JP2004310648A - Translation support system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To convert text data in data produced by use of software dealing with document data into a state allowing easy use in a translator such as a translation memory device and allowing execution of accurate translation, and to nearly automatically restore the translated text data to data in a state allowing use by original software. <P>SOLUTION: Character garbling is caused to a special character by a change of encoding of original text data in the original text data and primary original text data including the special character that is a function-controlling code to produce secondary original text data having a state allowing automatic retrieval of the special character as a garbled character portion, the special character in the data is automatically replaced by a specific character not imparting hindrance to parsing in the translator such as the translation memory device or is eliminated to produce tertiary original text data, and the tertiary original text data can be outputted to the translator. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
When translating text data in original data created by document design software such as DTP (Desk Top Publishing application) or other software, the present invention uses a translation memory device or a machine translation device. The present invention relates to a translation support system for performing translation smoothly.
[0002]
[Prior art]
In recent years, most of documents to be printed and documents exchanged through telecommunication lines such as the Internet use text design and plate data by using document design software such as DTP software to specify characters and layouts (printing patterns). ), And a document file containing text data and illustration data (referred to as content), character designation, characters, and illustration layout data (referred to as expression) is created. Depends on.
[0003]
By the way, when creating a document file in multiple languages, create a document in each language in advance, and add expressions such as layout data to contents such as text data and illustration data in each language corresponding to each language. There is a method of adding expressions, but in this method, in addition to the work of translation, the work of adding expressions for the number of languages must be performed, which is inefficient and impractical. It is not suitable for creating documents such as instruction manuals attached to products.
[0004]
Therefore, usually, a document file is created using a certain language, this document file is used as an original document file, text data (original text data) is extracted from the original document file, and another document is created based on the original text data. It is a common practice to create a translated document file by creating translated text data translated into the language of the original document and replacing it with the original text data, and using the expression in the original document file almost unchanged and almost unchanged .
[0005]
When translating text data, there is a primitive method in which the translator manually translates all documents and enters the translated text data.However, when the amount of text data to be translated is large, multiple However, there is a problem that the translation of the translated text data is not uniform because the translators share the translation work, and the texts of the translated text data have variations in terms and expressions.
[0006]
Therefore, the translation memory device is used to accumulate bilingual sentence data based on the relationship between the translated sentence created by the translator and the original sentence, and by using this accumulated data, translation work by multiple translators can be unified. In addition, it has become common practice to improve the translation quality of translation work by automatically generating translated sentences for expressions that are the same or similar to sentences that have been translated once. Attempts have been made to use such a translation memory device together with a machine translation device, which is said to have insufficient translation accuracy at present, to smoothly perform high-accuracy translation (see Patent Document 1).
[0007]
When translating text data in an original document file using a translation memory device or a machine translation device, the original document file is converted into an XML (Extensible) file using, for example, a function of DTP software (usually provided as a plug-in application). Output as a file with a standardized meta-language (meta tag) such as a Markup Language document (hereinafter referred to as a meta-language file). Translate
[0008]
Here, a translation memory device or a machine translation device analyzes a sentence by recognizing a sentence in units of one sentence, and replaces or changes the word order based on the analysis result. In the output meta language file, a meta language tag (meta tag) corresponding to the control code for the expression and a control code unique to the DTP software not included in the meta tag are inserted as special characters. It is possible to prevent the translation memory device and the machine translation device used for texts from recognizing as text, but the special characters set by the DTP software are confused with the text and recognized. Separation of sentences by translation memory device or machine translation device as a cause There are cases where a recognition not accurately performed.
[0009]
When a meta-language file described in the meta-language is displayed by a browser application, special characters depending on the DTP software are not displayed or are replaced with spaces, so that it is impossible to search or replace any of them. It is possible, and it is extremely complicated and impractical for a translator or editor to manually replace or delete special characters from a source file.
[0010]
Further, the above-described conventional technology relates to translation of text data in a document file handled by document design software, but is not limited to a document file, and includes a content including text data in a file and a display of the content. In a file containing a control code unique to the software, for example, a file of word processing software or spreadsheet software, if the text data in the file needs to be translated, the same problem as a document file occurs.
[0011]
In the case of word processing software, a control code depending on the software is removed, or such a control code is converted into a standardized control code such as an HTML (Hyper Text Markup Language) tag and output. It is thought that there are not many problems as many as the document files by the above-mentioned document design software because many of them have functions. However, files created by utilizing the functions depending on the software are not compatible with other formats such as HTML documents. If the file is output as a file, the function unique to the software cannot be sufficiently exhibited, and the same problem as the document file of the document design software described above occurs.
[0012]
[Patent Document 1]
JP-A-2002-278964 (pages 3 to 13, FIGS. 1 and 2)
[0013]
【Purpose】
An object of the present invention is to easily use text data in data created using document design software such as DTP software or other software for translation work using a translation memory device or a machine translation device. Can be converted to a state that allows accurate translation, and the translated text data is almost automatically converted to data that can be used with original software such as document design software, that is, data that contains expressions such as layout. It is an object of the present invention to provide a support system that can restore data in a dynamic manner.
[0014]
Configuration of the Invention
The present invention focuses on the fact that, when encoding and converting document data, a specific character in the document data is displayed as a meaningless character, and the content of the document cannot be read, a phenomenon called garbled characters occurs. It is configured.
[0015]
In other words, usually, when the document is garbled, the contents of the document cannot be read.Therefore, there have been many studies and proposals on how to prevent the garbled character from occurring. A completely new technology that replaces software-dependent special characters included in document data with specific characters that do not hinder translation work, or removes unnecessary special characters by causing garbledness. It was made based on a philosophical idea.
[0016]
Thus, the translation support system according to claim 1 of the present invention encodes original text data and primary original data including special characters, which are various function control codes depending on certain application software, by encoding the original text data. By causing the special characters to be garbled by the change, secondary text data in a state where the special characters can be automatically searched as garbled portions is generated, and the special characters in the data are translated memory device or machine translation. Tertiary original data is generated by automatically replacing or removing specific characters that do not interfere with the syntax analysis in a translation device such as a device, and the tertiary original data is returned to the original encoding system as it is or to the original encoding system. It has a configuration that allows output.
[0017]
In the translation support system according to a second aspect of the present invention, the primary source data is configured by meta language data in which a standardized meta language is added to source text data.
[0018]
The translation support system according to claim 3 of the present invention, wherein the primary original data is created by a document design apparatus running document design software as the application software, and the special character dependent on the document design software of the apparatus. And meta-language data obtained by adding a standardized meta-language to original text data.
[0019]
In the translation support system according to a fourth aspect of the present invention, when the translation device replaces the original text data in the secondary original data with the translated text data and inputs the primary translated data in which the specific character is left. The specific character left in the translation data from the primary translation data or from the primary translation data returned to the original encoding method is replaced with the original special character, and the secondary translation data for the application software is used. The configuration is such that it is output as
[0020]
In a translation support system according to a fifth aspect of the present invention, the secondary translation data is configured by meta-language data obtained by adding a standardized meta-language to translated text data.
[0021]
7. The translation support system according to claim 6, wherein the secondary translation data includes the special character that depends on the document design software of a document design device on which the document design software that is the application software runs. And meta-language data to which a meta-language is attached.
[0022]
【Example】
Hereinafter, embodiments of the translation support system according to the present invention will be described in detail based on specific examples shown in the accompanying drawings.
FIG. 1 shows a specific example of the system configuration. In FIG. 1, 1 denotes a support system according to the present invention, 2 denotes a document design device, and 3 denotes a translation memory device.
[0023]
The support system 1 includes first and second input / output interface units 4 and 5 for inputting and outputting data between the document design device 2 and the translation memory device 3, and the document design device 2 and the It performs encoding conversion of data input from the translation memory device 3 and various function control codes depending on application software for document design (hereinafter, referred to as document design software) running on the document design device 2. A central processing unit 6 for performing search / replacement between a special character and a predetermined specific character, an encode data storage unit 7 for storing the data for encoding conversion, and a correspondence relationship between the special character and the specific character. A character replacement data storage unit 8 for storing data is provided. , These two storage portions 7 and 8 is adapted to be able to perform the data update of change of additional or existing data of the new data from a separate input device via the central processing unit 6 either.
[0024]
Note that the support system 1 is configured by an electronic computer such as a personal computer, but the encoding change and the character search / replacement can be performed by using word processing software installed in the electronic computer, such as Microsoft Word (trade name) of Microsoft Corporation or a text editor. Alternatively, a function of an appropriate browse software can be used.
[0025]
The document design device 2 includes document design software such as DTP, for example, Adobe inDesign (product name), which is a commercially available application software of Adobe Systems Incorporated. The document design device 2 includes a main processing unit 2a that runs the software and document design software. Document data in a format unique to the software and document data (hereinafter referred to as meta-language data) to which a meta-language is added according to the XML (Extensible Markup Language) standard standardized by the W3C (World Wide Web Consortium). And a conversion processing unit 2b for performing mutual conversion and input / output of the data.
[0026]
Specifically, the document design device is configured by installing document design software on an electronic computer such as a personal computer, and the conversion processing unit 2b is software plugged into the document design software, for example, XMedia by Adobe Systems. (Merchandise name), and the input / output of the meta-language data is performed in an XML file in which the meta-language data is filed.
[0027]
The translation memory device 3 includes a control tag processing unit 3a, an analysis processing unit 3b, and a bilingual sentence database 3c. The control tag processing unit 3a includes a meta tag which is a control code in meta language data input from the outside. The original text data is identified, the parsing unit 3b analyzes the syntax of the original text data, refers to the data stored in the bilingual sentence database 3c, and selects an appropriate bilingual sentence in the parsing unit 3b. The bilingual sentence candidate is presented to the translator, and if no bilingual sentence is stored, the new bilingual sentence input by the translator is stored as new bilingual sentence data in the bilingual sentence database 3c.
[0028]
The translation memory device 3 is also constructed by installing translation memory software, for example, Trados (trade name) from a computer such as a personal computer, as in the case of the document design device 2 described above. Manages the bilingual sentence database 3c and analyzes, for example, WinAlign (trade name) and Translator's Workbench (trade name) of the Trados plug-in to the software, the analysis processing unit 3b, and Tag Editor (trade name of the Trados) ) Is used as the control tag processing unit 3a.
[0029]
As described above, the support system 1, the document design device 2, and the translation memory device 3 can use an electronic computer such as an individual personal computer, or a single electronic device in which software corresponding to each device is installed. It can be a computer, or it can be used in a network system in which a plurality of computers are interconnected. In any case, the computer system can be used between the support system 1, the document design device 2, and the translation memory device 3. Mutual data is transferred as a file or memory data in an appropriate format.
[0030]
Next, a specific example of the translation work in the translation support system of the present invention configured as described above will be described.
FIG. 2 shows that the support system 1 generates tertiary original data usable by the translation memory device 3 from the original document data (for example, English as shown in FIG. 2) created by the document design device 2. It is a flowchart showing a process up to this point.
[0031]
Since the special characters used in the original document data depend on the document design software as described above, in the following specific example, the above document data is created by Adobe inDesign (product name) of Adobe Systems Incorporated. It is assumed that the primary original data is output as an XML document file by XMedia (trade name) of Adobe Systems Incorporated.
[0032]
In FIG. 2, the original first step display (A1) shows the display state of the document created by the document design software on the document design device. In the document data, in addition to the original text displayed on the screen, special characters for various controls are hidden, and among these special characters, uniquely correspond to XML control tags (XML tags) as meta tags. Is automatically converted into an XML tag by the conversion processing unit 2b, and unspecified special characters are left in the data as they are, and thereafter, a specific encoding method, for example, an encoding standardized by the International Standards Organization (ISO) A file (XML file) is output as primary original data based on the UTF-8 standard.
[0033]
The primary text data output from the document design device 2 is processed by the central processing unit 6 of the support system 1, specifically, Microsoft's Microsoft Word (trade name) which is browsing software installed in the support system 1, for example, word processing software. ), An XML tag is displayed as in the original sentence second step display (A2).
[0034]
When the primary text data is changed to an encoding method of the Shift-JIS standard standardized (standardized) by the Japanese Industrial Standards (JIS) from the UTF-8, for example, by changing the encoding method, secondary text data is generated. Although the XML tag is displayed as it is as in the original third step display (A3), garbled characters that can be automatically searched are generated in the original text.
[0035]
FIG. 3 shows a specific example of the correspondence between garbled characters and special characters that occur when the encoding is changed from UTF-8 to Shift-JIS. In FIG. 3, a column of "original character string" is displayed on the document design device, a column of "UTF-8" is a special character included in the primary original data, and a column of "Shift-JIS" is after encoding conversion. And the column of “specific characters” indicate specific examples of the replacement character string of the garbled character.
[0036]
In FIG. 3, the line of “category A” is a character that needs to be replaced, and the line of “category B” needs to be recognized as a segment break of the original text in the translation memory device 3. Similar to "Category A", it is a character that needs to be replaced, and must not be recognized as a segment break of the original text in the translation memory device 3. Lines of "Category C" need to be replaced. The restoring process described below is not necessary. The character of "category C" is restored to a curled character or a character string by the function (or specification) of the document design device 2, and the line of "category D" Means that only the deletion is required, and neither replacement nor restoration is necessary. A.
[0037]
The garbled character generated as described above is searched by the support system, and the garbled character searched for is replaced with a garbled character in the column “Shift-JIS” and a replacement character string in the column “specific character” in FIG. The garbled character is replaced or removed with the character string of the specific character based on the correspondence relationship between the XML tag and the original text and the specific character as shown in the original fourth step display (A4) shown in FIG. Tertiary original data usable in the translation memory is generated and displayed.
[0038]
Note that the replacement character string of “specific character” shown in FIG. 3 is merely an example, and any character string that does not hinder the syntax analysis in the translation memory can be set. Yes, and it is also possible to use a single character or a character such as a symbol without being a so-called “character string”. Further, in FIG. 3, there is no character string between the start tag and the end tag of XML for the character garbled using the line feed symbol or the dot at the beginning of the line classified as “category A” as the original character “<R>”. Use of so-called empty tags such as </ R> and <D></D> has the advantage that it is easy to grasp the correspondence with garbled characters, and also controls the annotation text (REM text) and the like. It is also preferable to use a tag that is not performed or an empty element tag that has no content as a replacement character string.
[0039]
FIG. 4 shows that the support system 1 generates secondary translation data that can be used by the document design device 2 from the document data (for example, French text as shown in FIG. 4) created by the translation memory device 3. It is a flowchart showing a process up to this point.
[0040]
The tertiary original data is subjected to translation work such as syntax analysis and replacement with a bilingual sentence by the translation memory device 3 to become translated (for example, French text as shown in FIG. 4) text data and supported by the translation memory device 3. Output to the system 1 (translation first step display (B1)).
[0041]
The output data from the translation memory device 3 is primary translation data, which is the same encoding as that input to the translation memory device, that is, Shift-JIS in this embodiment, and includes an XML tag which is a meta tag. It is XML data and contains a specific character.
[0042]
Next, in the primary translation data described above, a specific character in the data is restored (re-replaced) as a garbled character of "Shift-JIS" (translation second step display (B2)). Note that, at this time, the restoring process that is not required for the categories C and D in FIG. 3 is not performed.
[0043]
For the data after the restoration processing, the encoding is changed from, for example, Shift-JIS to UTF-8, and the garbled character restored by this encoding conversion is returned to the original special character and output to the document design device 2 as secondary translation data. (The translated sentence third step is displayed (B3).
[0044]
In the document design device, the XML tag of the secondary translation data is returned to a special character which is a control code unique to the design device, or the special character included in the secondary translation data is used as it is to convert the translation text data. The content is displayed based on a predetermined expression together with other contents such as a plate (translation fourth step display (B4)).
[0045]
Note that display and printing in the document design device require manipulation and editing in the document design such as character length and forced line breaks when the original language is replaced with the translated language.
[0046]
In the above-described embodiment, a description has been given based on a specific example in which the translation memory device is used for the translation work from the original sentence to the translated sentence.However, the translation memory device and the machine translation device may be used in combination, In some cases, a machine translation device is used alone instead of a translation memory device.
[0047]
Although the encoding method has been described based on the relationship between “UTF-8” and “Shift-JIS”, another encoding method may be used instead of one encoding method, or two other encoding methods may be used. The encoding method may be used.In these cases, the garbled characters caused by the change of the encoding method are naturally different from those described in the present embodiment, but the correspondence between the special characters and the garbled characters is changed. This can be applied by registering in advance in the support system according to the encoding method to be used.
[0048]
【The invention's effect】
As described above, according to the translation support system of the present invention, garbled characters generated by encoding conversion of document data are intentionally generated, so that special characters depending on software included in the document data hinder translation work. Based on a completely new technical idea of being replaced with a specific character or removing unnecessary special characters, data in data created using document design software such as DTP software or other software is used. Text data can be converted to a state that can be easily used for translation work using a translation memory device or a machine translation device and can perform accurate translation, and the translated text data can be converted to a document design software or the like. Available in the original software or layer It can be restored almost automatically to the state of the data including the expression of bets like.
[0049]
Therefore, when translating a text in a document produced by the document design software, the time required for editing the document design accompanying this translation can be greatly reduced.
[Brief description of the drawings]
FIG. 1 is a configuration diagram showing an embodiment of a support system according to the present invention together with a relationship with other devices.
FIG. 2 is a flowchart showing a process of converting original text data of a document design device into data that can be used by a translation memory device.
FIG. 3 is a table showing a specific example of the correspondence between garbled characters and special characters that occur when the encoding method is changed from UTF-8 to Shift-JIS.
FIG. 4 is a flowchart showing a process until the translation data from the translation memory device is converted into data that can be used by the document design device.
[Explanation of symbols]
Reference Signs List 1 support system 2 document design device 3 translation memory device 4, 5 input / output interface unit 6 central processing unit 7 encoded data storage unit 8 character replacement data storage unit

Claims (6)

By changing the original text data and the primary text data including special characters, which are various function control codes depending on a certain application software, into the special characters by changing the encoding method of the original text data, Generates secondary text data in which special characters can be automatically searched for as garbled parts, and converts the special characters in the data to a specific one that does not hinder parsing in a translation device such as a translation memory device or a machine translation device. A translation support system in which tertiary original data is generated by automatically replacing or removing characters, and the tertiary original data can be output as it is or to the original encoding method to the translation device. 2. The translation support system according to claim 1, wherein the primary original data is meta-language data in which a standardized meta-language is added to original text data. 3. The primary text data is created by a document design device on which the document design software serving as the application software operates, and includes the special characters depending on the document design software of the device. The translation support system according to claim 1, wherein the translation support system is meta-language data to which is attached. When the translation device replaces the original text data in the secondary original data with the translated text data and inputs the primary translated data in which the specific character remains, the primary translated text data or the primary translated data is converted from the primary translated data. 4. The method according to claim 1, wherein the specific character remaining in the translation data from the original encoding method is replaced with an original special character and output as secondary translation data for the application software. Translation support system. 5. The translation support system according to claim 4, wherein the secondary translation data is meta-language data obtained by adding a standardized meta-language to translation text data. 6. The secondary translation data is meta-language data including the special character depending on the document design software of the document design apparatus on which the document design software serving as the application software operates, and having a meta-language added to the translated text data. Item 5. A translation support system according to item 4.

Images