Method for determining translator for entrusted translation and translation price and server-side server

Abstract

The invention relates to a method for determining a translator entrusted to translate and a translation price and a server-side server. The method comprises the steps of receiving an original text file input by a user, calculating feature vector similarities between a plurality of translation files of a plurality of translators and the original text file, and identifying related translation files based on the similarities. Then, a plurality of candidate translators are screened according to the number of the related translation files, and the translation unit price of each translator is calculated based on the lowest price and the highest price input by the translators and the number of the related translation files, so that a final translator is selected. Then, the translation price and the translation deadline are adjusted according to a preset rule and are sent to the user; and finally, transmitting the original text file to a translator according to the payment condition of the user.

Description

Technical Field

[0001] The present invention relates to a method and a server for determining a translator and a translation price for commissioned translation. The method receives an original text file input by a user, calculates the similarity of feature vectors between multiple translation files of multiple translators and the original text file, and identifies relevant translation files based on the similarity. Then, multiple candidate translators are screened according to the number of relevant translation files, and the translation unit price of each translator is calculated based on the minimum price and the maximum price input by the translator and the number of relevant translation files, so as to select the final translator. Subsequently, the translation price and the translation deadline are adjusted according to preset rules and sent to the user. Finally, according to the payment situation of the user, the original text file is transmitted to the translator. Background Art

[0002] With the expansion of the global market, multilingual communication has become a basic need in modern society. Especially in the commercial field of selling products or providing services internationally, and in the academic field of sharing international information, the professional requirements for multilingual communication are increasing day by day. However, due to the high labor cost of professional translators, in recent years, enterprises and institutions have chosen to outsource translation services to minimize labor costs instead of directly hiring in-house translators. However, due to the large demand for translation, it is often difficult to accurately evaluate the professionalism of translators, resulting in the disordered allocation of commissioned translation and the decline in the reliability of translation results.

[0003] Existing inventions related to translation agency services, such as Korean Patent Publication No. 10-2021-0050207, relate to a translation agency system and method. This invention ensures that the selected translator matches the professional field of the commissioned translation project by recommending translators meeting preset conditions to the applicant. However, this technology does not cover the requirements of the client for the cost or working hours of the translation project. In addition, in the agency service connecting the client and the translator, it is crucial to establish a high level of trust between the two parties. Therefore, a reasonable determination of the translation cost and working hours acceptable to both parties is an ideal solution.

[0004] Therefore, there is an urgent need to develop a technology that enables a client to select a translator who can offer a reasonable cost and an acceptable working time when commissioning translation. Summary of the Invention

[0005] The object of the present invention is to provide a method and a server for determining a translator for commissioned translation and a translation price. The method receives the original text file input by the user, calculates the similarity of the feature vectors between multiple translation files of each translator and the original text file, and identifies relevant translation files based on the similarity. Then, multiple candidate translators are screened according to the number of relevant translation files, and the translation unit price of each translator is calculated based on the minimum price and the maximum price input by the translator and the number of relevant translation files, so as to select the final translator. Subsequently, the translation price and the translation deadline are adjusted according to preset rules and sent to the user. Finally, according to the payment situation of the user, the original text file is sent to the translator.

[0006] To solve the above problems, an implementation of the present invention provides a method for determining a translator for commissioned translation and a translation price, which is executed on a server of a service end associated with an application program executed on a user terminal and a translator terminal, including: a translator information receiving step: receiving translator information including information on the type of translation language, the type of translation, and the range of the unit price of each word translation including the minimum price and the maximum price from the translator terminal; an original text file receiving step: receiving the type of translation language, the type of translation, and the original text file to be translated according to the input of the user on the user terminal; a feature vector extraction step: extracting the main keywords and related information of the entire text as feature vectors from a file containing a document through a trained feature extraction model based on an artificial neural network, and extracting a first feature vector from the above original text file; a similarity calculation step: for multiple translators that meet the type of translation language and the type of translation, extracting second feature vectors from multiple translation files of each translator stored in the database, and calculating the similarity between these second feature vectors and the above first feature vector; a candidate translator screening step: identifying translation files with a similarity reaching a preset value or more as relevant translation files, and screening a translator as a candidate translator when the number of relevant translation files of each translator reaches a preset number or more; a final translator screening step: calculating the translation unit price of each candidate translator according to the following [Formula 1], and selecting the candidate translator with the lowest translation unit price as the final translator; a translation price sending step: calculating a first translation price based on the translation unit price of the selected final translator and the number of words in the above original text file, and calculating a first translation deadline based on the number of words; then reducing the above first translation deadline and increasing the above first translation price according to preset rules to derive one or more second translation prices and second translation deadlines, and sending the information of the above first translation price, the first translation deadline, one or more second translation prices and second translation deadlines to the user terminal; an original text file sending step: sending the above original text file and the translation deadline information calculated based on the user payment information to the translator terminal.

[0007] [Formula 1]:

[0008] The translation unit price of the candidate translator = CostLOW + (CostHIGH - CostLow) * (NumberTOTAL - Number)

[0009] / NumberTOTAL

[0010] (wherein, CostLOW is the above-mentioned lowest price, CostHIGH is the above-mentioned highest price, NumberTOTAL is the sum of the respective numbers of relevant translation files of all candidate translators, and Number is the number of relevant translation files of this candidate translator)

[0011] In an implementation manner of the present invention, the above-mentioned preset rules are as follows: the above-mentioned one or more second translation deadlines are respectively derived based on a fixed unit time within a preset time range relative to the first translation deadline; the above-mentioned one or more second translation prices are respectively derived inversely proportional to the above-mentioned second translation deadlines and are limited within a preset lowest lower limit price and a preset highest upper limit price range; the above-mentioned preset lowest lower limit price is lower than the first translation price and not less than 0; the above-mentioned preset highest upper limit price is higher than the first translation price and is determined based on the shortest second translation deadline among the above-mentioned one or more second translation deadlines.

[0012] In an implementation manner of the present invention, the above-mentioned translation file belongs to the file translated by the translator for other original text files and contains text composed of the same language as the above-mentioned original text files.

[0013] In an implementation manner of the present invention, the method for determining the translator and the translation price further includes a template format conversion step for the original text file, and the above-mentioned template format conversion step includes: a template file receiving step: receiving a template file containing multiple sub-template formats according to the input of the user at the user terminal; a sub-template format detecting step: detecting multiple sub-template formats from the above-mentioned template file through a trained detection model based on an artificial neural network; an original text template format detecting step: detecting multiple original text template formats from the above-mentioned original text file through the above-mentioned detection model; an original text file conversion step: converting the multiple original text template formats contained in the above-mentioned original text file into multiple sub-template formats and inputting the detailed data input to the above-mentioned original text template formats into the corresponding sub-template formats.

[0014] In one implementation of the present invention, the above-mentioned original text file conversion step includes: for the original text template format that fails to correspond to the sub-template format, deleting the detailed data input to the original text template format according to the user input received from the user terminal, or inserting the detailed data of the original text template format into the detailed data insertion step of the sub-template format based on the template format matching table that matches the sub-template format according to the original text template format; for the sub-template format that fails to correspond to the original text template format, a notification reminder step of providing a relevant notification of the sub-template format to the user terminal.

[0015] In one implementation of the present invention, the above-mentioned template format matching table includes multiple original text template formats and multiple sub-template formats, and for each of the multiple original text template formats, at least one sub-template format is pre-matched; when at least one sub-template format matches a certain original text template format, it further includes a preset priority for the at least one sub-template format; the above-mentioned detailed data insertion step inserts the detailed data into the sub-template format based on the above-mentioned preset priority.

[0016] To solve the above problems, an implementation of the present invention provides a server that is associated with application programs running on a user terminal and an interpreter terminal, and executes a method for determining an interpreter for commissioned translation and a translation price, including: an interpreter information receiving unit: receiving interpreter information including information on the range of translation unit prices per word, including the types of translation languages, translation categories, and the minimum price and maximum price, from the interpreter terminal; a source document receiving unit: receiving, according to the input of the user on the user terminal, the types of translation languages, translation categories, and the source document to be translated; a feature vector extraction unit: extracting the main keywords and related information of the entire text as feature vectors from a file containing a document through a trained feature extraction model based on an artificial neural network, and extracting a first feature vector from the above source document; a similarity calculation unit: extracting second feature vectors from multiple translation files of each interpreter stored in the database for multiple interpreters that meet the types of translation languages and translation categories, and calculating the similarity between these second feature vectors and the above first feature vector; a candidate interpreter screening unit: identifying translation files with a similarity above a preset value as relevant translation files, and screening an interpreter as a candidate interpreter when the number of relevant translation files of each interpreter reaches a preset number or more; a final interpreter screening unit: calculating the translation unit price of each candidate interpreter according to the following [Formula 1], and selecting the candidate interpreter with the lowest translation unit price as the final interpreter; a translation price sending unit: calculating a first translation price based on the translation unit price of the selected final interpreter and the number of words in the above source document, and calculating a first translation deadline based on the number of words; then deriving one or more second translation prices and second translation deadlines while reducing the above first translation deadline and increasing the above first translation price according to a preset rule, and sending the information on the above first translation price, first translation deadline, one or more second translation prices, and second translation deadlines to the user terminal; a source document sending unit: sending the above source document and the translation deadline information calculated based on the user payment information to the interpreter terminal according to the user's payment situation. The above constitutes the provided server.

[0017] [Formula 1]:

[0018] The translation unit price of the candidate interpreter = CostLOW + (CostHIGH - CostLow) * (NumberTOTAL - Number)

[0019] / NumberTOTAL

[0020] (where CostLOW is the above minimum price, CostHIGH is the above maximum price, NumberTOTAL is the sum of the numbers of relevant translation files of all candidate interpreters, and Number is the number of relevant translation files of this candidate interpreter)

[0021] Effect of the Invention

[0022] In one implementation of the present invention, through a trained feature extraction model based on an artificial neural network, the feature vectors of the original text file commissioned by the user for translation are compared with the feature vectors of the translation file completed by the translator, and candidate translators are screened according to the similarity, thereby improving the reliability of the translation result and providing higher satisfaction for the user.

[0023] In one implementation of the present invention, relevant translation files with a feature vector similarity reaching a preset value or more are screened out from multiple translation files, and candidate translators are screened based on this, thereby improving the reliability of the translation result and providing higher satisfaction for the user.

[0024] In one implementation of the present invention, by using a specific formula with the minimum price and maximum price input by the translator on the server as variables, the final translator is selected from multiple candidate translators, thereby providing higher convenience for the translators using this server.

[0025] In one implementation of the present invention, based on preset rules, additional translation prices and translation deadlines are derived for the calculated translation price and translation deadline of the final translator, so that the user can select options that meet their own needs in terms of the cost and time of commissioned translation.

[0026] In one implementation of the present invention, when the document format used by the organization that needs to submit is different from the format of the original file, by using a template file made in the format used by this organization, the format of the original file is automatically converted into the corresponding format, thereby providing convenience for the user. Brief Description of the Drawings

[0028] Figure 1 Describes the components of the server for determining the translator and translation price in one implementation of the present invention.

[0029] Figure 2 Briefly illustrates the execution steps of the method for determining the translator and translation price in one implementation of the present invention.

[0030] Figure 3 Briefly illustrates the application program interface running on the translator terminal for registering translator information in one implementation of the present invention.

[0031] Figure 4 Briefly illustrates the application program interface running on the user terminal for the user to submit a translation commission in one implementation of the present invention.

[0032] Figure 5Briefly described is the process of extracting the first feature vector and the second feature vector in one implementation of the present invention.

[0033] Figure 6 Briefly described is the process of identifying relevant translation files of a translator based on the similarity of each translation file in one implementation of the present invention.

[0034] Figure 7 Briefly described is the process of screening multiple candidate translators in one implementation of the present invention.

[0035] Figure 8 Briefly described is the process of calculating the translation unit price of each translator based on a preset formula in one implementation of the present invention.

[0036] Figure 9 Briefly described is the process of deriving the first translation price, the first translation deadline, the second translation price, and the second translation deadline in one implementation of the present invention.

[0037] Figure 10 Briefly described is the application program interface running on the user terminal for selecting the translation price and the translation deadline in one implementation of the present invention.

[0038] Figure 11 Briefly described is the application program interface running on the user terminal for converting the template format of the original file in one implementation of the present invention.

[0039] Figure 12 Briefly described is the process of converting the template format of the original file in one implementation of the present invention.

[0040] Figure 13 Briefly described is the structure of the template format matching table in one implementation of the present invention.

[0041] Specific content for implementing the present invention

[0042] Various implementations and / or aspects will be disclosed below with reference to the accompanying drawings. In the following description, in order to facilitate the understanding of one or more aspects, a number of specific details will be disclosed. However, those of ordinary skill in the art should recognize that these aspects can also be implemented without involving specific details. The subsequent description and the drawings detail specific exemplary implementations of one or more aspects. However, these aspects are only examples, and there may be different implementations, and the described content is intended to cover all these aspects and their equivalent variations.

[0043] In addition, various aspects and features may be presented by a system including multiple devices, components, and / or modules. It should be understood that various systems may include additional devices, components, and / or modules, or may not include all devices, components, and / or modules in the discussion related to the accompanying drawings.

[0044] Terms such as "implementation", "example", "aspect", "exemplary", etc. used in this specification do not mean that certain aspects or designs described are superior to other aspects or designs, or have specific advantages. In addition, terms such as "part", "component", "module", "system", "interface", etc. generally refer to computer-related entities, such as hardware, the combination of hardware and software, or pure software.

[0045] In addition, it should be understood that the terms "include" and / or "comprise" indicate the presence of the described features and / or components, but do not exclude the presence or addition of one or more other features, components, and / or their combinations.

[0046] In addition, terms with ordinal numbers, such as "first", "second", etc., are used to describe various components, but do not mean a limitation on the components. These terms are only used to distinguish components. Preferably, without exceeding the scope of the claims of the present invention, the "first component" may be referred to as the "second component", and similarly, the "second component" may also be referred to as the "first component". In addition, the term "and / or" includes combinations of multiple related items, or any one of multiple related items.

[0047] In addition, in the implementation of the present invention, unless otherwise clearly defined, all terms, including technical terms or scientific terms, should be interpreted according to the meaning commonly understood by those skilled in the art. Commonly used terms should be consistent with the context of the related technology, and unless clearly defined in the implementation of the present invention, should not be interpreted as having an overly idealized or formalized meaning.

[0048] Figure 1 The components of the server (1000) for determining interpreters and translation prices in an implementation of the present invention are described. Figure 2 The execution steps of the method for determining interpreters and translation prices in an implementation of the present invention are briefly described.

[0049] As Figures 1 to 2As shown, the application programs running on the user terminal (3000) and the interpreter terminal (2000) are associated with the server (1000) and execute a method for determining an interpreter and a translation price for commissioned translation. The method includes the following steps: Interpreter information receiving step (S100): Receive interpreter information including information on the range of translation unit prices per word for each language type, translation category, and minimum and maximum prices from the interpreter terminal (2000). Source document receiving step (S200): Receive the language type, translation category, and the source document to be translated according to the input of the user on the user terminal (3000). Feature vector extraction step (S300): Extract the main keywords and related information of the entire text as feature vectors from the file containing the document through a trained feature extraction model based on an artificial neural network, and extract the first feature vector from the above source document. Similarity calculation step (S400): For multiple interpreters meeting the language type and translation category, extract the second feature vectors from multiple translation files of each interpreter stored in the database (1900), and calculate the similarity between these second feature vectors and the above first feature vector. Candidate interpreter screening step (S500): Identify the translation files with a similarity above a preset value as relevant translation files, and when the number of relevant translation files of each interpreter reaches or exceeds a preset number, screen this interpreter as a candidate interpreter. Final interpreter screening step (S600): Calculate the translation unit price of each candidate interpreter according to the following [Formula 1], and select the candidate interpreter with the lowest translation unit price as the final interpreter. Translation price sending step (S700): Calculate the first translation price based on the translation unit price of the selected final interpreter and the number of words in the above source document, and calculate the first translation deadline based on the number of words; then, while reducing the above first translation deadline according to a preset rule, increase the above first translation price to derive one or more second translation prices and second translation deadlines, and send the information on the above first translation price, first translation deadline, one or more second translation prices, and second translation deadlines to the user terminal (3000). Source document sending step (S800): Send the above source document and the translation deadline information calculated based on the user payment information to the interpreter terminal (2000) according to the user's payment situation; the above constitutes a method for determining an interpreter and a translation price.

[0050] [Formula 1]:

[0051] Translation unit price of candidate interpreter = CostLOW + (CostHIGH - CostLow) * (NumberTOTAL - Number)

[0052] / NumberTOTAL

[0053] (where CostLOW is the aforementioned lowest price, CostHIGH is the aforementioned highest price, NumberTOTAL is the sum of the relevant translation file quantities of all candidate interpreters, and Number is the relevant translation file quantity of this candidate interpreter)

[0054] Specifically, the method for determining an interpreter and a translation price in the present invention is executed on the server (1000) shown in Figure 1 In one implementation of the present invention, the server (1000) is connected to a user terminal (3000) and an interpreter terminal (2000). The interpreter terminal (2000) displays an interface for visually presenting a translation service through its own screen, and an interpreter using the interpreter terminal (2000) can operate by means of selection input on this interface. Similarly, the user terminal (3000) displays an interface for visually presenting a translation service through its own screen, and a user using the user terminal (3000) can operate by means of selection input on this interface.

[0055] The aforementioned server (1000) includes an interpreter information receiving unit (1100), a source file receiving unit (1200), a feature vector extraction unit (1300), a similarity calculation unit (1400), a candidate interpreter screening unit (1500), a final interpreter screening unit (1600), a translation price sending unit (1700), a source file sending unit (1800), and a database (1900) for storing multiple translation files of multiple interpreters.

[0056] The interpreter information receiving step (S100) is executed by the interpreter information receiving unit (1100), and receives interpreter information based on interpreter input from the interpreter terminal (2000). The aforementioned interpreter information includes the language types that this interpreter can translate, the translation categories corresponding to the professional fields of this interpreter, and the information on the unit price range for each word translation of this interpreter. Among them, the unit price range information includes the lowest price and the highest price for each word input by this interpreter.

[0057] The source file receiving step (S200) is executed by the source file receiving unit (1200), and receives the translation language type, translation category determined based on user input, and the source file uploaded by the user from the user terminal (3000). That is, the server (1000) sends the source file to an appropriate interpreter, and after the interpreter completes the translation, receives the translated file from the interpreter terminal (2000) and then sends it to the user terminal (3000), thereby performing a translation intermediary service.

[0058] The feature vector extraction step (S300) is executed by the feature vector extraction unit (1300), and the main keywords and related information are extracted from the original text file as feature vectors through a trained artificial neural network-based feature extraction model. The above first feature vector includes the main characteristics of the original text file through a vector structure, and when the original text file belongs to an academic category document, the first feature vector extracted from the original text file can represent the academic and technical field to which the original text file belongs.

[0059] The similarity calculation step (S400) is executed by the similarity calculation unit (1400), and multiple interpreters whose translation language types and translation categories input in the interpreter terminal (2000) match those input in the user terminal (3000) are screened out. For these interpreters, second feature vectors are extracted from multiple translation files of the interpreter stored in the database (1900) through the feature extraction model, and the similarity between the second feature vector and the above first feature vector is calculated. The interpreter with a higher similarity and more second feature vectors is more suitable for performing the commissioned translation of the original text file. Therefore, the server (1000) can screen multiple candidate interpreters through the following candidate interpreter screening step (S500).

[0060] The candidate interpreter screening step (S500) is executed by the candidate interpreter screening unit (1500), and the translation files with a similarity higher than a preset value are identified as relevant translation files. And if the number of relevant translation files of a certain interpreter reaches or exceeds the preset number, then this interpreter is selected as a candidate interpreter. The above preset value and preset number should be set within the range where the server (1000) can select an appropriate number of candidate interpreters for the user's commissioned translation.

[0061] The final interpreter screening step (S600) is executed by the final interpreter screening unit (1600), and the translation unit price of each candidate interpreter is calculated according to a preset formula, and the candidate interpreter with the lowest translation unit price is selected as the final interpreter. More specifically, the above preset formula takes the lowest price and the highest price input by the interpreter through the interpreter terminal (2000), and the number of relevant translation files obtained through the candidate interpreter screening step (S500) as parameters.

[0062] The translation price sending step (S700) is executed by the translation price sending unit (1700), the first translation price is calculated based on the translation unit price of the selected final interpreter and the number of words in the original text file, and the first translation period is calculated based on the number of words. At the same time, at least one second translation price and at least one second translation period are calculated according to preset rules, and the information of the first translation price, the first translation period, the second translation price and the second translation period is sent to the user terminal (3000).

[0063] On the other hand, the number of words in the source text file can be measured by the server (1000) through methods such as word tokenization, natural language processing, regular expressions, and stop word processing. A detailed description of the preset rules will be provided in Figure 9 and will be further elaborated in

[0064] The source text file sending step (S800) is executed by the source text file sending unit (1800). When the user makes a selection input and completes the payment for the first translation price, the first translation cycle, at least one second translation price, and at least one second translation cycle on the user terminal (3000), the source text file and the translation cycle information determined based on the user's payment information are sent to the translator terminal (2000) according to the user's payment confirmation result.

[0065] Figure 3 To illustrate an implementation manner of the present invention, an application program interface running on the translator terminal (2000) is used to register translator information.

[0066] As Figure 3 shown, the method includes a translator information receiving step (S100) of receiving translator information including the types of translation languages, translation categories, and the per-word translation unit price range information including the minimum price and the maximum price from the translator terminal (2000).

[0067] Specifically, the translator can input translator information through the interface displayed in the application program running on the translator terminal (2000). In an implementation manner of the present invention, the translator information includes the languages that the translator can translate, the categories of documents that can be translated, and the translation unit price range. In addition, the interface includes a first translation language selection layer (L1), a first translation category selection layer (L2), a translation unit price selection layer (L3), and an editing layer (L4).

[0068] The translator can select the languages that he / she can translate through the first translation language selection layer (L1) and confirm the selected languages. In an implementation manner of the present invention, the first translation language selection layer (L1) displays English, Chinese, and Korean selected and input by the translator.

[0069] The translator can select the categories of documents that he / she can translate through the first translation category selection layer (L2) and confirm the selected categories. In an implementation manner of the present invention, the first translation category selection layer (L2) displays the "academic" category selected and input by the translator.

[0070] The translator can input the minimum price and the maximum price of the per-word unit price of his / her translation work through the translation unit price selection layer (L3) to set the translation unit price range. In an implementation manner of the present invention, the server (1000) can limit the minimum price and the maximum price of the translation unit price within different preset ranges for input.

[0071] In addition, in one implementation of the present invention, as Figure 3 shown, the editing layer (L4) may include a modification layer and a saving layer, and the translator can modify and save their translator information at any time through the editing layer (L4). The server (1000) stores the corresponding translator information in the database (1900) for each translator and selects the final translator based on this information.

[0072] Figure 4 Briefly described is an application program interface executed on a user terminal (3000) according to one implementation of the present invention, so that the user can submit a translation request.

[0073] As Figure 4 shown, the present invention relates to a method for determining a translator and a translation price, including a source file receiving step (S200) of receiving the types of translation languages, translation categories, and the source file to be translated input by the user from the user terminal (3000).

[0074] Generally speaking, Figure 4 (a) describes one implementation of the present invention, in which the application program interface displays a preliminary selection interface (L5), and the user can initially select the type of language they wish to translate therein. Figure 4 (b) describes another implementation of the present invention, in which the application program interface displays a plurality of input interfaces, and the user can input the type of translation language, translation category, and upload the source file.

[0075] Specifically, the user can run an application program connected to the server (1000) through the user terminal (3000) and, when wishing to commission document translation, first select the type of translation language. In addition, the user's input is not limited to selecting the required translation language. As Figure 4 shown in (a), in one implementation of the present invention, it may also include a plurality of optional layers to provide options for various contents such as document proofreading and customer center consultation.

[0076] When the user selects the layer corresponding to "Chinese-Korean translation" in the above first selection layer (L5), the application program interface will display the implementation described in Figure 4 (b), including a second translation language selection layer (L6), a second translation category selection layer (L7), a source file upload layer (L8), and an order layer (L9).

[0077] The user can further select the specific type of translation language within the language category selected in the first selection layer (L5) through the above second translation language selection layer (L6). Preferably, when the user selects "Chinese-Korean translation" in the first selection layer (L5), Figure 4The second translation language selection layer (L6) shown in (b) will include "Chinese-Korean translation" and "Korean-Chinese translation"; if the user selects "Chinese-English translation" in the first selection layer (L5), the second translation language selection layer (L6) will include "Chinese-English translation" and "English-Chinese translation". In addition, "Chinese-Korean translation" means translating Chinese into Korean, and "Korean-Chinese translation" means translating Korean into Chinese.

[0078] The user can select a translation category through the above-mentioned second translation category selection layer (L7). To ensure more accurate translation results, it is recommended to select a suitable translation category according to the content of the original document as the above-mentioned translation category. In one implementation of the present invention, the translation categories include "Daily", "Standard", "Academic", and "Business", but in other implementations, the second translation category selection layer (L7) can also include more translation categories.

[0079] The user can upload the original document to be translated to the server (1000) through the above-mentioned original document upload layer (L8). In one implementation of the present invention, as Figure 4 shown in (b), for the convenience of the user, the original document upload layer (L8) can also include relevant information such as the size limit of the original document, supported file formats, and file name requirements.

[0080] Subsequently, the user can submit a translation request for the original document to the server (1000) through the above-mentioned order layer (L9). In one implementation of the present invention, the order layer (L9) can also include a liaison layer for connecting to the customer service center so that the user can directly contact the customer service when needed.

[0081] Figure 5 Describes the process of extracting the first feature vector and the second feature vector in one implementation of the present invention.

[0082] As Figure 5 shown, in the method of determining an interpreter and a translation price, using a trained artificial neural network feature extraction model, the core keywords and related information of the overall text of the document are extracted from the file containing the document and converted into feature vectors. In this process, a feature vector extraction step (S300) is performed to extract the first feature vector from the original document.

[0083] In addition, the translated document refers to a document that the interpreter has completed the translation and contains text in the same language as the original document.

[0084] Generally speaking, Figure 5 (a) Describes the first feature vector extracted from the original document, Figure 5 (b) Describes the second feature vector extracted from multiple translated documents of multiple interpreters.

[0085] Specifically, as Figure 5 (a) shows, the feature extraction model extracts the first feature vector from the original text file uploaded by the user. In addition, as Figure 5 (b) shows, the model also extracts the second feature vectors from the translation files of multiple translators stored in the database (1900) of the server (1000). That is to say, the system extracts multiple second feature vectors according to the number of translators stored in the database (1900) and the number of their translation files.

[0086] On the other hand, the feature extraction model is a model based on an artificial neural network. After training, it can extract the core keywords and related information of the overall text of the document from the file containing the document and convert them into feature vectors. In other words, the first feature vector is closely related to the characteristics of the core keywords of the original text file uploaded by the user, and each second feature vector is closely related to the characteristics of the core keywords of its corresponding translation file.

[0087] In addition, the translation file refers to the file that the translator has completed the translation, and preferably contains the text in the same language as the original text file.

[0088] Figure 6 The process of identifying the translation files related to the translator according to the similarity of the translation files is described.

[0089] As Figure 6 shown, in the method for determining the translator and the translation price, for multiple translators who meet the translation language type and translation category, the system extracts the corresponding second feature vectors from the multiple translation files of each translator stored in the database (1900) by using the feature extraction model, and calculates the similarity between it and the first feature vector, that is, performs the similarity calculation step (S400).

[0090] Specifically, the server (1000) calculates the similarity between the first feature vector and the second feature vectors corresponding to the multiple translation files of multiple translators through the similarity calculation step (S400). In one implementation, the similarity can adopt the conventional methods for calculating the similarity between vectors, such as Cosine Similarity or Euclidean Distance.

[0091] After calculating the similarity for multiple translation files, the server (1000) selects the translation files with similarity higher than the preset threshold from the multiple similarity calculation results through the candidate translator selection step (S500) and identifies them as related translation files. In one implementation of the present invention, as Figure 6As shown, the preset threshold is set at 70%. Preferably, among the multiple translation files of Translator A (Translation File #A1 to Translation File #A5), only the similarity of Translation File #A1 reaches 78%, so only this file is recognized as the relevant translation file of Translator A.

[0092] The server (1000) extracts relevant translation files highly similar to the characteristics of the original text file, and based on this, screens multiple candidate translators and finally determines the final translator, thereby improving the satisfaction of users and translators using the server (1000).

[0093] Figure 7 The process of screening multiple candidate translators in one implementation manner of the present invention is briefly described.

[0094] As Figure 7 shown, in the method for determining a translator and a translation price, for multiple translators corresponding to a specified translation language type and translation category, multiple translation files of each translator are extracted from the database (1900), and the similarity between the second feature vector and the first feature vector of these files is calculated through a feature extraction model, and the similarity calculation step (S400) is executed.

[0095] Specifically, in one implementation manner of the present invention, as Figure 7 shown in the table, it shows the results of the server (1000) performing the Figure 6 shown calculation process on multiple translators stored in the database (1900), and includes the number of relevant translation files of each translator. In the candidate translator selection step (S500), the server (1000) compares the number of relevant translation files of multiple translators stored in the database (1900) with a preset number. When the number of relevant translation files of a certain translator is greater than the preset number, this translator is selected as a candidate translator.

[0096] In one implementation manner, the preset number is 3. As Figure 7 shown, if the number of relevant translation files of Translator A is 1, the number of relevant translation files of Translator B is 6, the number of relevant translation files of Translator C is 4, the number of relevant translation files of Translator D is 0, and the number of relevant translation files of Translator E is 3, then the server (1000) can screen out Translator B, Translator C, and Translator E as candidate translators from Translator A to Translator E registered in the database (1900).

[0097] Figure 8 Describes one implementation manner of the present invention, the process of calculating the translation unit price of each translator based on a preset formula.

[0098] As Figure 8As shown, a method for determining interpreters and translation prices includes calculating the translation unit price of each candidate interpreter (according to [Formula 1] below), and selecting the interpreter with the lowest translation unit price among the candidate interpreters as the final interpreter in the final interpreter selection stage (S600).

[0099] [Formula 1]:

[0100] The translation unit price of the candidate interpreter = CostLOW + {(CostHIGH - CostLow) * (NumberTOTAL - Number) / NumberTOTAL}

[0101] (where CostLOW is the lowest unit price, CostHIGH is the highest unit price, NumberTOTAL is the total number of relevant translation files of all candidate interpreters, and Number is the number of relevant translation files of this candidate interpreter)

[0102] Specifically, based on Figure 3 the lowest unit price and the highest unit price of the candidate interpreters received by the translation unit price selection layer (L3) as shown, and the number of relevant translation files of each candidate interpreter, the server (1000) automatically calculates the translation unit price of this candidate interpreter using the following [Formula 1], thus providing convenience for the interpreters using the server (1000), which is one of the technical features of the present invention.

[0103] Formula 1

[0104]

[0105] In the above [Formula 1], CostLOW represents the lowest price input by the candidate interpreter through the interpreter terminal (2000), CostHIGH represents the highest price input by the candidate interpreter through the interpreter terminal (2000), NumberTOTAL represents the total number of relevant translation files of all candidate interpreters respectively, and Number represents the number of relevant translation files of this candidate interpreter.

[0106] In addition, the technical feature of the present invention is that for an interpreter with more relevant translation files having a higher similarity to the feature vector, the translation difficulty is relatively lower when processing the original text file submitted by the user, and the translation unit price is calculated accordingly. Therefore, the server (1000) will make the translation unit price of the candidate interpreter with a larger number of relevant translation files closer to the lowest price input by this interpreter, while the translation unit price of the candidate interpreter with a smaller number of relevant translation files is closer to the highest price input by this interpreter.

[0107] Figure 8 describes the execution process of calculating the translation unit price for each of the selected multiple candidate interpreters in the implementation manner as shown. As Figure 7 shownFigure 8 As shown, in this implementation, the lowest price input by Translator B through the translation unit price selection layer (L3) is 40 yuan, and the highest price is 50 yuan; the lowest price input by Translator C through the translation unit price selection layer (L3) is 15 yuan, and the highest price is 40 yuan; the lowest price input by Translator E through the translation unit price selection layer (L3) is 25 yuan, and the highest price is 50 yuan. Based on the above [Formula 1], the calculation is as follows:

[0108] The translation unit price of Translator B = 40 + (50 - 40) * {(6 + 4 + 3) - 6} / 6 + 4 + 3 = 40 + 10 * 7 / 13 ≒ 45

[0109] The translation unit price of Translator C = 15 + (40 - 15) * {(6 + 4 + 3) - 4} / 6 + 4 + 3 = 40 + 10 * 7 / 13 ≒ 32

[0110] The translation unit price of Translator E = 25 + (50 - 25) * {(6 + 4 + 3) - 3} / 6 + 4 + 3 = 40 + 10 * 7 / 13 ≒ 44

[0111] The results are as shown above.

[0112] The above calculation shows that the server (1000) can select Translator C with the lowest calculated translation unit price as the final translator from among multiple candidate translators, namely Translator B, Translator C, and Translator E. In other words, the present invention comprehensively considers the translation information input by the translator himself and the expected difficulty of the commissioned translation, thereby selecting the final translator. In this way, the server (1000) can provide higher satisfaction for both the user and the translator.

[0113] Figure 9 Describes the process of calculating the first translation price, the first translation deadline, the second translation price, and the second translation deadline in an implementation manner of the present invention.

[0114] As Figure 9 shown, the present invention provides a method for determining a translator and a translation price, including the following steps: calculating a first translation price based on the translation unit price of the selected final translator and the number of words in the original text file, and calculating a first translation deadline based on the number of words; then, according to a preset rule, while shortening the first translation deadline, increasing the first translation unit price, thereby calculating at least one second translation price and a second translation deadline, and sending information on the first translation price, the first translation deadline, at least one second translation price, and at least one second translation deadline to the user terminal (3000) (translation price sending step S700).

[0115] In addition, the above-mentioned preset rules stipulate that at least one second translation deadline is derived from the first translation deadline based on a fixed unit deadline within a preset deadline range. In addition, at least one second translation price is inversely derived from the above-mentioned second translation deadline, but it should be maintained within the range of a preset minimum lower limit price and a maximum upper limit price. Among them, the preset minimum lower limit price is lower than the first translation price and not less than zero; the preset maximum upper limit price is higher than the first translation price and is determined according to the shortest deadline among the above-mentioned at least one second translation deadline.

[0116] Generally speaking, Figure 9 (a) describes the derivation process of the first translation price and the first translation deadline, while Figure 9 (b) describes the derivation process of the second translation price and the second translation deadline.

[0117] Specifically, as Figure 9 (a) shows, the server (1000) derives the first translation price by calculating the product of the translation unit price of the selected final translator and the number of words in the original text file, and derives the first translation deadline based on the number of words in the original text file. More specifically, the first translation deadline can be directly proportional to the number of words in the original text file, but in order to consider the translator's schedule, the system presets a minimum lower limit deadline, so the first translation deadline should not be lower than this minimum lower limit deadline.

[0118] In addition, for the translation task of the original text file, the shorter the translation deadline, the higher the translation price; the longer the translation deadline, the lower the translation price. Therefore, the server (1000) can, based on the first translation price and the first translation deadline, and in accordance with the preset rules, derive at least one second translation price and at least one second translation deadline according to different changes in the translation deadline.

[0119] More specifically, the preset rules stipulate that at least one second translation deadline is derived from the first translation deadline based on a fixed unit deadline within a preset deadline range. In addition, at least one second translation price is inversely derived from the above-mentioned second translation deadline, but it should be maintained within the range of a preset minimum lower limit price and a maximum upper limit price. Among them, the preset minimum lower limit price is lower than the first translation price and not less than 0; and in order to prevent the translation price from being too high and bringing a burden to the user, the preset maximum upper limit price is higher than the first translation price and is determined according to the shortest deadline among at least one second translation deadline. In other words, as Figure 9 (b) shows, within the preset deadline range, the second translation price should be maintained within the fixed range of the maximum upper limit price.

[0120] Ideally, as Figure 9In the graph shown in (b), the second translation deadline changes continuously. However, in order for users to more intuitively perceive the impact of changes in the translation deadline on the translation price, the server (1000) derives at least one second translation deadline based on a fixed unit deadline within a preset deadline range.

[0121] Figure 10 Briefly describes the interface of the application program executed on the user terminal (3000) for selecting the translation price and translation deadline in one implementation of the present invention.

[0122] Briefly speaking, Figure 10 (a) of describes one implementation of the present invention, that is, the interface for selecting the translation price and translation deadline for entrusting translation displayed on the user terminal (3000), while Figure 10 (b) of describes another implementation of the present invention, that is, the interface for selecting the translation job speed for entrusting translation displayed on the user terminal (3000).

[0123] Specifically, after the user submits a translation order request through the order layer (L9) shown in (b) of Figure 4 , in one implementation of the present invention, as shown in (a) of Figure 10 , the order information layer (L10), the translation deadline setting layer (L11), and the payment amount layer (L13) can be displayed on the screen of the user terminal (3000).

[0124] The user can confirm the information selected and input by himself in the second translation language selection layer (L6) and the second translation category selection layer (L7) shown in (b) of Figure 4 through the above-mentioned order information layer (L10), and can view the number of characters automatically measured by the server (1000) in the original text file uploaded through the original text file upload layer (L8).

[0125] The user can select and input his desired translation deadline from one or more translation deadlines generated by the server (1000) through the above-mentioned translation deadline setting layer (L11). In one implementation of the present invention, the user can select a translation deadline from D-Day (the first translation deadline) or D-3, D-2, D+1, etc. (one or more second translation deadlines) through the scrolling element included in the translation deadline setting layer (L11), and in the translation deadline setting layer (L11), the translation delivery date determined according to the translation deadline selected by the user can be recorded.

[0126] The user can confirm the first translation price or the second translation price calculated based on the translation deadline determined in the translation deadline setting layer (L11) through the above-mentioned payment amount layer (L13). Since the translation price displayed in the payment amount layer (L13) will change in real time when the user selects and enters the translation deadline through the payment amount layer (L13), the user can refer to the translation price displayed in the payment amount layer (L13) to determine an appropriate translation deadline in the translation deadline setting layer (L11).

[0127] In addition, after the user submits a translation order request through the order layer (L9) shown in Figure 4 (b), in another implementation manner of the present invention, an order information layer (L10), an urgent selection layer (L12), and a payment amount layer (L13) as shown in Figure 10 (b) can be displayed on the screen of the user terminal (3000).

[0128] In another implementation manner of the present invention, the server (1000) provides the user with an opportunity to select a first translation price and a first translation deadline, or to select a second translation price and a second translation deadline. The above-mentioned urgent selection layer (L12) includes: a standard layer that displays the translation delivery date based on the first translation price and the first translation deadline; an urgent layer that displays the translation delivery date based on the second translation price and the second translation deadline; and a display element that displays the layer selected by the user. The second translation deadline included in the urgent layer is a deadline advanced compared to the first translation deadline, and the second translation price included in the urgent layer is an amount that changes according to the first translation price and the second translation deadline. The user can select one of the standard layer and the urgent layer and make a payment.

[0129] In addition, as shown in Figure 4 (b) and Figure 10 (b), in another implementation manner of the present invention, when the user pays for the translation order, the user can make the final payment after enjoying the coupons and membership discounts provided by the server (1000).

[0130] That is to say, through the Figures 1 to 10 series of execution processes described above, the present invention calculates the reasonable fees and working hours required for the translation task entrusted by the user, and selects a translator who can provide a highly reliable translation result, which is a technical feature of the present invention.

[0131] Figure 11 Briefly described is the application program interface running on the user terminal (3000) for converting the template format of the original document in one implementation manner of the present invention.

[0132] When a user wishes to submit a paper belonging to the "academic" category to a university or institution, it is usually necessary to modify and submit it in the format required by that university or institution. However, when the user needs to adjust the paper with the same content according to different formats, it may be inconvenient and even lead to a series of problems such as delays in the research progress.

[0133] Therefore, the technical feature of the present invention lies in converting various template formats included in the original document. The above template format refers to the basic structure and design framework used when writing a document. The server (1000) can convert the user's original document according to the template format required by the university or institution that the user plans to submit to.

[0134] As Figure 11 shown, in an implementation manner of the present invention, the application program interface running on the user terminal (3000), in addition to including the second translation language selection layer (L6), the second translation category selection layer (L7), the original document upload layer (L8) and the order layer (L9) as shown in Figure 4 (b), can also additionally display a template file upload layer (L14).

[0135] The user can upload a template file containing multiple sub-template formats through the template file upload layer (L14) of the user terminal (3000). That is to say, the server (1000) converts the original template format adopted by the original document into the sub-template formats included in this template file.

[0136] On the other hand, even if the template file upload layer is already displayed on the interface of the user terminal (3000), but when the user does not upload a template file, the system can send a notification to the user terminal (3000) to confirm whether to upload. If the user still does not upload a template file, the server (1000) will not convert the template format of the original document, but directly send the original document to the selected final translator.

[0137] In addition, as Figure 11 shown, in an implementation manner of the present invention, the upload capacity of the template file is limited to within 20MB. However, in other implementation manners, the upload capacity may not be limited to within 20MB.

[0138] As Figure 12 shown, an implementation manner of the present invention outlines the process of converting the template format of the original document.

[0139] As Figure 12As shown, the method for determining a translator and translation price further includes a template format conversion step of converting the template format of the original document, and the template format conversion step includes: a template file receiving step for receiving a template file containing a plurality of target template formats according to a user input of a user terminal (3000); a target template format detecting step of detecting the plurality of target template formats from the template file by using a trained artificial neural network model; an original document template format detecting step of detecting a plurality of original document template formats from the original document by using the detecting model; and an original document conversion step of converting the plurality of original document template formats included in the original document into the plurality of target template formats, and inputting the detailed data input into the plurality of original document template formats into the corresponding target template formats.

[0140] In addition, the original document conversion step further includes: for an original document template format that does not correspond to any target template format, deleting the detailed data input into this original document template format according to the user input received by the user terminal (3000), or inserting the detailed data of this original document template format into the detailed data in the target template format based on a template format matching table in which the original document template format matches the target template format; and a notification reminder step of providing a notification to the user terminal (3000) about the lack of this target template format for a target template format that does not correspond to the original document template format.

[0141] The template file receiving step is executed by a server (1000), and receives a template file containing a plurality of target template formats from the user terminal (3000) according to a user input. Preferably, the template file receiving step can be executed simultaneously with the original document receiving step (S200).

[0142] The target template format detecting step is executed by a server (1000), and detects a plurality of target template formats from the template file by using a trained artificial neural network model. In addition, the original document template format detecting step is executed by a server (1000), and detects a plurality of original document template formats from the original document by using the detecting model. As Figure 12 shown, in one implementation manner of the present invention for the target template format and the original document template format, the plurality of target template formats include a title, author information, a table of contents, an introduction, a body text, a conclusion, and references, while the plurality of original document template formats include a title, author information, a table of contents, an abstract (Abstract, summary), an introduction, a body text, a conclusion, and an appendix.

[0143] As Figure 12 shown, the original document conversion step converts the plurality of original document template formats into the plurality of target template formats, and inputs the detailed data input into the plurality of original document template formats into the corresponding target template formats. In other words, as Figure 12In an implementation shown, for parts that exactly match the target template format, such as the title, author information, table of contents, introduction, body, and conclusion, the detailed data can be directly input.

[0144] However, for the original file template formats such as "abstract" or "appendix" that do not correspond to any target template format, since these are not essential elements of the document format required by the user, the server (1000) can, according to the user input received from the user terminal (3000), delete the detailed data contained in the "appendix", or insert the detailed data in the "abstract" into the target template format preset to correspond to the original file template format (such as Figure 12 in, corresponding to "conclusion").

[0145] In addition, since target template formats such as "references" that do not correspond to any original file template format are essential elements of the document format required by the user, the server (1000) provides a notice of the lack of this target template format to the user terminal (3000) through a notification reminder step.

[0146] Figure 13 Briefly describes the template format matching table in an implementation of the present invention.

[0147] As Figure 13 shown, the template format matching table contains multiple original template formats and multiple target template formats, and one or more preset target template formats are matched for each original template format. When an original template format matches multiple target template formats, the matching table also contains the preset priorities for these target template formats. In the detailed data insertion stage, the system inserts the detailed data into the corresponding target template formats according to these priorities.

[0148] Specifically, in the Figure 12 implementation shown, for an original template format such as "abstract" that cannot be matched to a target template format, the server (1000) inserts its detailed data into the preset target template format. However, in some cases, this preset target template format may not be included in the list of multiple target template formats. Therefore, the server (1000) generates a template format matching table containing multiple original template formats, multiple target template formats, and their priorities, and inserts the detailed data of the original template format into the appropriate target template format based on this table.

[0149] As Figure 13The table shown is the template format matching table in the implementation of the present invention. Among them, multiple target template formats that match the original text template format "Abstract" include "Introduction", "Body", and "Conclusion", and a priority is preset for each target template format. That is to say, the server (1000) will first insert the detailed data corresponding to "Abstract" into the target template format corresponding to "Conclusion". However, if the target template format corresponding to "Conclusion" is not included in the template file uploaded by the user, this detailed data will be inserted into the target template format corresponding to the next-priority "Body".

[0150] When there is no target template format in the template format matching table that matches the original text template format, the server (1000) can directly delete the detailed data of this original text template format without the user input.

[0151] In one implementation of the present invention, by using a feature extraction model based on a trained artificial neural network, the feature vectors of the original text file submitted by the user are compared with the feature vectors of the translated text file completed by the translator, and candidate translators are screened according to the similarity, thereby improving the reliability of the translation result and enhancing the user's satisfaction.

[0152] In one implementation of the present invention, relevant translated text files with feature vector similarity reaching a preset value or more are screened out from multiple translated text files, and candidate translators are selected based on this, thereby improving the reliability of the translation result and enhancing the user's satisfaction.

[0153] In one implementation of the present invention, by using a specific formula with the lowest price and the highest price input by the translator in the server as variables, the final translator is selected from multiple candidate translators, thereby improving the convenience of the translator using this server.

[0154] In one implementation of the present invention, for the translation price and translation cycle calculated for the final translator, additional translation prices and translation cycles are further derived based on preset rules, thereby providing the user with multiple choice opportunities in terms of the cost and time of commissioning translation.

[0155] In one implementation of the present invention, when the document format required by the submitting agency is different from the format of the original text file, by providing a template file that meets the format requirements of this agency, the format of the original text file is automatically converted into the corresponding format, thereby improving the convenience of the user.

[0156] As described above, although the implementation of the present invention has been illustrated by specific examples and drawings, those of ordinary skill in the art can make various modifications and variations based on the above. For example, the described techniques can be executed in a method sequence different from that described, and / or the components of the system, structure, device, circuit, etc. can be combined or combined in a method different from that described, or replaced by other components or equivalents, so as to achieve appropriate results. Therefore, different implementations, different implementations, and equivalents of the following claims all fall within the scope of the following claims.

Claims

1. A method for determining a translator and a translation price for a commissioned translation executed on a server associated with an application running on a user terminal and a translator terminal, comprising: Translator information receiving step: receiving translator information including translation language type, translation category, and translation unit price range information charged by word including minimum price and maximum price from the translator terminal; The original document receiving step: receiving the translation language type, translation category and the original document to be translated from the user terminal according to the user input; Feature vector extraction step: using a feature extraction model based on an artificial neural network, the model has been trained to extract main keywords and related information of the entire document text from the file containing the document as a feature vector, extracting a first feature vector from the original document; Similarity calculation step: for multiple translators who meet the above translation language types and translation categories, extract multiple existing translation files corresponding to them from the database, extract the second feature vector of each translation file through the above feature extraction model, and calculate the similarity between these second feature vectors and the first feature vector; Candidate translator screening step: identifying translation files whose similarity reaches a preset threshold as relevant translation files, and screening each translator as a candidate translator when the number of relevant translation files of the translator reaches a preset standard; Final translator selection step: Calculate the translation unit price of each candidate translator according to the following [Formula 1], and select the candidate translator with the lowest translation unit price as the final translator; The translation price sending step: calculating a first translation price based on the translation unit price of the selected final translator and the number of words in the original document, calculating a first translation period based on the number of words, and increasing the first translation price while reducing the first translation period according to a preset rule, thereby deriving one or more second translation prices and a second translation period, and sending information of the first translation price, the first translation period, the one or more second translation prices and the second translation period to the user terminal; Original file sending step: according to the user's payment situation, the original file and the translation period information calculated based on the user's payment information are sent to the translator terminal; The translation unit price of the above candidate translators is based on the number of relevant translation documents of the candidate translators, the estimated difficulty of the entrusted translation and part of the translator information input by the above candidate translators, and is calculated according to the following [Formula 1]; The above method for determining translators and translation prices further includes a template format conversion step, wherein: Template file receiving step: receiving a template file including a plurality of sub-template formats from a user terminal according to user input; Sub-template format detection step: detecting multiple sub-template formats from the template file through a trained artificial neural network-based detection model; Original template format detection step: using the above detection model, multiple original template formats are detected from the original document; The original document conversion step: converting the multiple original document template formats contained in the original document into the multiple sub-document template formats, and inputting the detailed data input into the original document template formats into the corresponding sub-document template formats; the above is a method for determining the translator and the translation price; [Formula 1]: The translation unit price of the candidate translator = CostLOW + (CostHIGH-CostLow)*(NumberTOTAL-Number) / NumberTOTAL Among them, CostLOW is the above-mentioned lowest price, CostHIGH is the above-mentioned highest price, NumberTOTAL is the sum of the number of relevant translation files of all candidate translators, and Number is the number of relevant translation files of the candidate translator.

2. The method according to claim 1, characterized in that The above preset rules include: The one or more second translation deadlines are derived based on fixed unit time within a preset deadline range relative to the first translation deadline; The one or more second translation prices are derived in inverse proportion to the second translation period, and are limited to a preset minimum lower limit price and a preset maximum upper limit price; The above preset minimum price is lower than the first translation price and is not less than 0; The preset maximum upper limit price is higher than the first translation price and is determined based on the method for determining the translator and the translation price according to the shortest second translation period among the one or more second translation periods.

3. The method according to claim 1, characterized in that: The above translation files: Files that the translator has translated from other original documents; A method for determining translators and translation prices for texts in the same language as the original documents.

4. The method according to claim 1, characterized in that: The above original document conversion steps include: For an original template format that cannot be matched to a sub-template format, according to a user input received from a user terminal, the detailed data input to the original template format is deleted, or based on a template format matching table that matches the sub-template format according to the original template format, the detailed data of the original template format is inserted into the detailed data of the sub-template format; For the sub-template format that cannot correspond to the original template format, a notification reminder step of providing the user terminal with relevant notification of the sub-template format; the above constitutes a part of the method for determining the translator and the translation price.

5. The method according to claim 4, characterized in that: The above template format matching table includes: Contains multiple original template formats and multiple sub-template formats; Pre-matching at least one sub-template format for each of a plurality of original template formats; When at least one sub-template format matches a certain original template format, it further includes a preset priority for the at least one sub-template format; The detailed data inserting step inserts the detailed data into the sub-template format based on the preset priority. The above constitutes a part of the method for determining the translator and the translation price.

6. A service-side server associated with an application running on a user terminal and a translator terminal and executing a method for determining a translator and a translation price for a commissioned translation, comprising: The translator information receiving unit receives translator information including the translation language type, translation category, and the minimum price and maximum price range of each word translation unit price from the translator terminal; Original document receiving unit: receiving the translation language type, translation category and original document to be translated according to the user input in the user terminal; Feature vector extraction unit: extracting main keywords and related information of the entire text as feature vectors from the file containing the document through a trained feature extraction model based on an artificial neural network, and extracting a first feature vector from the original document; A similarity calculation unit: for multiple translators who meet the translation language type and translation category, extracting second feature vectors from multiple translation files of each translator stored in the database, and calculating the similarity between these second feature vectors and the first feature vector; Candidate translator screening unit: identifies translation files with similarity exceeding a preset value as related translation files, and screens each translator as a candidate translator when the number of related translation files of each translator reaches a preset number or more; Final translator selection department: calculate the translation unit price of each candidate translator according to the following [Formula 1], and select the candidate translator with the lowest translation unit price as the final translator; A translation price sending unit: calculating a first translation price based on the translation unit price of the final selected translator and the number of words in the original document, and calculating a first translation period based on the number of words; Then, according to a preset rule, the first translation period is reduced while the first translation price is increased, one or more second translation prices and a second translation period are derived, and information about the first translation price, the first translation period, the one or more second translation prices and the second translation period is sent to a user terminal; Original document sending unit: according to the user's payment situation, sends the original document and the translation deadline information calculated based on the user's payment information to the translator's terminal; The translation unit price of each candidate translator is calculated based on the partial translator information input by the candidate translator and the number of relevant translation documents, combined with the estimated difficulty of the entrusted translation, and in accordance with the following [Formula 1]; The above method for determining translators and translation prices includes the steps of converting the original document into a template format, wherein: Steps to receive template files: Receiving a template file including a plurality of sub-template formats according to an input of a user at a user terminal; Sub-template format detection step: detecting multiple sub-template formats from the template file through a trained artificial neural network-based detection model; Original template format detection step: using the detection model to detect multiple original template formats from the original document; The original document conversion step: converting the multiple original document template formats in the original document into multiple sub-template formats, and inputting the detailed data input into the original document template format into the corresponding sub-template format. The above constitutes the characteristics of the server at the service end; [Formula 1]: The translation unit price of the candidate translator = CostLOW + (CostHIGH-CostLow)*(NumberTOTAL-Number) / NumberTOTAL Among them, CostLOW is the above-mentioned lowest price, CostHIGH is the above-mentioned highest price, NumberTOTAL is the sum of the number of relevant translation files of all candidate translators, and Number is the number of relevant translation files of the candidate translator.

Images