JP2012230460A - Machine translation system, method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a machine translation system, method, and a program for creating translation of original text to be translated in consideration of format information included in the original text to be translated, even when a translation example is used for example translation.SOLUTION: When a translation example is registered in a translation example database 28, example registration means 34 also registers format information included in original text to be registered acquired by format information acquisition means 33. When the registered example is used, format information checking means 36 checks format information included in original text to be translated against format information included in the translation example retrieved by translation example retrieval means 35. When the two pieces of format information match with each other, translation creation means 37 creates translation of the original text to be translated by using translation of the translation example.

Description

  Embodiments described herein relate generally to a machine translation system, a method, and a program for translating using a translation example that is a parallel translation of an original sentence in a first language and a translated sentence in a second language.

  In a machine translation system that translates an original sentence in a first language into a sentence in a second language, there is a technique called example translation using a translation example. In the example translation, a plurality of translation examples consisting of parallel translations of the original text in the first language and the translation text in the second language are stored in advance in the translation example database, and a translation example similar to the input original text to be translated (hereinafter referred to as a similar example). ) And if the translation target text exactly matches the retrieved similar example, the similar example translation is output as a translation of the translation target original.

  If there is a difference (hereinafter referred to as “difference”) between the original text to be translated and the retrieved similar example original text, the difference is determined in terms of words, the phrase of the similar example translation corresponding to the difference in the similar example original text is determined, and A translation is created by replacing a phrase with a translation corresponding to a difference phrase in the original text to be translated.

  Since the translation examples stored in the translation example database have only text information, only the text of the translation target text is used at the time of translation of the example, and the format information is not used even if the translation target text has the format information. For example, in a Word document or an HTML document, format information is given to each sentence, but at present, these format information is not used for example translation. For this reason, when the original text having a different format is translated, the translated text becomes the same regardless of the format information, and the translation accuracy may be lowered.

  For example, in the case of a translation example in which the translation example source text is “solve trouble” and the translation example translation is “Canceling Trouble”, if the example translation is performed using the translation example database including the translation example, All translations of “Troubleshooting” are “Canceling Trouble”. Although this translation is suitable as a translation of the title sentence, it is not perfect as a translation of a normal sentence, resulting in a decrease in translation accuracy.

  Therefore, as a translation example suitable for the translation of the passage, if the translation example original text is “Troubleshooting” and the translation example translation is “Cancel a trouble”, a new translation example is registered. Since the new translation examples are preferentially applied, all “solve the trouble” in the sentence becomes “Cancel a trouble”.

  In order to change the translation method according to the format of the original text to be translated, the input sentence data to be translated is divided one sentence at a time, and the sentence is identified as a title, a bullet, or a text from the format of each sentence. In some cases, translation processing is performed in accordance with the determined sentence type, and the sentence data is translated into conformity to the sentence type according to the sentence type of each sentence (see, for example, Patent Document 1). However, this patent document 1 does not consider the format information of the original text to be translated in the example translation.

  Thus, at present, in the example translation, the translation examples stored in the example database have only text information, and a translation is created only with the text information of the original sentence. For this reason, in a document in which format information is given to each sentence, all original sentences having the same text become the same translated sentence, resulting in a decrease in translation accuracy.

Japanese Examined Patent Publication No. 4-12505

  A problem to be solved by the present invention is a machine translation system, method, and program for creating a translation of an original text to be translated in consideration of the format information of the original text to be translated even when the example is translated using the translation example Is to provide.

  A machine translation system according to an embodiment of the present invention includes a machine translation program, a machine translation dictionary for translating an original text in a first language to be translated into a translated text in a second language for translation, a machine translation rule, and an original text in the first language. Example database for translation consisting of translation examples composed of translations of translations in the second language and a memory storing a format information internal code correspondence table consisting of pairs of format information possessed by various applications and internal codes corresponding to the format information An apparatus and an arithmetic control unit that executes a machine translation program.

  The format information acquisition means acquires the format information of the original text in the first language when registering the translation example that is a translation of the original text in the first language input from the input device and the translated text in the second language in the translation example database. Then, the format information of the first language is displayed and output in the applied format column of the translation example registration screen displayed on the display device.

  The translation example registration means converts the format information set by the input device to the internal code based on the format information internal code correspondence table for the format information displayed in the applicable format column of the translation example registration screen, and includes it. Then, the translation examples are registered in the translation example database.

  The translation example retrieval unit retrieves a translation example similar to the first language original text from the translation example database using the original text in the first language input from the input device as a search key. The format information collating unit collates the format information possessed by the original sentence of the translation example similar to the original sentence of the first language retrieved by the translation example retrieving unit and the format information possessed by the original sentence of the first language. The translation creation means creates a translation of the original text in the first language using the translation example when the original text in the first language matches the original text in the translation example in the collation of the format information matching means.

1 is a functional block diagram of a machine translation system according to an embodiment of the present invention. The block block diagram which shows the hardware constitutions of the machine translation system which concerns on embodiment of this invention. The flowchart which shows the content of the example translation process of the machine translation system which concerns on embodiment of this invention. The flowchart which shows the content of the processing formula information acquisition process S1 of FIG. The block diagram which shows an example of cooperation translation with the machine translation system which concerns on embodiment of this invention, and arbitrary applications. Explanatory drawing of an example of the translation example registration screen in embodiment of this invention. Explanatory drawing of an example of the format information registration screen in embodiment of this invention. The flowchart which shows the content of the example translation process S3 of FIG. It is explanatory drawing of an example of the translation document in the case of example translation in the machine translation system which concerns on embodiment of this invention. Explanatory drawing when the application in which the original text to be translated exists and the machine translation system exist in different processing environments.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a functional block diagram of a machine translation system 11 according to an embodiment of the present invention, and FIG. 2 is a block configuration diagram showing a hardware configuration of the machine translation system according to the embodiment of the present invention.

  In FIG. 2, the machine translation system 11 is realized by installing a software program such as a machine translation program in a general computer and executing the software program in the processor 13 of the arithmetic control device 12.

  The arithmetic control device 12 performs various arithmetic operations related to machine translation. The arithmetic control device 12 has a processor 13 and a memory 14. A memory translation program 15 is stored in the memory 14, and the processor 13 performs processing. When executed, the work area 16 is used. The calculation results of the calculation control device 12 are output to the display device 18, the printing device 19, and the disk drive 20, which are output devices 17, and are also output to the communication network via the communication control device 21.

  The input device 22 is used to input information to the arithmetic control device 12, and includes, for example, a mouse 23, a keyboard 24, a disk drive 20, and a communication control device 21. For example, the mouse 23 and the keyboard 24 are connected via the display device 18. Then, various commands are input to the arithmetic and control unit 12, and the keyboard 24, the disk drive 20, and the communication control unit 21 input a document to be translated.

  That is, the disk drive 20 inputs / outputs a file of a document to be translated to / from a storage medium, and the communication control device 21 connects the machine translation system 11 to a communication network such as the Internet or a LAN. The communication control device 21 is a device such as a LAN card or a modem, and data transmitted / received to / from the communication network via the communication control device 21 is transmitted / received to / from the arithmetic control device 12 as an input signal or an output signal. Further, a hard disk drive (HDD) 25 is provided for storing a calculation dictionary of the calculation control device 12 and a translation dictionary storing knowledge and rules necessary for translation.

  FIG. 1 is a functional block diagram of a machine translation system 11 according to an embodiment of the present invention. Each functional block in the arithmetic and control unit 12 shown in FIG. 2 corresponds to each program constituting the machine translation program 15 described above. That is, when the processor 13 executes each program constituting the machine translation program 15, the arithmetic control device 12 functions as each functional block. Each block of the storage device 26 corresponds to the storage area of the memory 14 and the hard disk drive 25 in the arithmetic and control unit 12.

  The storage device 26 includes a machine translation dictionary for translating the original text of the first language to be translated into a translated text of the second language for translation, a dictionary rule section 27 storing machine translation rules, the original text of the first language and the first text. A translation example database 28 in which a plurality of translation examples consisting of parallel translations with bilingual translations, a format consisting of pairs of format information held by various application programs (hereinafter referred to as applications) and internal codes corresponding to the format information An information internal code correspondence table 29 is stored.

  The input processing unit 30 of the arithmetic control device 12 performs input processing on the original text of the first language to be translated and information necessary for the operation input from the input device 22 and outputs the input information to the translation unit 31.

  The translation unit 31 of the arithmetic and control unit 12 uses the translation example in the translation example database 28, the machine translation dictionary and machine translation rule in the dictionary rule unit 27, and the format information internal code in the format information internal code correspondence table 29. Machine translation of the original text of the first language to be translated input from.

  The display processing unit 32 of the arithmetic control device 12 includes a translation of the first language source text to be translated, a translated second language text, a translation of the first language text and the second language translation text when registering the translation example, The format information to be registered together with the parallel translation is subjected to display processing and displayed on the display device 18.

  The translation unit 31 obtains format information of the original text in the first language when registering the translation of the original text in the first language and the translated text in the second language as a translation example in the translation example database 28; The format information acquired by the format information acquisition means 33 is converted into an internal code based on the format information internal code correspondence table 29, and the translation of the original translation in the first language and the translation in the second language is translated as an example for translation. Example registration means 34 for registering in the example database 28, translation example search means 35 for searching a translation example similar to the translation target original text from the translation example database 28 using the translation target original text in the first language as a search key, and translation example search means The format information of the similar example retrieved in 35 is converted into an internal code based on the format information internal code correspondence table 29, and the format information internal code of the original text to be translated Format information matching means 36 for matching, translation creating means 37 for creating a translation of the original text to be translated from the translation of the similar example whose format information is matched by the matching in the format information matching means 36, and matching in the format information matching means 36 When there is no similar example with matching format information, the machine translation dictionary of the dictionary rule unit 27 and the translation means 38 for translating the original text to be translated using the machine translation rules.

  In the machine translation system according to the embodiment of the present invention, when registering a translation example in the translation example database 28, the format information possessed by the registered original text is obtained by the format information obtaining unit 33, and the example registration unit 34 translates together with the format information. The example is registered in the translation example database 28. Then, when using the example registered in the translation example database 28, the translation example search means 35 searches for the translation example from the translation example database 28, and the format information collating means 36 determines that the searched original text to be translated is format information. If there is a match, the format information of the original text to be translated is matched with the format information of the example, and if they match, the translation creation means 37 creates a translation of the source text to be translated using the translation of the translation example To do.

  FIG. 3 is a flowchart showing the contents of the example processing for translation of the machine translation system according to the embodiment of the present invention. First, the format information acquisition means 33 performs format information acquisition processing (S1). That is, the format information acquisition means 33 acquires format information of translation examples registered in advance so that the format information can be used in the example translation. Then, the example registration means 34 performs a translation example registration process with format information (S2). That is, the example registration unit 34 registers the translation example in the translation example database 28 together with the format information.

  Next, the translation example search means 35, the format information collation means 36, and the translation creation means 37 perform example translation (S3). That is, the translation example search unit 35, the format information collating unit 36, and the translation creation unit 37 perform the example translation process using the translation example with format information registered in the translation example database 28.

  FIG. 4 is a flowchart showing the contents of the format information acquisition process S1 of FIG. In the format information acquisition process, first, the format information acquisition unit 33 determines whether the original document of the translation example to be registered is a document type having format information (S11).

  In this case, it is necessary to import into the machine translation system information on the document to be translated (original text text, format information). Since the document to be translated is a translatable (text can be acquired) document having a file format of an arbitrary application, it is necessary to coordinate the application with the machine translation system.

  There are various ways to link this application with the machine translation system. For example, data communication interfaces (technology) between applications include DDE (Dynamic Data Exchange), OLE (Object Linking and Embedding), OCX (OLE Custom Control), ActiveX, and COM (Component Object Model).

  FIG. 5 is a configuration diagram illustrating an example of cooperative translation between the machine translation system and any application according to the embodiment of the present invention. As shown in FIG. 5, in order to realize cooperative translation, a macro or a plug-in (hereinafter referred to as a macro plug-in) is incorporated into the application 39. These are generally realized as a toolbar or menu command of the application 39. This macro plug-in communicates with the translation unit 31 of the machine translation system 11 via the linkage interface 40 such as ActiveX or COM, and exchanges data. DDE, OLE, OCX, ActiveX, COM, etc. are mature technologies that have been used in many fields and have already been introduced in various materials.

  Whether or not the document type has format information in step S11 in FIG. 4 is determined from the type of application. For example, since documents of applications such as Word (registered trademark), Excel (registered trademark), and PowerPoint (registered trademark) have format information, in the case of these applications, the document type has format information. Is determined. Instead of determining from the type of application, the determination may be made from the type of document to be translated. Documents with extensions DOC, XLS, PPT, HTML, XML, and the like have format information. If they are these documents, it is determined that the document type has format information.

  Next, when the document type has format information, the format information acquisition unit 33 acquires the format information of the original text of the translation example to be registered by the macro plug-in incorporated in the application (S12).

  Then, the format information acquisition means 33 sets the acquired format information on the translation example registration screen (S13). In other words, as will be described later, the format information of the original text of the translation example to be registered is displayed and output in the application format field of the translation example registration screen. Thus, the user can confirm the format information and can update the format information through the translation example registration screen.

  On the other hand, when it is determined in step S11 that the document type is not a document type having format information, initial information is set on the translation example registration screen (S14). This is because the setting information of the format information can be updated through the translation example registration screen even if the document type has no format information.

  FIG. 6 is an explanatory diagram showing an example of a translation example registration screen according to the embodiment of the present invention. As shown in FIG. 6, the translation example registration screen is provided with an original text field, a translated text field, an application format field, a registration button, and a cancel button. In the original text column, input the original text “Troubleshooting” of the translation example to be registered, and in the translated text field, enter the translated text “Canceling Trouble” of the translation example to be registered. The translation of the example for translation may be a translation generated by machine translation or a translation manually input by the user.

  The applied format is a format in which the registered translation example is applied in the example translation. In FIG. 6, as the applied format, “Table”, “Bullet”, “Auto Shape”, “Title”, “Heading 1” of the Word document are used. ”,“ Heading 2 ”,“ Heading 3 ”,“ Heading 4 ”, and“ Body ”. Select the application format by entering a check mark in the □ column of this application format.

"(No format)" is selected when not selecting all the applicable formats in these lists. "All formats" is selected when all the applicable formats in these lists are selected. Is to be selected. The registration button is operated when registering the translation example, and the cancel button is operated when canceling the process.
If the candidate in the list of applicable formats is a document type having format information, the format information acquisition unit 33 displays a list of formats in the document. On the other hand, if the document type does not have the format information, the format information acquisition means 33 displays the format information of the initial condition determined in advance. For example, if the document type does not have format information, “table”, “itemized item”, “auto shape”, “title”, and “text” are displayed as format information of predetermined initial conditions. Regardless of the original document type of the translation example to be registered, the format information of the initial conditions may always be displayed as a list.

  In the example of FIG. 6, the translation example database 28 is that the original sentence of the translation example is “to solve the trouble”, the translation sentence of the translation example is “Canceling Trouble”, and “Heading 2” is selected as the application format. The case of registering is shown.

Next, the format information is converted into an internal code by the example registration means 34 based on the format information internal code correspondence table and registered in the translation example database. As shown in Table 1, the format information internal code correspondence table is a table in which format information and internal codes are associated in advance.

  As shown in Table 1, the format information “(no format)” to “text” is associated with the internal codes “NONE” to “NONE” in advance. That is, “(no format)” is “NONE”, “all formats” is “ALL”, “table” is “TABLE”, “bullet” is “ITEMS”, “auto shape” is “SHAPE”, “title” "Is" TITLE "," Heading 1 "is" TITLE "," Heading 2 "is" TITLE "," Heading 3 "is" TITLE "," Heading 4 "is" TITLE "," Text "is" NONE " Each is associated.

  Here, when the format information acquisition means 33 detects new format information that is not in the format information internal code correspondence table, it displays a format information registration screen that prompts the user to perform new registration. FIG. 7 is an explanatory diagram of an example of a format information registration screen.

  Now, it is assumed that the format information acquisition means 33 detects new format information “table number” that is not in the format information internal code correspondence table as format information. In this case, a case will be described in which registration is performed by inputting the internal code “CHARTNUM” for the newly detected format “chart number”.

  If the format information acquisition means 33 detects new format information that is not in the format information internal code correspondence table, it displays a format information registration screen shown in FIG. The format information registration screen has a format field, an internal code field, a registration button, and a cancel button. In the format field, the format information format “table number” to be newly registered is input, and in the internal code field. Then, input the internal code “CHARTNUM” of the format information to be newly registered.

  The internal code may be input by the user, or based on the result of verifying various applications, the internal code corresponding to the format information that may appear is given to the machine translation system in advance, and the format information registration screen The user may be allowed to select.

The registration button is operated when registering format information, and the cancel button is operated when canceling processing. As a result, the format information whose internal code is “CHARTNUM” is registered for the newly detected format “chart number”. Table 2 shows the format information internal code correspondence table after registration. As shown in Table 2, a format “Chart Number” and an internal code “CHARTNUM” are added to Table 1.

  As described above, when the format information acquisition unit 33 detects a format that is not registered in the format information internal code correspondence table, the user can register it in the format information internal code correspondence table. Note that a tool for editing the format information internal code correspondence table may be prepared to provide functions such as registration, change, and deletion.

  In the example of FIG. 6, the format information acquisition means 33 displays a list of formats acquired from the Word document in the applied format column of the translation example registration screen, generates a list of formats, and the original text of the translation example to be registered is Check the corresponding format (heading 2) in the format information list according to the format information. Note that the format information acquisition unit 33 may check a fixed format (for example, “all formats”) other than the corresponding format (heading 2) as an initial value. This improves the accuracy of the user's format specification and reduces the format specification effort.

  When the user clicks the registration button, the example registration means 34 executes the example registration process S2 with translation information in FIG. 3, and the application format set on the translation example registration screen is the translation example database together with the translation of the original text and the translated text. 28 is registered. In this case, the application format set on the translation example registration screen is converted into an internal code based on the format information internal code correspondence table and registered in the translation example database 28.

  Thereby, the registration information in the translation example database 28 is as follows. SRC represents the original text, TGT represents the translated text, and STY represents the format information.

SRC: Solving trouble
TGT: Canceling Trouble
STY: TITLE
In addition to the original text SRC, the translated text TGT, and the format information STY, fields (information) such as a registrant, a registration date / time, an updater, and an update date / time may be registered.

  Next, a translation process for performing example translation using the translation examples registered in this way will be described.

  FIG. 8 is a flowchart showing the contents of the example translation process S3 of FIG. As shown in FIG. 8, in the example translation process, when a translation request is issued for an original sentence in the first language to be translated, the translation means 38 performs dictionary lookup and morpheme on the input original sentence to be translated. Analysis is performed (S31), and various information such as part-of-speech of each morpheme, utilization type, various attributes, and translation are obtained.

  Subsequently, the translation example search means 35 searches the translation example database 28 for a translation example search (S32). That is, it is searched whether translation examples having the same original text or similar original text as the input original text are accumulated.

  At this stage, morphological analysis results and various information for each word in the original text to be translated have been acquired. Information has also been acquired. With these pieces of information, even if the translation example does not completely match the input original text by one character, it is possible to perform a vibration absorption process that does not recognize the difference regarding the fluctuation of these notations as a difference.

  On the other hand, it is possible to make the search successful only when there is an exact translation example that does not absorb the fluctuation of the notation. In addition, even if there is a difference between characters and words, a difference character count or a lower limit value of the ratio of the difference word count that is regarded as the same value may be provided, and the possibility of search may be controlled by a so-called similarity. For example, the similarity may be obtained from the number of matching characters among the characters in the input original text and the characters in the example original text, and those having a similarity greater than or equal to a predetermined value may be searched as similar examples. In this case, the completely matching translation example is a similarity example having a similarity of 100%. In this way, translation examples that use similar words / characters are extracted. In this case, it is not necessary to perform the example search S2 after the dictionary lookup and the morphological analysis. Therefore, it may be performed at the beginning of the process of FIG.

  Next, the translation example search means 35 determines whether or not the similar example can be searched by the example search (S33). If the similar example cannot be searched, the translation unit 38 performs normal translation using the machine translation dictionary and the machine translation rule of the dictionary rule unit 27 and creates a translation (S36). On the other hand, when the similar example can be searched, the format information collating means 36 collates the format information of the retrieved translation example with the format information of the original text to be translated (S34).

  As described above, the format information of the source text to be translated is sent to the machine translation system together with the text of the source text by the macro plug-in embedded in the application. , Already registered in the translation example registration screen described above. Accordingly, the format information collating means 36 converts the format information held in the translation target text into an internal code based on the format information internal code correspondence table shown in Table 1, and collates the format information (internal code) (S34). . That is, it is determined whether the format information of the original text to be translated matches the format information of the translation example (S35), and if the format information matches, a translation is created using the translation example (S36). On the other hand, if the format information does not match, the process returns to step S32, and an example search is performed again for other similar examples.

FIG. 9 is an explanatory diagram of an example of a translated document when the example is translated. The original text to be translated is a Word document, which indicates a document having “title”, “heading 1”, and “text” as format information. In this case, the format information acquisition means 33 acquires the format information shown in Table 3 from the original text to be translated.

  IDs in Table 3 are sentence numbers. “XX System Instruction Manual” is format information “Title”, “Before Installation” is format information “Heading 1”, “Install” is format information “Heading 1”, “Getting Started” is format information “ “Heading 1” and “Resolve trouble” are format information “Heading 1”, and “Resolve trouble” is format information “Body”. Assume that the following example is registered in the translation example database.

SRC: Solving trouble
TGT: Canceling Trouble
STY: TITLE
In the translation of the 15th sentence “Troubleshooting”, the similar example can be searched in step S32 of FIG. 8, and the format code of the original sentence and the internal code of the format information of the similar example are both “TITLE” in step S34. Therefore, this similar example is used and the translation is “Canceling Trouble”.

  On the other hand, in the translation of the 17th sentence “solve the trouble”, the similar example can be searched in step S32, and the format information of the original sentence does not match the format information of the example in step S34. Does not apply. Therefore, the translation unit 38 performs normal translation using the machine translation dictionary and the machine translation rule of the dictionary rule unit 27, and the translated sentence becomes “Cancel a trouble”.

  In this way, when registering a translation example, the format information to which the example can be applied is specified and registered, and the format information of the source text to be translated is compared with the format information registered in the example when translating the example. If they match, the translation accuracy is improved by creating a translation using the translation example.

  In the above description, the application and machine translation system where the translation target text exists coexist in the same processing environment. However, the application where the translation target text exists and the machine translation system exist in different processing environments. You may make it the structure to carry out.

  FIG. 10 is an explanatory diagram in the case where the application in which the original text to be translated exists and the machine translation system exist in different processing environments. For example, FIG. 10 shows a case where the application 39 in which the translation target text exists exists in the client 41 and the machine translation system 11 exists in the server 42. The client 41 and the machine translation system 11 are connected via a network 43. That is, the server 42 can provide the client 41 with a machine translation function in a form called ASP, SaaS, or cloud computing.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 11 ... Machine translation system, 12 ... Arithmetic control device, 13 ... Processor, 14 ... Memory, 15 ... Machine translation program, 16 ... Work area, 17 ... Output device, 18 ... Display device, 19 ... Printing device, 20 ... Disk drive , 21 ... communication control device, 22 ... input device, 23 ... mouse, 24 ... keyboard, 25 ... hard disk drive, 26 ... storage device, 27 ... dictionary rule part, 28 ... translation example database, 29 ... format information internal code correspondence table , 30 ... input processing section, 31 ... translation section, 32 ... display processing section, 33 ... format information acquisition means, 34 ... example registration means, 35 ... translation example search means, 36 ... format information collation means, 37 ... translation creation means 38 ... translation means, 39 ... macro plug-in, 40 ... cooperation interface, 41 ... client, 42 ... server, network 43

Claims (5)

Machine translation program, machine translation dictionary for translating an original sentence in a first language to be translated into a second language translation sentence for translation, and an example for translation comprising a translation between the original sentence in the first language and the translated sentence in the second language Example database for translation, storage device storing a format information internal code correspondence table consisting of pairs of format information possessed by various applications and internal codes corresponding to the format information, and arithmetic control for executing the machine translation program In a machine translation system comprising an apparatus, the machine translation program includes:
Format information for acquiring the format information of the original text in the first language when the translation example that is a translation of the original text in the first language and the translated text in the second language input from the input device is registered in the translation example database Acquisition means;
A translation example registration means for converting the format information into an internal code based on the format information internal code correspondence table and including the translation example in the translation example database;
A translation example search means for searching a translation example similar to the original sentence of the first language to be translated from the translation example database using the original sentence of the first language to be translated input from the input device as a search key;
Format information collating means for collating the format information possessed by the original text of the translation example similar to the original text of the first language to be translated with the format information possessed by the original text of the first language to be translated;
When the format information of the original text of the first language to be translated matches the format information of the original text of the translation example in the collation by the format information collating means, the translation of the original text of the first language based on the translation example Machine translation system comprising a translation creation means for creating a sentence.   When the bilingual translation of the original text in the first language and the translation text in the second language is registered in the translation example database, the format information acquisition means displays the translation target document in the application format column of the translation example registration screen. The machine translation system according to claim 1, wherein an application format corresponding to the type of application is provided.   The format information acquisition means displays a format information registration screen when format information that does not exist in the format information internal code correspondence table is detected, and displays a pair of format information and internal code input from the input device as the format information. The machine translation system according to claim 1, wherein the machine translation system is registered in the information internal code correspondence table. Machine translation program, machine translation dictionary for translating an original sentence in a first language to be translated into a second language translation sentence for translation, and an example for translation comprising a translation between the original sentence in the first language and the translated sentence in the second language Example database for translation, storage device storing a format information internal code correspondence table consisting of pairs of format information possessed by various applications and internal codes corresponding to the format information, and arithmetic control for executing the machine translation program Operating a computer equipped with a device,
Obtaining the format information of the original text in the first language when registering the translation example that is a translation of the original text in the first language and the translated text in the second language input from the input device in the translation example database;
Converting the acquired format information of the first language into an internal code based on the format information internal code correspondence table;
Including it, register the translation example in the translation example database,
When an original sentence in the first language to be translated is input from the input device, a translation example similar to the original sentence in the first language to be translated is searched from the translation example database using the original sentence in the first language as a search key. And
Collating the format information possessed by the original text of the translation example similar to the retrieved original text of the first language to be translated with the format information possessed by the original text of the first language;
Machine translation that creates a translation of the original text of the first language based on the translation example when the format information of the original text of the first language to be translated matches the format information of the original text of the translation example How to provide functionality. Machine translation program, machine translation dictionary for translating an original sentence in a first language to be translated into a second language translation sentence for translation, and an example for translation comprising a translation between the original sentence in the first language and the translated sentence in the second language Example database for translation, storage device storing a format information internal code correspondence table consisting of pairs of format information possessed by various applications and internal codes corresponding to the format information, and arithmetic control for executing the machine translation program Used in a computer equipped with a device,
A function of acquiring format information of the original text in the first language when the translation example that is a translation of the original text in the first language and the translation in the second language input from the input device is registered in the translation example database; ,
A function of converting the acquired format information of the first language into an internal code based on the format information internal code correspondence table;
Including it, the function of registering the translation example in the translation example database;
When an original text in the first language to be translated is input from the input device, a translation example similar to the original text in the first language is searched from the translation example database using the original text in the first language to be translated as a search key. Function to
A function for collating the format information of the original text of the translation example similar to the retrieved first language text of the translation target with the format information of the original text of the first language;
A function for creating a translation of the original text of the first language based on the translation example when the format information of the original text of the first language to be translated matches the format information of the original text of the translation example. A machine translation program for realizing

Images