KR102731381B1 - The Method and Service Server That Determine The Translator and Translation Price for A Translation Request - Google Patents

Abstract

The present invention relates to a method for determining a translator and a translation price for a translation request, and a service server for doing so, which receives an original text file input by a user, calculates a similarity between a plurality of translation files of each of a plurality of translators and a feature vector of each of the original text files, identifies a related translation file among the plurality of translation files based on the similarity, selects a plurality of candidate translators based on the number of related translation files, selects a final translator based on a translation unit price for each translator calculated based on a minimum price and a maximum price input by the translator and the number of related translation files, transmits to the user a translation price and translation period that change according to a preset rule, and transmits the original text file to the translator depending on whether the user has made payment.

Description

{The Method and Service Server That Determine The Translator and Translation Price for A Translation Request}

The present invention relates to a method for determining a translator and a translation price for a translation request, and a service server for doing so, which receives an original text file input by a user, calculates a similarity between a plurality of translation files of each of a plurality of translators and a feature vector of each of the original text files, identifies a related translation file among the plurality of translation files based on the similarity, selects a plurality of candidate translators based on the number of related translation files, selects a final translator based on a translation unit price for each translator calculated based on a minimum price and a maximum price input by the translator and the number of related translation files, transmits to the user a translation price and translation period that change according to a preset rule, and transmits the original text file to the translator depending on whether the user has made payment.

As the global market expands, multilingual communication is required for modern people. In particular, high expertise in multilingual communication is required in the business field that sells products or services internationally or in the academic field that shares information internationally. However, due to high labor costs for professional translators, companies and organizations are minimizing labor costs by using translation services through outsourcing instead of directly hiring in-house interpreters and translators. However, due to high demand, they are indiscriminately entrusting translation requests without understanding the expertise of translators, which is lowering the reliability of translation results.

Conventional inventions related to translation brokerage services include translation brokerage systems and methods, such as Korean Patent Publication No. 10-2021-0050207. While the invention recommends a translator who satisfies preset conditions among a number of translators who are judged to have the same specialty as the genre of the translation project for which translation work has been requested to the applicant, it does not suggest a description of the cost or working time for the translation project requested by the applicant. Meanwhile, in the brokerage service that connects the applicant and the translator as described above, a high level of trust between the applicant and the translator is required, and accordingly, it is desirable to set a reasonable cost and working time that can satisfy both the applicant and the translator.

Therefore, there is a need to develop a technology that can select a translator who can provide a reasonable cost and work time for the translation request requested by the client.

Republic of Korea Publication Patent No. 10-2021-0050207 (May 7, 2021)

The present invention provides a method for determining a translator and a translation price for a translation request, and a service server for doing so, which receives an original text file input by a user, calculates a similarity between a plurality of translation files of each of a plurality of translators and a feature vector of each of the original text files, identifies a related translation file among the plurality of translation files based on the similarity, selects a plurality of candidate translators based on the number of related translation files, selects a final translator based on a translation unit price for each translator calculated based on a minimum price and a maximum price input by the translator and the number of related translation files, transmits to the user a translation price and translation period that change according to a preset rule, and sends the original text file to the translator depending on whether the user has made payment.

In order to solve the above-mentioned problem, in one embodiment of the present invention, a method for determining a translator and a translation price for a translation request performed in a service server associated with an application performed in a user terminal and a translator terminal, the method comprising: a translator information receiving step for receiving translator information including a translation language type, a translation category, and per-word translation price range information including a minimum price and a maximum price from a translator terminal; an original text file receiving step for receiving, from the user terminal, a translation language type, a translation category, and an original text file to be translated according to a user's input; a feature vector extraction step for extracting a first feature vector from the original text file through an artificial neural network-based feature extraction model learned to extract information related to main keywords of the entire text of a document as a feature vector from a file including the document; a similarity calculation step for calculating a similarity between each of second feature vectors extracted through the feature extraction model from a plurality of translation files of the corresponding translator pre-stored in a database and the first feature vector, for each of a plurality of translators corresponding to the translation language type and the translation category; A candidate translator selection step in which a translation file with a similarity level greater than or equal to a preset value is identified as a related translation file, and in which case the number of related translation files of each translator is greater than or equal to a preset number, the translator is selected as a candidate translator; a final translator selection step in which the translation unit price of each candidate translator is calculated using the following [Formula 1], and the candidate translator with the lowest translation unit price among the candidate translators is selected as the final translator; The present invention provides a method for determining a translator and a translation price, including: a translation price transmission step for calculating a first translation price based on a translation unit price of a selected final translator and the number of words in the original text file, calculating a first translation period based on the number of words, and increasing the first translation price while decreasing the first translation period according to a preset rule to derive one or more second translation prices and one or more second translation periods, and transmitting information about the first translation price, the first translation period, the one or more second translation prices, and the one or more second translation periods to the user terminal; and an original text file transmission step for transmitting information about the translation period based on the original text file and the user's payment information to the translator terminal based on the result of confirmation of whether the user has made a payment.

[Formula 1]:

Translation cost of candidate translator = CostLOW+(CostHIGH-CostLow)*(NumberTOTAL-Number)/NumberTOTAL

(Here, CostLOW is the minimum price above, CostHIGH is the maximum price above, NumberTOTAL is the sum of the number of related translation files for each candidate translator, and Number is the number of related translation files for that candidate translator.)

In one embodiment of the present invention, the preset rule may be such that each of the one or more second translation periods is derived based on a certain unit period within a preset period range with respect to the first translation period, and each of the one or more second translation prices is derived inversely proportional to the second translation period, but within a range of a preset minimum lower price to a preset maximum upper price, and the preset minimum lower price is lower than the first translation price and is a value equal to or greater than 0, and the preset maximum upper price is higher than the first translation price and may be determined based on a second translation period having a lowest price among the one or more second translation periods.

In one embodiment of the present invention, the translation file corresponds to a file in which the translator has completed translation of another original text file, and may include text composed in the same language as the original text file.

In one embodiment of the present invention, the method for determining the translator and the translation price may further include a template format conversion step for converting the template format of the original text file, and the template format conversion step may include: a template file receiving step for receiving a template file including a plurality of source template formats from a user terminal according to a user input; a source template format detection step for detecting the plurality of source template formats from the template file through a learned artificial neural network-based detection model; an original text template format detection step for detecting the plurality of original text template formats from the original text file through the detection model; and an original text file conversion step for converting the plurality of original text template formats included in the original text file into the plurality of source template formats and inputting detailed data input into each of the plurality of original text template formats into a source template format corresponding to the corresponding original text template format.

In one embodiment of the present invention, the original text file conversion step may include a detailed data insertion step of deleting detailed data entered in the original text template format according to a user input received from the user terminal for an original text template format that does not correspond to the source template format, or inserting detailed data of the original text template into the source template format based on a template format matching table in which source template formats are matched to each original text template format; and a notification provision step of providing a notification of the source template format to the user terminal for the source template format that does not correspond to the original text template format.

In one embodiment of the present invention, the template format matching table includes a plurality of original template formats and a plurality of source template formats, and matches at least one preset source template format to each of the plurality of original template formats, and further includes a preset priority for each of the at least one source template format when at least one source template format is matched to the original template format, and the detailed data insertion step can insert the detailed data into the source template format based on the preset priority.

In order to solve the above-mentioned problem, in one embodiment of the present invention, a service server which is associated with an application executed in a user terminal and a translator terminal and which performs a method of determining a translator and a translation price for a translation request, comprises: a translator information receiving unit which receives translator information including a translation language type, a translation category, and a translation price range per word including a minimum price and a maximum price from a translator terminal; an original text file receiving unit which receives, from the user terminal, a translation language type, a translation category, and an original text file to be translated according to a user's input; a feature vector extracting unit which extracts a first feature vector from the original text file through an artificial neural network-based feature extraction model which is trained to extract information related to main keywords of the entire text of a document as a feature vector from a file including the document; a similarity calculating unit which calculates a similarity between each of second feature vectors extracted through the feature extraction model from a plurality of translation files of a corresponding translator pre-stored in a database and the first feature vector, for each of a plurality of translators corresponding to the translation language type and the translation category; A candidate translator selection unit that identifies translation files whose similarity is greater than or equal to a preset value as related translation files, and selects each translator as a candidate translator when the number of related translation files of each translator is greater than or equal to the preset number; a final translator selection unit that calculates the translation unit price of each candidate translator by the following [Formula 1], and selects the candidate translator with the lowest translation unit price among the candidate translators as the final translator; a translation price transmission unit that calculates 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, calculates a first translation period based on the number of words, and derives one or more second translation prices and second translation periods by decreasing the first translation period according to a preset rule and increasing the first translation price, and transmits information about the first translation price, the first translation period, the one or more second translation prices, and the one or more second translation periods to the user terminal; A service server is provided, including an original file transmission unit that transmits the original file and information about the translation period based on the user's payment information to the translator terminal according to the result of verification of whether the user has made payment.

[Formula 1]:

Translation cost of candidate translator = CostLOW+(CostHIGH-CostLow)*(NumberTOTAL-Number)/NumberTOTAL

(Here, CostLOW is the minimum price above, CostHIGH is the maximum price above, NumberTOTAL is the sum of the number of related translation files for each candidate translator, and Number is the number of related translation files for that candidate translator.)

In one embodiment of the present invention, by comparing the feature vector of an original file for which a user has requested translation with the feature vector of a translation file performed by a translator through a learned artificial neural network-based feature extraction model and selecting a candidate translator based on similarity, the reliability of the translation result can be improved and the effect of providing high satisfaction to the user can be achieved.

In one embodiment of the present invention, by identifying a related translation file among a plurality of translation files whose feature vector similarity is greater than a preset value and selecting a candidate translator based on the similarity, the reliability of the translation result can be improved and a high level of satisfaction can be provided to the user.

In one embodiment of the present invention, by selecting a final translator among a plurality of candidate translators through a specific formula that uses as variables the minimum price and maximum price input by the translator to the service server, the effect of providing high convenience to the translator using the service server can be achieved.

In one embodiment of the present invention, by deriving additional translation prices and translation periods based on preset rules for the translation prices and translation periods calculated for the final translator, the user can have the effect of providing an opportunity to select the option he or she wants in terms of cost and time for a translation request.

In one embodiment of the present invention, when the format of a document handled by an institution requiring submission is different from the format of the original file, the format of the original file is automatically converted into the corresponding format through a template file composed of the format handled by the institution, thereby providing convenience to the user.

FIG. 1 illustrates components of a service server that performs a method for determining a translator and a translation price according to one embodiment of the present invention.
Figure 2 schematically illustrates the execution steps of a method for determining a translator and a translation price according to one embodiment of the present invention.
FIG. 3 schematically illustrates a screen of an application executed on a translator terminal for registering translator information according to one embodiment of the present invention.
FIG. 4 schematically illustrates a screen of an application executed on a user terminal for a user's translation request according to one embodiment of the present invention.
FIG. 5 schematically illustrates a process of extracting a first feature vector and a second feature vector according to one embodiment of the present invention.
FIG. 6 schematically illustrates a process of identifying a translator's related translation file based on the similarity of each translation file according to one embodiment of the present invention.
Figure 7 schematically illustrates a process for selecting multiple candidate translators according to one embodiment of the present invention.
Figure 8 schematically illustrates a process of deriving a translation unit price per translator based on a preset formula according to one embodiment of the present invention.
FIG. 9 schematically illustrates a process for deriving a first translation price, a first translation period, a second translation price, and a second translation period according to one embodiment of the present invention.
FIG. 10 schematically illustrates a screen of an application executed on a user terminal to select a translation price and translation date according to one embodiment of the present invention.
FIG. 11 schematically illustrates a screen of an application executed on a user terminal for converting a template format of an original file according to one embodiment of the present invention.
Figure 12 schematically illustrates a process of converting a template format of an original file according to one embodiment of the present invention.
Figure 13 schematically illustrates a template format matching table according to one embodiment of the present invention.

Hereinafter, various embodiments and/or aspects are now disclosed with reference to the drawings. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a general understanding of one or more aspects. It will be recognized, however, by one skilled in the art that such aspect(s) may be practiced without these specific details. The following description and the annexed drawings set forth specific exemplary aspects of one or more aspects in detail. It should be understood, however, that these aspects are exemplary and that any of the various methods of the principles of the various aspects may be utilized, and the description is intended to encompass all such aspects and their equivalents.

Additionally, various aspects and features will be presented by means of systems that may include a number of devices, components, and/or modules, etc. It is also to be understood and appreciated that various systems may include additional devices, components, and/or modules, etc., and/or may not include all of the devices, components, modules, etc. discussed in connection with the drawings.

The terms "embodiment," "example," "aspect," and "example" as used herein are not to be construed as implying that any aspect or design described is better or advantageous over other aspects or designs. The terms "part," "component," "module," "system," and "interface," as used below, generally refer to computer-related entities, and can refer to, for example, hardware, a combination of hardware and software, or software.

Additionally, it should be understood that the terms "comprises" and/or "comprising" imply the presence of the features and/or components, but do not preclude the presence or addition of one or more other features, components and/or groups thereof.

Also, terms including ordinal numbers such as first, second, etc. may be used to describe various components, but the components are not limited by the terms. The terms are only used to distinguish one component from another. For example, without departing from the scope 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. The term and/or includes a combination of a plurality of related described items or any item among a plurality of related described items.

In addition, in the embodiments of the present invention, unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the present invention belongs. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning they have in the context of the relevant technology, and shall not be interpreted in an ideal or overly formal meaning unless explicitly defined in the embodiments of the present invention.

FIG. 1 illustrates components of a service server (1000) that performs a method for determining a translator and a translation price according to one embodiment of the present invention, and FIG. 2 schematically illustrates execution steps for a method for determining a translator and a translation price according to one embodiment of the present invention.

As illustrated in FIGS. 1 and 2, a method for determining a translator and a translation price for a translation request performed in a service server (1000) associated with an application performed in a user terminal (3000) and a translator terminal (2000) comprises: a translator information receiving step (S100) of receiving translator information including a translation language type, a translation category, and a translation price range per word including a minimum price and a maximum price from a translator terminal (2000); an original text file receiving step (S200) of receiving, from the user terminal (3000), a translation language type, a translation category, and an original text file to be translated according to a user's input; a feature vector extraction step (S300) of extracting a first feature vector from the original text file through a feature extraction model based on an artificial neural network learned to extract information related to main keywords of the entire text of a document from a file including the document as a feature vector; For each of a plurality of translators corresponding to the above translation language type and translation category, a similarity calculation step (S400) for calculating the similarity between each of the second feature vectors extracted through the feature extraction model from the plurality of translation files of the corresponding translator pre-stored in the database (1900) and the first feature vector; a candidate translator selection step (S500) for identifying translation files with the similarity greater than or equal to a preset value as related translation files, and selecting the corresponding translator as a candidate translator if the number of related translation files of each translator is greater than or equal to a preset number; a final translator selection step (S600) for calculating the translation unit price of each candidate translator by the following [Formula 1], and selecting the candidate translator with the lowest translation unit price among the candidate translators as the final translator; A method for determining a translator and a translation price, comprising: a translation price transmission step (S700) for 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, calculating a first translation period based on the number of words, and decreasing the first translation period according to preset rules while increasing the first translation price to derive one or more second translation prices and one or more second translation periods, and transmitting information on the first translation price, the first translation period, the one or more second translation prices, and the one or more second translation periods to the user terminal (3000); and an original text file transmission step (S800) for transmitting information on the translation period based on the original text file and the user's payment information to the translator terminal (2000) based on the result of confirmation of whether the user has made a payment.

[Formula 1]:

Translation cost of candidate translator = CostLOW+(CostHIGH-CostLow)*(NumberTOTAL-Number)/NumberTOTAL

(Here, CostLOW is the minimum price above, CostHIGH is the maximum price above, NumberTOTAL is the sum of the number of related translation files for each candidate translator, and Number is the number of related translation files for that candidate translator.)

Specifically, the method for determining a translator and a translation price of the present invention is performed in the service server (1000) illustrated in FIG. 1, and in one embodiment of the present invention, the service server (1000) is connected to a user terminal (3000) and a translator terminal (2000). The translator terminal (2000) displays an interface that visually displays a translation service through its screen, and a translator using the translator terminal (2000) can operate the interface by performing an input, such as a selection input, for the interface. Similarly, the user terminal (3000) displays an interface that visually displays a translation service through its screen, and a user using the user terminal (3000) can operate the interface by performing an input, such as a selection input, for the interface.

The above service server (1000) includes a translator information receiving unit (1100), an original text file receiving unit (1200), a feature vector extraction unit (1300), a similarity calculation unit (1400), a candidate translator selection unit (1500), a final translator selection unit (1600), a translation price transmission unit (1700), an original text file transmission unit (1800), and a database (1900) that stores a plurality of translation files of each of a plurality of translators.

The translator information receiving step (S100) is performed by the translator information receiving unit (1100) and receives translator information according to the translator's input from the translator terminal (2000). The translator information includes the types of translation languages that the translator can translate, the translation category corresponding to the translator's specialty, and the translator's translation price range information per word, and the translation price range information includes the minimum price and maximum price per word input by the translator.

The original text file reception step (S200) is performed by the original text file reception unit (1200), and receives the translation language type, translation category, and original text file uploaded according to the user's input from the user terminal (3000). That is, the service server (1000) performs a translation intermediary service that transmits the original text file to an appropriate translator, receives the translation file that the translator has completed from the translator terminal (2000), and then transmits it back to the user terminal (3000).

The feature vector extraction step (S300) is performed by the feature vector extraction unit (1300), and extracts a first feature vector from the original text file through an artificial neural network-based feature extraction model that is trained to extract information related to main keywords of the entire text of the document as feature vectors. The first feature vector includes the main characteristics of the original text file through the composition of the vector, and if the original text file includes a document corresponding to an academic category, the first feature vector extracted from the original text file can represent the academic technical field of the original text file.

The similarity calculation step (S400) is performed by the similarity calculation unit (1400), and derives a plurality of translators whose translation language type and translation category entered by the translator into the translator terminal (2000) match the translation language type and translation category entered by the user into the user terminal (3000), and extracts a second feature vector from a plurality of translation files of the corresponding translator stored in the database (1900) through the feature extraction model for the multiple translators, and calculates the similarity between the corresponding second feature vector and the corresponding first feature vector. Preferably, a translator who has a large number of second feature vectors whose similarity is calculated to be relatively high may be appropriate for performing a translation request for the corresponding original text file. Therefore, the service server (1000) can select a plurality of candidate translators through the following candidate translator selection step (S500).

The candidate translator selection step (S500) is performed by the candidate translator selection unit (1500), and identifies translation files with a similarity level greater than or equal to a preset value as related translation files, and if the number of related translation files for each translator is greater than or equal to a preset value, the translator is selected as a candidate translator. It is preferable that the preset value and the preset number are determined to a degree that allows the service server (1000) to select a plurality of candidate translators in order to select an appropriate candidate translator for the user's translation request.

The final translator selection step (S600) is performed by the final translator selection department (1600), and the translation unit price of each candidate translator is calculated by a preset formula, and the candidate translator with the lowest translation unit price among the candidate translators is selected as the final translator. More specifically, the preset formula uses the minimum price and maximum price input by the translator through the translator terminal (2000) and the number of related translation files derived through the candidate translator selection step (S500) as parameters.

The translation price transmission step (S700) is performed by the translation price transmission unit (1700), and calculates 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, calculates a first translation period based on the number of words, and derives one or more second translation prices and one or more second translation periods based on the first translation price, the first translation period, and preset rules, and transmits the results to the user terminal (3000).

Meanwhile, the number of words in the original text file can be measured by the service server (1000) through word tokenization, natural language processing, use of regular expressions, and stop word processing, and a detailed description of the preset rules will be described later in FIG. 9.

The original file transmission step (S800) is performed by the original file transmission unit (1800), and when the user performs payment by selecting and inputting the translation period and translation price into the user terminal (3000) for the first translation price, the first translation period, one or more second translation prices, and one or more second translation periods, the original file and information on the translation period based on the user's payment information are transmitted to the translator terminal (2000) based on the result of the confirmation of whether the user has made payment.

FIG. 3 schematically illustrates a screen of an application executed on a translator terminal (2000) for registering translator information according to one embodiment of the present invention.

As shown in Fig. 3, the method for determining a translator and a translation price includes a translator information receiving step (S100) for receiving translator information including translation language type, translation category, and translation price range information per word including a minimum price and a maximum price from a translator terminal (2000).

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

A translator can select a language that he or she can translate into and confirm the selected language through the first translation language selection layer (L1). In one embodiment of the present invention, the first translation language selection layer (L1) displays English, Chinese, and Korean selected by the translator.

A translator can select a category of documents that he or she can translate through the first translation category selection layer (L2) and check the selected category. In one embodiment of the present invention, the 'academic' category selected by the translator is displayed on the screen in the first translation category selection layer (L2).

A translator can set a translation price range by entering the minimum price and maximum price of the translation price (translation price per word) for his/her translation work through the above-mentioned translation price selection layer (L3). Preferably, as an embodiment of the present invention, the service server (1000) can limit the minimum price and maximum price of the translation price to be entered within different preset ranges.

In addition, as one embodiment of the present invention illustrated in FIG. 3, the editing layer (L4) may include a modification layer and a storage layer, and a translator may modify and store his or her translator information at any time through the editing layer (L4), and the service server (1000) stores the translator information for each of a plurality of translators in a database (1900) and selects the final translator based on the information.

FIG. 4 schematically illustrates a screen of an application executed on a user terminal (3000) for a user's translation request according to one embodiment of the present invention.

As shown in Fig. 4, the method for determining a translator and a translation price includes an original text file receiving step (S200) for receiving a translation language type, a translation category, and an original text file to be translated from a user terminal (3000) according to a user's input.

Schematically, Fig. 4 (a) illustrates a screen of an application that displays a primary selection layer (L5) through which a user can primarily select the type of language he or she wants to translate, as an embodiment of the present invention, and Fig. 4 (b) illustrates a screen of an application that outputs a plurality of layers through which a user can input a translation language type, a translation category, and an original text file, as an embodiment of the present invention.

Specifically, when a user executes an application connected to a service server (1000) through a user terminal (3000) and requests translation of a document, the user can first select the type of translation language. In addition, the user is not limited to selecting and inputting the type of language he or she wants to translate, and as an embodiment of the present invention, multiple layers can be included that allow selection of various contents such as document proofreading or customer center inquiries, such as the primary selection layer (L5) illustrated in (a) of Fig. 4.

When a user selects and inputs a layer corresponding to 'Chinese-Korean translation' within the primary selection layer (L5), the screen of the application can output a second translation language selection layer (L6), a second translation category selection layer (L7), an original text file upload layer (L8), and an order layer (L9) as shown in one example of (b) of FIG. 4.

The user can select a more specific translation language type from the language type selected in the primary selection layer (L5) through the second translation language selection layer (L6). Preferably, if the user selects a layer corresponding to 'Chinese-Korean translation' in the primary selection layer (L5), the second translation language selection layer (L6) shown in (b) of Fig. 4 may include 'Chinese-Korean translation' and 'Korean-Chinese translation', and if the user selects a layer corresponding to 'Chinese-English translation' in the primary selection layer (L5), the second translation language selection layer (L6) may include 'Chinese-English translation' and 'English-Chinese translation'. In addition, preferably, 'Chinese-Korean translation' means translating a document from Chinese to Korean, and 'Korean-Chinese translation' means translating a document from Korean to Chinese.

The user can select and input a translation category through the second translation category selection layer (L7), and in order to ensure the completeness of the translation result, it is preferable to select a category related to the content included in the original text file as the translation category. In one embodiment of the present invention, the translation categories are limited to 'daily life', 'standard', 'academic', and 'business', but in other embodiments, the second translation category selection layer (L7) may further include other translation categories.

A user can upload a source file to be translated to the service server (1000) through the source file upload layer (L8). In one embodiment of the present invention as shown in (b) of Fig. 4, the source file upload layer (L8) may further include descriptions of the limited capacity of the source file, the limited extension of the source file, and the file name for the convenience of the user.

Thereafter, the user can order a translation request for the original file to the service server (1000) through the order layer (L9), and as an embodiment of the present invention, the order layer (L9) may further include a customer center contact layer that connects to the customer center for the convenience of the user.

FIG. 5 schematically illustrates a process of extracting a first feature vector and a second feature vector according to one embodiment of the present invention.

As illustrated in FIG. 5, the method for determining a translator and a translation price includes a feature vector extraction step (S300) of extracting a first feature vector from the original text file through a feature extraction model based on an artificial neural network trained to extract information related to main keywords of the entire text of a document from a file including the document as a feature vector.

In addition, the above translation file corresponds to a file that the translator has completed translation of another original text file, and includes text composed in the same language as the original text file.

Schematically, Fig. 5 (a) illustrates a first feature vector extracted from an original text file, and Fig. 5 (b) illustrates a second feature vector extracted from multiple translation files for each of multiple translators.

Specifically, the feature extraction model extracts a first feature vector for an original text file uploaded by a user, as illustrated in (a) of FIG. 5, and extracts a second feature vector for each translation file of a plurality of translators stored in a database (1900) of a service server (1000), as illustrated in (b) of FIG. 5. That is, a plurality of second feature vectors are extracted based on the number of a plurality of translators stored in the database (1900) and the number of a plurality of translation files for the corresponding translators.

Meanwhile, the feature extraction model corresponds to an artificial neural network-based model that is trained to extract information related to the main keywords for the entire text in the document from the file containing the document as a feature vector. That is, the first feature vector is closely related to the characteristics of the main keywords of the original text file uploaded by the user, and each of the plurality of second feature vectors is closely related to the characteristics of the main keywords of the corresponding translation file.

In addition, it is preferable that the above translation file corresponds to a file that the translator has completed translation of another original text file, and includes text composed in the same language as the original text file.

FIG. 6 schematically illustrates a process of identifying a translator's related translation file based on the similarity of each translation file according to one embodiment of the present invention.

As illustrated in FIG. 6, the method for determining a translator and a translation price includes a similarity calculation step (S400) for calculating the similarity between the first feature vector and each of the second feature vectors extracted through the feature extraction model from the plurality of translation files of the corresponding translators stored in the database (1900), for each of the plurality of translators corresponding to the translation language type and translation category.

Specifically, the service server (1000) calculates the similarity between the first feature vector and the second feature vectors for the multiple translation files of each of the multiple translators through the similarity calculation step (S400). As an example, the similarity may be calculated through a conventional method for measuring the similarity between vectors, such as cosine similarity or Euclidean distance.

After multiple similarities are calculated for multiple translation files, the service server (1000) identifies translation files having a similarity higher than a preset value among the multiple similarities as related translation files through a candidate translator selection step (S500). In one embodiment of the present invention illustrated in Fig. 6, the preset value corresponds to 70%, and the related translation file for translator A is only one translation file #A1 among translation files #A1 to #A5, which has a similarity of 78%.

The service server (1000) can provide the effect of providing high satisfaction to users and translators using the service server (1000) by deriving related translation files that are very similar in characteristics to the original file and selecting multiple candidate translators and final translators based on the translated files.

Figure 7 schematically illustrates a process for selecting multiple candidate translators according to one embodiment of the present invention.

As illustrated in FIG. 7, the method for determining a translator and a translation price includes a similarity calculation step (S400) for calculating the similarity between the first feature vector and each of the second feature vectors extracted through the feature extraction model from the plurality of translation files of the corresponding translators stored in the database (1900), for each of the plurality of translators corresponding to the translation language type and translation category.

Specifically, as an embodiment of the present invention, the table illustrated in FIG. 7 is a result of the service server (1000) performing the execution process illustrated in FIG. 6 for each of a plurality of translators stored in the database (1900), and further includes the number of related translation files for each translator. Through the candidate translator selection step (S500), the service server (1000) compares the number of related translation files of a translator among a plurality of translators stored in the database (1900) with a preset number, and if the number of power translation files is greater than the preset number, selects the translator as a candidate translator.

As an example, if the preset number is 3, and as shown in FIG. 7, the number of related translation files of Translator A is 1, the number of related translation files of Translator B is 6, the number of related translation files of Translator C is 4, the number of related translation files of Translator D is 0, and the number of related translation files of Translator E is 3, the service server (1000) can select Translator B, Translator C, and Translator E as candidate translators among Translators A to E registered in the database (1900).

Figure 8 schematically illustrates a process of deriving a translation unit price per translator based on a preset formula according to one embodiment of the present invention.

As shown in Fig. 8, the method for determining a translator and a translation price includes a final translator selection step (S600) in which the translation unit price of each candidate translator is calculated using the following [Formula 1], and the candidate translator with the lowest translation unit price among the candidate translators is selected as the final translator.

[Formula 1]:

Translation cost of candidate translator = CostLOW+{(CostHIGH-CostLow)*(NumberTOTAL-Number)/NumberTOTAL}

(Here, CostLOW is the minimum price above, CostHIGH is the maximum price above, NumberTOTAL is the sum of the number of related translation files for each candidate translator, and Number is the number of related translation files for that candidate translator.)

Specifically, the present invention provides convenience to translators using the service server (1000) by automatically calculating the translation unit price of a candidate translator based on the following [Mathematical Formula 1], which uses as parameters the minimum and maximum prices of candidate translators received through the translation unit price selection layer (L3) illustrated in FIG. 3 and the number of related translation files of each of a plurality of candidate translators.

Preferably, in the above [Mathematical Formula 1], CostLOW is the minimum price input by the candidate translator through the translator terminal (2000), CostHIGH is the maximum price input by the candidate translator through the translator terminal (2000), NumberTOTAL is the sum of the number of related translation files of each of all candidate translators, and Number corresponds to the number of related translation files of the candidate translator.

In addition, the present invention determines that the difficulty of a user's translation request for an original text file will be relatively lower when a translator has a relatively large number of related translation files that are highly related to the similarity of feature vectors, and calculates the translation unit price accordingly. Accordingly, the service server (1000) calculates the translation unit price of a candidate translator closer to the minimum price between the minimum price and the maximum price input by the candidate translator as the number of related translation files of the candidate translator is relatively large, and calculates the translation unit price of the candidate translator closer to the maximum price between the minimum price and the maximum price input by the candidate translator as the number of related translation files of the candidate translator is relatively small.

Fig. 8 illustrates a process of calculating a translation unit price for each of a plurality of candidate translators selected in one embodiment illustrated in Fig. 7. As illustrated in Fig. 8, in one embodiment where Translator B inputs a minimum price of 40 won and a maximum price of 50 won through the translation unit price selection layer (L3), Translator C inputs a minimum price of 15 won and a maximum price of 40 won through the translation unit price selection layer (L3), and Translator E inputs a minimum price of 25 won and a maximum price of 50 won through the translation unit price selection layer (L3), based on the above [Mathematical Formula 1],

B Translator's translation cost = 40+(50-40)*{(6+4+3)-6}/6+4+3 = 40+10*7/13 ≒ 45

C translator's translation cost = 15+(40-15)*{(6+4+3)-4}/6+4+3 = 40+10*7/13 ≒ 32

E-translator's translation cost = 25+(50-25)*{(6+4+3)-3}/6+4+3 = 40+10*7/13 ≒ 44

It is produced as follows.

Preferably, the service server (1000) can select Translator C, who has the lowest translation unit price among multiple candidate translators, Translator B, Translator C, and Translator E, as the final translator. That is, the present invention can provide high satisfaction to both users and translators using the service server (1000) by selecting the final translator for the translation request by considering both the translator information entered by the translator himself and the expected difficulty level of the translation request.

FIG. 9 schematically illustrates a process for deriving a first translation price, a first translation period, a second translation price, and a second translation period according to one embodiment of the present invention.

As illustrated in FIG. 9, the method for determining a translator and a translation price includes a translation price transmission step (S700) in which a first translation price is calculated based on the translation unit price of the selected final translator and the number of words in the original text file, a first translation period is calculated based on the number of words, and the first translation period is decreased according to a preset rule while increasing the first translation unit price to derive one or more second translation prices and one or more second translation periods, and information on the first translation price, the first translation period, the one or more second translation prices, and the one or more second translation periods is transmitted to the user terminal (3000).

In addition, the above preset rule is that each of the one or more second translation periods is derived based on a certain unit period within a preset period range with respect to the first translation period, and each of the one or more second translation prices is derived inversely proportional to the second translation period, but is derived within a range of a preset minimum lower price to a preset maximum upper price, and the preset minimum lower price is lower than the first translation price and is a value equal to or greater than 0, and the preset maximum upper price is higher than the first translation price and is determined based on a second translation period with the lowest price among the one or more second translation periods.

Briefly, Fig. 9 (a) illustrates the process of deriving the first translation price and the first translation period, and Fig. 9 (b) illustrates the process of deriving the second translation price and the second translation period.

Specifically, as illustrated in (a) of FIG. 9, the service server (1000) calculates the first translation price by multiplying the translation unit price of the selected final translator and the number of words in the original text file, and derives the first translation period based on the number of words in the original text file. More specifically, the first translation period may be derived in proportion to the number of words in the original text file, but in order to take into account the translator's schedule, etc., it is preferable that a minimum period be set in advance and the first translation period be derived higher than the minimum period.

In addition, for translation work on original files, it is desirable that the translation price increases as the translation period becomes shorter, and the translation price decreases as the translation period becomes longer. Therefore, the service server (1000) can derive one or more second translation prices and one or more second translation periods through changes in the translation price according to the translation period according to a rule set in advance based on the first translation price and the first translation period.

More specifically, the above preset rule includes a rule that the one or more second translation periods are derived based on a certain unit period within a preset period range with respect to the first translation period, and that the one or more second translation prices are derived inversely proportional to the second translation period, but must be derived within a range from a preset minimum lower price to a preset maximum upper price. In addition, the preset minimum lower price is lower than the first translation price and is equal to or greater than 0, and in order to prevent the translation price from becoming too high and becoming a burden to the user, the preset maximum upper price is higher than the first translation price and is determined based on a second translation period that is the lowest among the one or more second translation periods. That is, as in one embodiment illustrated in (b) of FIG. 9, it is preferable that the second translation price is derived consistently as the maximum upper price within the preset period.

Preferably, in the graph shown in (b) of Fig. 9, the second translation period changes continuously, but so that the user can easily recognize the change in translation price according to the translation period, the service server (1000) derives one or more second translation periods based on a certain unit period within a preset period range.

FIG. 10 schematically illustrates a screen of an application executed on a user terminal (3000) to select a translation price and translation date according to one embodiment of the present invention.

In outline, Fig. 10 (a) illustrates a screen output from a user terminal (3000) that allows a user to select a translation price and translation date for a translation request as an embodiment of the present invention, and Fig. 10 (b) illustrates a screen output from a user terminal (3000) that allows a user to select a translation work speed for a translation request as another embodiment of the present invention.

Specifically, after a user places an order request for a translation request through the order layer (L9) illustrated in (b) of FIG. 4, as an embodiment of the present invention, an order information layer (L10), a translation due date setting layer (L11), and a payment amount layer (L13) may be displayed on the screen of a user terminal (3000) as illustrated in (a) of FIG. 10.

The user can check the information selected and entered in the second translation language selection layer (L6) and the second translation category selection layer (L7) shown in (b) of Fig. 4 through the above order information layer (L10), and can check the number of characters automatically measured by the service server (1000) from the original text file uploaded in the original text file upload layer (L8).

The user can select and input a desired translation period from one or more translation periods derived by the service server (1000) through the translation period setting layer (L11). As an embodiment of the present invention, the user can select one of the translation periods, such as D-Day (first translation period) or D-3, D-2, D+1, etc. (one or more second translation periods), through a scroll element included in the translation period setting layer (L11), and the translation period setting layer (L11) can record a translation period determined according to the translation period selected by the user.

The user can check the first translation price or the second translation price determined according to the translation period determined through the translation due date setting layer (L11) through the payment amount layer (L13). The translation price displayed in the payment amount layer (L13) changes in real time each time the user selects and inputs a translation period through the payment amount layer (L13), so that the user can consider the translation price displayed in the payment amount layer (L13) and determine an appropriate translation due date in the translation due date setting layer (L11).

In addition, after a user places an order request for a translation request through the order layer (L9) illustrated in (b) of FIG. 4, as another embodiment of the present invention, an order information layer (L10), an express selection layer (L12), and a payment amount layer (L13) may be displayed on the screen of a user terminal (3000) as illustrated in (b) of FIG. 10.

In another embodiment of the present invention, the service server (1000) provides the user with an opportunity to select a first translation price and a first translation period or a second translation price and a second translation period. The express selection layer (L12) includes a standard layer including a translation period according to the first translation price and the first translation period, an express layer including a translation period according to the second translation price and the second translation period, and a display element that displays the layer selected by the user. The second translation period included in the express layer corresponds to a period that is brought forward by a preset period with respect to the first translation period, and the second translation price included in the express layer corresponds to an amount that varies according to the first translation price and the second translation period. The user can select and input either the standard layer or the express layer and proceed with payment.

In addition, as shown in (b) of FIG. 4 and (b) of FIG. 10, in another embodiment of the present invention, when paying for an order for a translation request, the user can proceed with the final payment after receiving the benefits of coupons and membership discounts provided by the service server (1000).

That is, the present invention has as its technical feature the ability to derive a reasonable cost and work time for a translation request made by a user through a series of execution processes described above in FIGS. 1 to 10, and to select a translator who can produce a highly reliable translation result.

FIG. 11 schematically illustrates a screen of an application executed on a user terminal (3000) for converting a template format of an original file according to one embodiment of the present invention.

Preferably, if a user wishes to submit a paper that falls under the 'Academic' category to a university or institution, the paper needs to be revised and submitted according to the format required by the university or institution. However, if the user revises the same paper to fit different formats, it can be inconvenient, and various problems such as delays in the progress of research based on the paper can occur.

Therefore, the present invention has a technical feature of converting a plurality of template formats included in an original text file. The template format refers to a form including a basic structure and design used when writing a document, and the service server (1000) can convert the user's original text file to a template format required by the university or institution to which the user wishes to submit.

As an embodiment of the present invention illustrated in FIG. 11, the screen of an application executed on a user terminal (3000) may additionally display a template file upload layer (L14) in addition to the second translation language selection layer (L6), the second translation category selection layer (L7), the original file upload layer (L8), and the order layer (L9) illustrated in FIG. 4 (b).

A user can upload a template file including multiple source template formats through the template file upload layer (L14) using a user terminal (3000). That is, the service server (1000) converts the original template format included in the original file into the source template format included in the template file.

Meanwhile, if a template file upload layer is displayed on the screen of the user terminal (3000) but the user has not uploaded a template file, a notification can be provided to the user terminal (3000) to confirm whether or not to upload the template file, and if the user does not upload the template file thereafter, the service server (1000) transmits the original file to the selected final translator without converting the template format of the original file.

Meanwhile, in one embodiment of the present invention illustrated in FIG. 11, the upload capacity of a template file is limited to 20 MB or less, but in other embodiments, the upload capacity may not be limited to 20 MB or less.

Figure 12 schematically illustrates a process of converting a template format of an original file according to one embodiment of the present invention.

As illustrated in FIG. 12, the method for determining the translator and the translation price further includes a template format conversion step for converting the template format of the original text file, and the template format conversion step includes: a template file receiving step for receiving a template file including a plurality of source template formats from a user terminal (3000) according to a user's input; a source template format detection step for detecting the plurality of source template formats from the template file through a learned artificial neural network-based detection model; an original text template format detection step for detecting the plurality of original text template formats from the original text file through the detection model; and an original text file conversion step for converting the plurality of original text template formats included in the original text file into the plurality of source template formats, and inputting detailed data input into each of the plurality of original text template formats into a source template format corresponding to the corresponding original text template format.

In addition, the original text file conversion step includes a detailed data insertion step of deleting detailed data entered in the original text template format according to a user input received from the user terminal (3000) for the original text template format that does not correspond to the source template format, or inserting detailed data of the original text template into the source template format based on a template format matching table in which the source template formats are matched to the original text template formats; and a notification provision step of providing a notification of the source template format to the user terminal (3000) for the source template format that does not correspond to the original text template format.

The template file receiving step is performed by the service server (1000), and receives a template file including multiple source template formats from the user terminal (3000) according to the user's input. Preferably, the template file receiving step can be performed simultaneously with the original file receiving step (S200).

The source template format detection step is performed by the service server (1000), and detects the plurality of source template formats from the template file through a learned artificial neural network-based detection model. In addition, the original text template format detection step is performed by the service server (1000), and detects the plurality of original text template formats from the original text file through the detection model. The source template format and the original text template format illustrated in FIG. 12 are embodiments of the present invention, and the plurality of source template formats include a title, author information, a table of contents, an introduction, a main text, a conclusion, and references, and the plurality of original text template formats include a title, author information, a table of contents, an abstract, an introduction, a main text, a conclusion, and an appendix.

The above original file conversion step, as illustrated in FIG. 12, converts the plurality of original text template formats into the plurality of source template formats, and inputs the detailed data entered in each of the plurality of original text template formats into the source template format corresponding to the original text template format. That is, as an example illustrated in FIG. 12, for the title, author information, table of contents, introduction, main text, and conclusion where the original text template format and the source template format exactly match, the detailed data is input as is.

However, since original template formats that do not correspond to the source template format, such as 'abstract' and 'appendix', are not essential elements for the document format desired by the user, the service server (1000) can delete detailed data included in the original template format, such as 'appendix', or insert detailed data in the source template format preset for each original template format, such as 'abstract' (in the case of FIG. 12, 'conclusion'), based on the user's input received from the user terminal (3000).

In addition, since a source template format that does not correspond to the original template format, such as 'references', is absolutely necessary for the document format desired by the user, the service server (1000) provides a notification of the absence of the corresponding source template format to the user terminal (3000) through the notification provision step.

Figure 13 schematically illustrates a template format matching table according to one embodiment of the present invention.

As illustrated in FIG. 13, the template format matching table includes a plurality of original text template formats and a plurality of source template formats, matches one or more preset source template formats to each of the plurality of original text template formats, and further includes a preset priority for each of the one or more source template formats when one or more of the source template formats is matched to the original text template format, and the detailed data insertion step inserts the detailed data into the source template format based on the preset priority.

Specifically, in one embodiment illustrated in FIG. 12, for an original template format that does not correspond to a source template format, such as 'Green', the service server (1000) inserts detailed data of the original template format into a preset source template format, but a situation may occur where the preset source template format is not included in the plurality of minority template formats. Therefore, the service server (1000) creates a template format matching table including a plurality of original template formats, a plurality of source template formats, and priorities, and inserts detailed data of the original template format into the minority template format based on the template format matching table.

The table illustrated in FIG. 13 is a template format matching table in one embodiment of the present invention, in which a plurality of source template formats matching the original template format corresponding to 'abstract' include 'introduction', 'body', and 'conclusion', and a priority is preset for each source template format. That is, the service server (1000) inserts the detailed data of the original template format corresponding to 'abstract' into the source template format corresponding to 'conclusion' with the highest priority, but if the template file uploaded by the user does not include the source template format corresponding to 'conclusion', it inserts the detailed data into the source template format corresponding to 'body' with the next highest priority.

Preferably, if there is no source template format matching the original template format in the template format matching table, the service server (1000) can delete the detailed data of the original template format regardless of the user's input.

In one embodiment of the present invention, by comparing the feature vector of an original file for which a user has requested translation with the feature vector of a translation file performed by a translator through a learned artificial neural network-based feature extraction model and selecting a candidate translator based on similarity, the reliability of the translation result can be improved and the effect of providing high satisfaction to the user can be achieved.

In one embodiment of the present invention, by identifying a related translation file among a plurality of translation files whose feature vector similarity is greater than a preset value and selecting a candidate translator based on the similarity, the reliability of the translation result can be improved and a high level of satisfaction can be provided to the user.

In one embodiment of the present invention, by selecting a final translator among a plurality of candidate translators through a specific formula that uses as variables the minimum price and maximum price input by the translator to the service server, the effect of providing high convenience to the translator using the service server can be achieved.

In one embodiment of the present invention, by deriving additional translation prices and translation periods based on preset rules for the translation prices and translation periods calculated for the final translator, the user can have the effect of providing an opportunity to select the option he or she wants in terms of cost and time for a translation request.

In one embodiment of the present invention, when the format of a document handled by an institution requiring submission is different from the format of the original file, the format of the original file is automatically converted into the corresponding format through a template file composed of the format handled by the institution, thereby providing convenience to the user.

Although the embodiments have been described above by way of limited examples and drawings, those skilled in the art can make various modifications and variations from the above description. For example, appropriate results can be achieved even if the described techniques are performed in a different order from the described method, and/or components of the described system, structure, device, circuit, etc. are combined or combined in a different form from the described method, or are replaced or substituted by other components or equivalents. Therefore, other implementations, other embodiments, and equivalents of the claims also fall within the scope of the claims described below.

Claims (7)

A method for determining a translator and a translation price for a translation request performed on a service server associated with an application performed on a user terminal and a translator terminal,
A translator information receiving step for receiving translator information including translation language type, translation category, and translation price range per word including minimum price and maximum price from a translator terminal;
A source file receiving step for receiving the translation language type, translation category, and source file to be translated from the user terminal according to the user's input;
A feature vector extraction step of extracting a first feature vector from the original text file through an artificial neural network-based feature extraction model trained to extract information related to main keywords of the entire text of a document from a file containing the document as a feature vector;
A similarity calculation step for calculating the similarity between the first feature vector and each of the second feature vectors extracted through the feature extraction model from the plurality of translation files of the corresponding translators stored in the database, for each of the plurality of translators corresponding to the above translation language type and translation category;
A candidate translator selection step in which translation files with a similarity level greater than or equal to a preset value are identified as related translation files, and when the number of related translation files of each translator is greater than or equal to a preset number, the translator is selected as a candidate translator;
The final translator selection stage calculates the translation unit price of each candidate translator using the following [Formula 1], and selects the candidate translator with the lowest translation unit price among the candidate translators as the final translator;
A translation price transmission step of 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, calculating a first translation period based on the number of words, and increasing the first translation price while decreasing the first translation period according to preset rules to derive one or more second translation prices and second translation periods, and transmitting information about the first translation price, the first translation period, the one or more second translation prices, and the one or more second translation periods to the user terminal;
Including an original file transmission step for transmitting the original file and information about the translation period based on the user's payment information to the translator terminal according to the result of verification of whether the user has made payment;
The translation unit price of each of the above candidate translators is calculated by the following [Formula 1], taking into account the expected difficulty of the translation request based on the number of translation files related to the candidate translator, and some of the translator information entered by the candidate translator.
The method for determining the above translator and translation price further includes a template format conversion step for converting the template format of the above original file,
The above template format conversion step is,
A template file receiving step for receiving a template file including multiple source template formats according to user input from a user terminal;
A source template format detection step for detecting the plurality of source template formats from the template file using a detection model based on a learned artificial neural network;
A source text template format detection step for detecting multiple source text template formats from the source text file through the above detection model; and
A method for determining a translator and a translation price, comprising: an original text file conversion step of converting a plurality of original text template formats included in the original text file into the plurality of source template formats, and inputting detailed data entered into each of the plurality of original text template formats into a source template format corresponding to the original text template format.

[Formula 1]:
Translation cost of candidate translator = CostLOW+(CostHIGH-CostLow)*(NumberTOTAL-Number)/NumberTOTAL
(Here, CostLOW is the minimum price above, CostHIGH is the maximum price above, NumberTOTAL is the sum of the number of related translation files for each candidate translator, and Number is the number of related translation files for that candidate translator.)
In claim 1,
The above established rules are:
Each of the above 1 or more second translation periods is derived based on a certain unit period within a period range set for the above first translation period,
Each of the above 1 or more second translation prices is derived inversely proportional to the second translation period, but is derived within the range of the preset minimum lower price and the preset maximum upper price.
The above preset minimum price is lower than the first translation price and is a value greater than or equal to 0,
A method for determining a translator and a translation price, wherein the above-described maximum price is higher than the above-described first translation price and is determined based on the lowest second translation period among the one or more second translation periods.
In claim 1,
The above translation file is,
This is a file that the translator has completed translation of for another original file.
A method for determining a translator and a translation price, wherein the text comprises the same language as the original text file.
delete In claim 1,
The above original file conversion step is,
For a source template format that does not correspond to a source template format, a detailed data insertion step of deleting detailed data entered in the original template format according to a user input received from the user terminal, or inserting detailed data of the original template into the source template format based on a template format matching table in which the source template format is matched to the original template format; and
A method for determining a translator and a translation price, comprising: a notification provision step for providing a notification of a source template format that does not correspond to the original text template format to the user terminal;
In claim 5,
The above template format matching table is,
Contains multiple source template formats and multiple source template formats;
Matching one or more preset source template formats to each of the above multiple original text template formats,
If one or more source template formats match the original text template format, each of the one or more source template formats further includes a preset priority.
The above detailed data insertion step is a method for determining a translator and translation price by inserting the above detailed data into the source template format based on the above preset priority.
A service server that is associated with an application running on a user terminal and a translator terminal and performs a method of determining a translator and a translation price for a translation request.
A translator information receiving unit that receives translator information including translation language type, translation category, and translation price range per word including minimum price and maximum price from a translator terminal;
A source file receiving unit that receives the translation language type, translation category, and source file to be translated from a user terminal according to user input;
A feature vector extraction unit that extracts a first feature vector from the original text file through an artificial neural network-based feature extraction model that is trained to extract information related to main keywords of the entire text of the document from the file containing the document as a feature vector;
A similarity calculation unit that calculates the similarity between the first feature vector and each of the second feature vectors extracted through the feature extraction model from the plurality of translation files of the corresponding translators stored in the database, for each of the plurality of translators corresponding to the above translation language type and translation category;
A candidate translator selection committee that identifies translation files with a similarity level greater than or equal to a preset value as related translation files, and selects each translator as a candidate translator when the number of related translation files of each translator is greater than or equal to a preset number;
The final translator selection committee calculates the translation unit price of each candidate translator using the following [Formula 1] and selects the candidate translator with the lowest translation unit price among the candidate translators as the final translator;
A translation price transmitter which calculates 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, calculates a first translation period based on the number of words, and increases the first translation price while decreasing the first translation period according to preset rules to derive one or more second translation prices and second translation periods, and transmits information about the first translation price, the first translation period, the one or more second translation prices, and the one or more second translation periods to the user terminal;
Including an original file transmission unit that transmits the original file and information about the translation period based on the user's payment information to the translator terminal according to the result of verification of the user's payment;
The translation unit price of each of the above candidate translators is calculated by the following [Formula 1], which takes into account the expected difficulty of the translation request derived from a portion of the translator information entered by the candidate translator and the number of related translation files of the candidate translator.
The method for determining the above translator and translation price includes a template format conversion step for converting the template format of the above original file,
The above template format conversion step is,
A template file receiving step for receiving a template file including multiple source template formats according to user input from a user terminal;
A source template format detection step for detecting the plurality of source template formats from the template file using a detection model based on a learned artificial neural network;
A source text template format detection step for detecting multiple source text template formats from the source text file through the above detection model; and
A service server, characterized in that it includes an original text file conversion step of converting a plurality of original text template formats included in the original text file into the plurality of source template formats, and inputting detailed data entered into each of the plurality of original text template formats into a source template format corresponding to the original text template format.

[Formula 1]:
Translation cost of candidate translator = CostLOW+(CostHIGH-CostLow)*(NumberTOTAL-Number)/NumberTOTAL
(Here, CostLOW is the minimum price above, CostHIGH is the maximum price above, NumberTOTAL is the sum of the number of related translation files for each candidate translator, and Number is the number of related translation files for that candidate translator.)

Images