JPH0589163A - Machine translation system - Google Patents

Abstract

PURPOSE:To easily discriminate a character string consisting of a character type peculiar to the second language mixed in the document of the first language with the other character string in a translated sentence. CONSTITUTION:A machine translation device translating the sentence of given first language into the sentence of the second language which is different from the first language is provided with a selection means 6 discriminating the character type in the character string in the sentence of the first language and selecting the character string of the character type which is not peculiar to the first language, a processing means 7 processing the character string so that it can be discriminated when the character type of the selected character string is discriminated and it agrees with the character type peculiar to the second language, a translation means 5 translating the processed sentence of the first language into the sentence of the second language, and a discrimination support processing means 7 processing the character string applied to the processed character string so that it can be identified by the processing means in the translated sentence in such a way that it arises in a system which can discriminate it from the other character string in the translated sentence.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a machine translation apparatus for translating first language data into second language data.

[0002]

2. Description of the Related Art In a translation between language pairs having different language-specific character types, such as a translation from Japanese to English and a translation from English to Japanese, a document specific to the second language is unique to a document of the first language. When character strings of different character types are mixed, conventional machine translation technology does not make any special efforts to translate the character strings, so different characters that could be clearly distinguished from other character strings in the original document were recognized. There was a problem that the string was merged with other character strings in the translated sentence, and a translated sentence that was difficult to read (difficult to understand) was generated.

An example thereof will be described in the case that occurs when translating from Japanese into English. Suppose there is an English word alphabet string continue in a Japanese document. This is a verb as an English word, but as long as it exists in a Japanese document, it cannot have any syntactic function other than a noun phrase.

Therefore, when a sentence including such an English word is translated into an English sentence, the verb continue occurs at a position in the syntax that should function as a noun phrase without being naked.
As a result, the translated English text is very difficult (and sometimes misleading) to understand. A similar problem arises when translating from English to Japanese. For example, if a Japanese character string such as Hiragana, Katakana, or Kanji is found in an English document, the Japanese character string found in the English sentence is generated because there is generally no space between words in Japanese. Not only is it very hard to read because it is assimilated into another character string in the translated Japanese translation, but the translation that is generated is also not well-formed.

[0005]

As described above, in the conventional machine translation device, when translating between languages such as English and Japanese that have different language-specific character types, the first language document has If the character strings of the two language-specific character types are mixed, the character string will be assimilated with other character strings in the generated translation to the second language, and as a result, the translated text will be very difficult to understand. However, there was a problem that the format was not good.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a second language mixed in a document of the first language, which is a source of confusion. It is an object of the present invention to provide a machine translation device capable of easily distinguishing a character string composed of a specific character type from other character strings in a translated sentence.

[0007]

In order to achieve the above object, the present invention is configured as follows. That is, in a machine translation device for translating a given first language sentence into a different second language sentence, a character type not unique to the first language is determined by determining the character type in the character string in the first language sentence. Selecting means for selecting the character string of, and a processing means for distinguishing the character string of the selected character string and recognizing the character string when they match the character type peculiar to the second language,
A translation unit that translates the sentence in the first language that has been processed by the processing unit into a sentence in the second language, and a character string corresponding to a character string that is identifiable by the processing unit in the translated sentence. Is provided with a discrimination support processing means for processing so as to occur in a format distinguishable from other character strings in the translated text.

Secondly, a selection instructing means for selecting a desired format from among a plurality of characterization formats is provided, and the discrimination support processing means can be identified by the processing means in the sentence of the first language. A character string corresponding to the processed character string has a function of characterizing and outputting in a format designated by the selection instructing means so that it can be distinguished from other character strings of the second language in the translated document. ..

[0009]

In the present apparatus configured as described above, the text of the given first language is selected by the selection means to determine the character type in the character string in the text of the first language, and the character of the character type not unique to the first language is determined. A string is selected, and the processing means discriminates the character type of the selected character string, and if the character type matches the character type peculiar to the second language, processes the character string so that the character string can be identified. The sentence in the first language that has been subjected to such processing is translated by the translation means, and the discrimination support processing means translates the character string corresponding to the character string that is identifiable by the processing means in the translated text. Process to occur in a format that can be distinguished from other character strings in the sentence.

As a result, the character string composed of the character type peculiar to the second language included in the sentence of the first language is in a predetermined format or a specified format that can be distinguished from other character strings after being translated into the second language. It will be characterized and born.

As described above, according to the machine translation apparatus of the present invention, when translating, the character type discriminating means is not a character string peculiar to the first language mixed in the document of the first language, but from the character type peculiar to the second language. Is extracted, processed to be identifiable, then translated into a second language, and the discrimination support processing means for this translated sentence is a character string corresponding to the identifiably processed character string in the translated sentence. Characterize it so that it can be distinguished from the character string. Therefore, the character string is prominent in the translated text, and the translated text is easy to understand. Also, in order to distinguish the character string from other character strings, if there are multiple ways to realize the display format, such as italicized character format, enclosing format in quotation marks, emphasized character format, etc. By making it possible to specify which one to use, it is possible to absorb variations in user requirements and reduce the burden of calibration.

Therefore, according to the present invention, a character string consisting of character types unique to the second language mixed in the document of the first language, which is a source of confusion, is replaced with another character string in the translated sentence. A machine translation device that can be easily distinguished can be provided.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The embodiments of the present invention will be described below with reference to the drawings, taking the translation from Japanese to English as an example. In the machine translation device of the present invention, the input unit for inputting the data in the first language, the dictionary recording unit for recording the dictionary data necessary for translation, and the data in the first language input from the input unit are A translation unit that refers to the dictionary data and translates the data into a second language data, and an output unit that outputs the second language data obtained by the translation unit. The translation unit includes a character type discriminating unit that discriminates the character type of each character string in the original text of the first language, and the character type of the character string discriminated by the character type discriminating unit is not unique to the first language. When the character string is unique to the two languages, it is generated by using a means (discrimination support processing means) that can distinguish the character string from other character strings in the translated sentence.

Further, the discrimination support processing means included in the translation unit has a plurality of realization methods as candidates, and the user can select / designate which method is to be adopted from the input unit, and according to the designation by the user. The character string is generated by the realization method described above.

FIG. 1 shows a block diagram of an embodiment of the present invention. In FIG. 1, 1 is an input unit, 2 is a display unit, 3 is a document recording unit, 4 is a dictionary recording unit, 5 is a translation unit, 6 is a character type discrimination unit, 7 is a discrimination support processing unit, and 8 is a translation control unit. is there.

Of these, the input unit 1 instructs the translation control unit 8 to translate the original sentence (in this case, the Japanese sentence).
Is used to specify which realization method is to be used as a discrimination support generation means, which will be described later in detail, and a keyboard or the like is usually used. Further, with respect to the input of the original text, input by OCR (optical character reading device) or the like can be considered.

The display unit 2 translates the Japanese sentence input by the input unit 1 and the English sentence translated by the translation unit 5 into the translation control unit 8.
It is intended to be displayed under the control of
It can display underline, italic, etc., for example, CR
T displays, plasma displays, liquid crystal displays, etc. can be used. Further, it also has a function of displaying the candidates of the realization method as the discrimination support generation means.

Under the control of the translation control unit 8, the document recording unit 3 is capable of recording and reading each of the documents in the first language and the second language, and can read the Japanese-English machine translation system. For example, in addition to being able to record each data of the Japanese sentence which is the first language and the English sentence which is the second language independently, it is possible to associate the data of the Japanese sentence with the data of the English sentence as the translation result. It is possible to record.

The dictionary recording unit 4 stores the words of the first language, Japanese, as knowledge information necessary for the translation processing in the translation unit 5.
It records information such as a vocabulary dictionary in which idioms and English translations, which are the second language, are associated with each other and records information such as headwords, parts of speech, conjugations, concepts, and grammatical functions in both Japanese and English. The information described in can also be recorded as information.

Under the control of the translation control unit 8, the translation unit 5 translates the Japanese sentence, which is the data of the first language, into the English sentence, which is the data of the second language, while referring to the information in the dictionary recording unit 4. It is a thing. In addition to the conventional translation unit, it has a character type discrimination unit 6 and a discrimination support processing unit 7.

Of these, the character type discriminating unit 6 includes a character string consisting of alphabetic characters peculiar to the second language English included in the document of the first language Japanese, and the alphabetic characters such as symbols and numbers. This is to distinguish and extract a mixed character string.

When the character string extracted by the character type discriminating unit 6 is generated in the translated sentence, the discrimination supporting processing unit 7 discriminates the character string from other character strings in the translated sentence. Generation processing is performed using the processing means. At this time,
If the user has previously designated the method of realizing the discrimination support from the input unit 1, the processing according to the designated content is performed.

The translation control unit 8 controls the entire apparatus.

FIG. 2 is a view showing an example of the key arrangement of the input section 1. The input unit 1 includes a character key 201 for inputting Japanese / English sentences including numbers and symbols, a translation instruction key 202, an edit key 203, and a function key 2 as function keys.
04, cursor keys 205, and other keys.

FIG. 3 is a diagram showing an example of the screen layout of the display unit 2. The input original text is displayed in the original text display area 301 on the left side of the screen. The translated text obtained as a result of the translation processing is displayed in the translated text display area 302 on the right side of the screen. The edit area 303 at the top of the screen can be used to display information necessary for various edits.

FIG. 4 is a flow chart for explaining the operation of the translation controller 8. The operation of this system will be described with reference to this flowchart.

The translation control unit 8 specifies the document to be translated among the documents recorded in the document recording unit 3,
The operation is started when a translation start command is input from the input unit 1. Before starting translation, first let the user specify how to process and generate alphabetic character strings that appear in the original Japanese text of the translation target document or character strings mixed with alphabets and symbols and numbers in the translated text. (Step S401).

This is because the translation control unit 8 controls the display unit 2 to display the translation specific character string display format setting screen, and the user operates the screen with the input unit 1 to select and specify the translation control unit. 8 recognizes the setting contents, and advances the processing in accordance with the setting contents.

FIG. 5 is an example of a translation specific character string display format setting screen presented to the user at that time. In this embodiment, as options, "generate with italic characters", "generate with bold type characters", "enclose in quotation marks", "underline.", Etc. are presented. Of course, "no designation"
It can also be

Here, if the user specifies something, the translation control unit 8 sets the content as the discrimination support generating means (step S402). On the other hand, if no designation is made, "generate in italics" is set as the default value in the discrimination support processing means (step S40).
3).

Since these processes determine the generation processing means for the alphabetic character string in the original text and the character string in which the alphabet, the symbols, and the numbers are mixed, the translation control unit 8 determines the data of the first language ( Data A) is read from the document recording unit 3 (step S
404). Of course, this can also be performed by inputting a document directly from the input unit 1 and translating it.

The translation control unit 8 first takes out the first sentence (first paragraph) of the data A (step S405), transfers it to the translation unit 5, and causes the translation unit 5 to perform a translation process to obtain a translation result. Is obtained (step S406). The translation process will be described later with reference to FIG.

This process is repeated until the last sentence (final paragraph) of the data A is completed (steps S407 and S4).
08) and when the translation is completed up to the last sentence of the data A, the data A and the translation result are recorded in the document recording unit 3 (step S409), and the process is terminated.

FIG. 6 shows a step S406 in FIG.
That is, it is a flowchart for explaining the flow of processing of the translation unit 5.

The translation unit 5 starts its operation when the document data to be translated is transferred from the translation control unit 8. When the document data to be translated is transferred, in the translation unit 5, the character type discriminating unit 6 first discriminates the character type of each word in the original sentence (step S601), and the word consisting of the alphabetic character string which is the focused character string. , And a word with a mixture of alphabets and symbols / numbers are extracted, and the extracted word is marked.

Since the transfer method is used in the present embodiment, the translation unit 5 thereafter performs the analysis process for translation (step S602), the conversion process (step S603), and the generation process (step S604) to obtain the translation result. Obtain the data at the stage before outputting. Then, step S601.
The discrimination support processing unit 7 performs the discrimination support processing according to the set content of the discrimination support that has been set on the words and phrases marked in (step S605).

In the discrimination support process, if the previous setting is "generate in italic characters", the character string is generated to be italic characters, and "generate in bold type characters" is set. If so, the character string is generated in bold font, and if "enclose in quotation marks." Is set, the character string is double-byte.
It is generated in the form surrounded by quotation marks such as quotation marks, single quotation marks and various parentheses.
If "underline." Is set, the character string is generated with an underline.

FIG. 7 shows an example of generation when these options are set. The original text (Japanese text) to be translated in this case is "There are highlighted."
The translation results are shown in Fig. 7 (1) when "generate in italics" is set, and in (2) in Fig. 7 when "generate in boldface." , (3) of FIG. 7 is set when “enclose in quotation marks.”
(4) is an example of a translated sentence when “underline.” Is set.

The display unit 2 receives the processed data generated by the translation unit 5 under the control of the translation control unit 8, and the character string of interest is characterized as shown in FIG. The generated translated sentence is displayed, so that the English words described in alphabet in the original sentence are displayed on the display unit 2 so that this portion can be clearly understood in the translated sentence. Of course, even when printing is performed by the printer, it is possible to perform printing while maintaining such characteristics under the control of the translation controller 8.

As described above, the present apparatus is a machine translation apparatus for translating a given sentence in a first language into a sentence in a different second language, and selects and designates a desired form from a plurality of characterizing forms. A selection instructing means, a selection means for determining a character type of a character string in a sentence of the first language to select a character string of a character type not unique to the first language, and a character type of the selected character string; When it matches the character type peculiar to the second language, the processing means for processing the character string so that it can be identified, the translation means for translating the text of the first language processed by this processing means into the text of the second language, A character string corresponding to a character string that has been processed to be identifiable by the processing means in the translated text is processed to occur in the selected and specified format that can be distinguished from other character strings in the translated text. Discrimination support It is constructed by providing a means.

Then, the text of the given first language is selected by the selecting means to determine the character type of the character string in the text of the first language, and the character string of the character type not unique to the first language is selected. When the character type of the selected character string is discriminated and if it matches the character type peculiar to the second language, the character string is processed so as to be identifiable. Then, the sentence in the first language subjected to such processing is translated by the translation means, and the discrimination support processing means determines that the character string corresponding to the character string processed by the processing means in the translated text is identifiable. The processing is performed so that it occurs in a format that can be distinguished from other character strings in the translated text.

As a result, the character string composed of the character type unique to the second language, which is included in the sentence of the first language, can be distinguished from other character strings after being translated into the second language in a predetermined format or a designated format. It will be characterized and aroused.

Therefore, the character string using the character type of the second language contained in the sentence written in the first language is characterized so that it can be distinguished in the output when translated into the second language. Can be displayed.

The present invention is not limited to the above embodiment. For example, in the above-described embodiment, as a candidate for the method for realizing the discrimination support processing means, "generate in italics", "generate in bold font", "enclose in quotation marks", "underline However, the present invention is not limited to these, and many other variations of typefaces and fonts can be applied, and inversion characters, high brightness, and shading can also be used. Also, the types of quotation marks may be subdivided and selected. Furthermore, it is also possible to apply the discrimination support process only when the character string actually exists as an English word, and the user can determine which type of character string the main discrimination support process is applied to. You may make it specify.

In addition to this, in the present embodiment, the translation processing is described based on the transfer method performed by the procedures of "analysis", "conversion", and "generation", but the translation processing can also be performed by the pivot method. This is because the present invention performs the discrimination support process immediately before the output of the translated sentence, and is not affected by the translation process method.

In short, the present invention can be implemented with various modifications without departing from the spirit thereof.

[0047]

As described above, according to the present invention, the unique character types of the individual languages of the language pair to be translated, such as the translation between Japanese and English, are different, and If a character string consisting of a character type unique to the second language is mixed in the language document, the character string can be generated in a format that is easily distinguishable from other character strings in the translated text of the second language. , Which will not cause confusion and will be able to generate translated text that is easy to understand, and will also be a format that suits the user's preference from among a wide variety of formats as an expression format that makes it easy to distinguish other character strings. It is also possible to provide a machine translation device that is remarkably easy to use.

[Brief description of drawings]

FIG. 1 is an overall block diagram of an embodiment of the present invention.

FIG. 2 is a diagram showing an example of a key layout of the input unit shown in FIG.

FIG. 3 is a diagram showing an example of a layout of the display unit shown in FIG.

FIG. 4 is a flowchart showing an example of a processing flow of a translation control unit used in an embodiment of the present invention.

FIG. 5 is a diagram showing a display example of a display unit in FIG. 4 when the user specifies the content of the generation processing means of the alphabetic character string occurring in the Japanese original sentence.

FIG. 6 is a flowchart showing details of translation processing in FIG.

FIG. 7 is a diagram showing an example of data for explaining processing according to an embodiment of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Input part, 2 ... Display part, 3 ... Document recording part, 4 ... Dictionary recording part, 5 ... Translation part, 6 ... Character type discrimination part, 7 ... Discrimination support processing part, 8 ... Translation control part.

Claims (4)

[Claims] 1. A machine translation device for translating a given sentence in a first language into a sentence in a second language which is different from the given sentence, wherein the character type in the character string in the sentence in the first language is determined to be the first language. A selecting means for selecting a character string of a non-unique character type; a processing means for distinguishing the character type of the selected character string and processing the character string so that the character string can be identified when the character type is unique to the second language; The translation means for translating the sentence in the first language that has been processed by the processing means into the sentence in the second language, and the character string corresponding to the character string that is distinguishably processed by the processing means in the translated text are A machine translation device, comprising: a discrimination support processing means for processing so as to occur in a format distinguishable from other character strings in a translated sentence. 2. A machine translation device for translating a given sentence in a first language into a sentence in a second language which is different from the first sentence, wherein the character type in the character string in the sentence in the first language is determined to be the first language. A selecting means for selecting a character string of a non-unique character type; a processing means for distinguishing the character type of the selected character string and processing the character string so that the character string can be identified when the character type is unique to the second language; The translation means for translating the sentence in the first language that has been processed by the processing means into the sentence in the second language, and the character string corresponding to the character string that is distinguishably processed by the processing means in the translated text are Discrimination support processing means for processing so as to occur in a form distinguishable from other character strings in the translated text, and outputting the translation result processed by the discrimination support processing means, and characterizing the character strings. Output Machine translation apparatus, characterized in that which is configured by providing a means. 3. A machine translation device for translating a given sentence in a first language into a sentence in a different second language, and a selection instruction means for selecting a desired one of a plurality of characterization forms, Selection means for determining a character type in a character string in a sentence of the first language and selecting a character string of a character type not specific to the first language, and a character type specific to the second language by determining the character type of the selected character string And a translation means for translating the sentence in the first language processed by the processing means into a sentence in the second language, and A machine translation apparatus comprising: a discrimination support processing means for processing a character string corresponding to a character string processed in a distinguishable manner by the processing means so as to be processed in the selected and specified format. 4. A machine translation device for translating a given sentence in a first language into a sentence in a different second language, and a selection instruction means for selecting and designating a desired form among a plurality of characterization forms. Selecting means for determining a character type in a character string in a sentence of the first language and selecting a character string of a character type not specific to the first language; and a character type of the selected character string for determining a character type of the second language When the character string matches the character string, a processing unit that processes the character string in a distinguishable manner, a translation unit that translates the sentence in the first language processed by the processing unit into a sentence in the second language, and a translation unit in the translated sentence. In the above, the character string corresponding to the character string processed by the processing means is outputted to the discrimination support processing means for processing so as to be in the selected and designated format, and the translation result processed by the discrimination support processing means. In addition, the machine translation device is characterized by being provided with means for characterizing and outputting the character string.