obtain fused translation data; the fused translation data is used to input the target translation model for translation analysis.

6. The method according to claim 5, characterized in that The method further comprises: determining an abnormal translation model from the translation models based on the candidate translation data; generating an abnormal translation prompt word corresponding to the abnormal translation model based on a translation difference between the candidate translation data corresponding to the abnormal translation model and the candidate translation data corresponding to the target translation model; The business data and the abnormal translation prompt word are input into the abnormal translation model for language translation to obtain updated translation data; the updated translation data is used to input into the abnormal translation model for translation analysis.

7. The method according to claim 1, characterized in that The method further comprises: Querying the translation result corresponding to the business data from the language translation package corresponding to the target language; When the query fails, the step of obtaining the initial translation prompt word corresponding to the target language is executed until the target translation data is obtained; The language translation package corresponding to the target language is updated based on the target translation data.

8. A language translation device for business data, characterized in that: The device comprises: Data acquisition module, used to obtain business data to be translated and the target language; An initial translation module, configured to obtain initial translation prompt words corresponding to the target language, input the business data and the initial translation prompt words into a translation model for language translation, and obtain initial translation data corresponding to the business data; A translation suggestion module, configured to obtain a reflection translation prompt word, input the business data, the initial translation data, and the reflection translation prompt word into the translation model for translation analysis, and obtain translation suggestion data; The improved translation module is used to obtain improved translation prompt words, input the business data, the initial translation data, the translation suggestion data and the improved translation prompt words into the translation model for improved translation, and obtain target translation data corresponding to the business data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images