JP7223450B2 - Automatic translation device and automatic translation program - Google Patents

Description

The present invention relates to an automatic translation apparatus for automatically translating an original text in a first language, for example Japanese, into a second language such as English, and a program used for the automatic translation.

Japanese sentences often contain delimited words surrounded by pairs of delimiters (hereinafter simply referred to as delimiters), for example parenthesized words surrounded by parentheses. For example, like "Tanaka", the reading of the kanji character is inserted after the kanji character of the person's name as a phrase with delimiters (brackets). Also, there are cases where the English katakana notation after a specific word is inserted as a word with delimiters (brackets), such as "patent". Also, when inserting the English katakana notation as a phrase with delimiters (parentheses) after the English notation, such as "Patent", and vice versa, like "Patent", English After the katakana notation, the English notation may be inserted as a phrase with delimiters (brackets).

Furthermore, as the phrases with delimiters, numbers and alphabets are used as phrases with parentheses, such as Task (1) and Task (a). (apples, strawberries, peaches, etc.) exemplify the preceding phrase by means of bracketed phrases, as in Socrates (ancient Greek philosopher) and Plato (427 BC-347 BC). There are some that use parenthetical phrases to explain the

When automatically translating a Japanese sentence into English, for example, the translation results for phrases with delimiters are mostly the above item numbers and symbols, examples of the preceding phrase, and explanations of the preceding phrase. not a problem. However, in the case of the kanji reading of the above person's name, the English katakana notation, etc., the translation result of the automatic translation is "Tanaka (Tanaka)" or "Patent (Patent)", and the same word is There was a problem that it was meaninglessly repeated and redundant.

Patent Document 1 (Japanese Laid-Open Patent Publication No. 11-3334) discloses an invention designed to solve the problem of phrases with delimiters when automatically translating a Japanese sentence into a foreign language. In the translation device disclosed in Patent Document 1, "information (correspondence table information) about knowledge such as bracket expression patterns, classification of bracket expressions, and actions for individual bracket expressions is stored, and It is possible to analyze parenthesis expressions and avoid translating redundant or unnecessary parts.

JP-A-11-3334

However, in the invention of Patent Document 1, it is necessary to create and store in advance correspondence table information in which parenthesis expression patterns, their parenthesis expression classifications, and actions for individual parenthesis expressions are associated with each other. It is necessary to analyze the parenthesis expression in the text and determine which of the correspondence table information the parenthesis expression pattern or classification corresponds to, which complicates the processing.

In addition, there is also the problem that processing cannot be performed for bracket expression patterns that are not registered in advance in the correspondence table information.

Parentheses have been explained above as delimiters, but the above problem is that the reading of kanji and English katakana notation are not used as delimiters other than parentheses, such as -○○-. This also occurs in the case of . In addition, in parentheses, when the words in parentheses are represented by ○, ', and various symbols such as "○" are included. In addition to the above-mentioned hyphens, delimiters other than parentheses include ·○·, |○|, *○*, ~○~, etc., and the same problem arises.

In view of the above problems, the present invention provides a method for automatically translating an original text in a first language such as Japanese containing phrases with delimiters such as parentheses into a second language. To provide an automatic translation device and an automatic translation program that can obtain a translation result that can recognize

In order to solve the above problems, the invention of claim 1 is
An automatic translation device that translates from a first language to a second language,
detection means for detecting delimited phrases enclosed in delimiters from the original text in the first language;
attribute determination means for determining an attribute of a word within the delimiter with respect to a word immediately preceding the delimiter;
outputting the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter; a translation means for adding and outputting annotations according to the attributes of the words in the delimiters;
To provide an automatic translation device comprising

According to the first aspect of the invention configured as described above, when the detection means detects a phrase with a delimiter such as parentheses, the translation means outputs the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter. is output, and at least the result of translation of the word within the delimiter is added with an annotation according to the attribute of the word within the delimiter for the word immediately preceding the delimiter determined by the attribute determining means.

As a result, when automatically translating an original text in a first language such as Japanese that includes words with delimiters such as parentheses into a second language, it is possible to easily grasp how the original text was. A translation result is obtained.

According to the present invention, when automatically translating an original text in a first language including phrases with delimiters such as parentheses into a second language, an automatic translation device capable of easily grasping what the original text was like; An automatic translation program can be provided.

1 is a block diagram showing a configuration example of a first embodiment of an automatic translation device according to the present invention; FIG. It is a figure used in order to demonstrate the operation|movement of 1st Embodiment of the automatic translation apparatus by this invention. FIG. 3 is a diagram showing part of a flowchart for explaining an example of the flow of processing operations of the first embodiment of the automatic translation device according to the present invention; FIG. 3 is a diagram showing part of a flowchart for explaining an example of the flow of processing operations of the first embodiment of the automatic translation device according to the present invention; It is a figure for demonstrating the example of the automatic translation result of 1st Embodiment of the automatic translation apparatus by this invention. FIG. 4 is a block diagram showing a configuration example of a second embodiment of an automatic translation device according to the present invention; It is a figure used in order to demonstrate the structure of one part of the structural example of 2nd Embodiment of the automatic translation apparatus by this invention. FIG. 10 is a diagram showing part of a flowchart for explaining an example of the flow of processing operations of the automatic translation device according to the second embodiment of the present invention; FIG. 10 is a diagram showing part of a flowchart for explaining an example of the flow of processing operations of the automatic translation device according to the second embodiment of the present invention; FIG. 10 is a diagram for explaining an example of an automatic translation result of the second embodiment of the automatic translation device according to the present invention; FIG. 10 is a diagram for explaining an example of an automatic translation result of another embodiment of the automatic translation device according to the present invention;

An embodiment of an automatic translation device according to the present invention will be described below with reference to the drawings. This automatic translation device may be composed of a dedicated device, but in this embodiment, for example, a personal computer, a tablet type terminal, and a high-performance mobile phone terminal called a smart phone are equipped with the program for the automatic translation device according to the present invention. consists of installing an embodiment of The following embodiment assumes that the first language is Japanese and the second language is a foreign language such as English.

[First Embodiment]
FIG. 1 is a block diagram for explaining a configuration example of a first embodiment of an automatic translation device according to the present invention. In the block diagram of FIG. 1, in order to facilitate understanding of the embodiment of the automatic translation apparatus according to the present invention, a part (main part) of the processing executed by the automatic translation program is also shown as a functional block. be.

The automatic translation apparatus of this embodiment provides a memory 102, a display interface 103, an operation input unit interface 104, an external input interface 105, and an original text memory 106 via a system bus 100 to a control unit 101 composed of a computer (processor). , a translation condition reception unit 107, a morphological analysis unit 108, a translation execution unit 109, a translation dictionary storage unit 110, a translation result memory 111, and a delimiter-attached phrase processing unit 112 are connected to each other.

The memory 102 stores programs and data for supporting the processing executed by the control unit 101 . Programs include programs for automatic translation.

A display 121 such as an LCD (Liquid Crystal Display) is connected to the display interface 103 .

Further, the operation input unit interface 104 is connected to an operation input unit 122 configured by a touch sensor arranged over the display screen of the LCD constituting the display 121 . On the display screen of the display 121, software buttons (icon buttons) included in various input screens are displayed. 104 to the control unit 101 . The control unit 101 performs processing according to an operation input from the operation input unit 122 . Note that the operation input unit is not limited to one using a touch sensor as in this example, and may of course be configured with a keyboard or a group of operation buttons. Also, voice input using a microphone may be used.

The external input interface 105 is composed of, for example, a USB memory interface, a card-type memory interface, or the like. Also, the external input interface 105 may be composed of an optical disc drive. In this embodiment, through the external input interface 105, original Japanese text data stored in a USB memory, an optical disc, or the like is taken in. FIG.

The original text memory 106 stores Japanese original text data input by the user through the operation input unit 122 and Japanese original text data captured through the external input interface 105 .

The translation condition accepting unit 107 stores the translation conditions designated by the user through the operation input unit 122 (in addition to direct input as character input, including selection designation). In this case, the translation conditions include the specification of the original Japanese text to be translated, and the foreign languages such as English, French, German, Spanish, Russian, Chinese, Korean, etc., from the specified Japanese original text. In addition to specification of the foreign language to which language the words are to be translated, selection specification of whether or not to perform redundancy elimination processing for words with delimiters is included. Here, the process of removing redundancy for words with delimiters means that when the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter are the same, the delimiter and the This is a process for not including the translation result of the words and phrases within the delimiters in the final automatic translation result.

The automatic translation apparatus of this embodiment does not always remove the redundancy for words with delimiters, but performs it when the user selects to "remove the redundancy". to The reason why the user can select and specify in this way is that there are cases where the translation must be a word-for-word translation (word-for-word translation, letter-for-word translation) of the original text without allowing free translation. It is what I did. Note that a list of words immediately preceding the delimiter and words with delimiters may be presented so that the user can select and specify from the list.

The morphological analysis unit 108 divides the original Japanese text to be translated into a string of morphemes (minimum units having the meaning of the language), and converts each part of speech (nouns, verbs, conjunctions, delimiters such as parentheses, punctuation marks, etc. ). The morphological analysis unit 108 sends analysis result information consisting of a sequence of morphemes of the analysis result and respective parts of speech to the translation execution unit 109 . The morphological analysis unit 108 also detects symbols such as delimiters and punctuation marks as morphemes, and thus has the function of delimiter-attached phrase detection means.

The translation execution unit 109 uses the analysis result information of the original Japanese text from the morphological analysis unit 108 and the bilingual dictionary between the foreign language and the Japanese stored in the translation dictionary storage unit 110 to translate the specified text. Automatically translates the original Japanese text into the specified foreign language. In the translation dictionary storage unit 110, bilingual dictionaries with Japanese words and phrases are stored in advance for foreign languages such as English, French, German, Spanish, Russian, Chinese, and Korean. . The bilingual dictionary may exist on a server on the Internet and be updated as needed.

A translation result memory 111 temporarily stores the automatic translation result generated by the translation execution unit 109 . Then, the automatic translation result in the translation result memory 111 is displayed on the display screen of the display 121 through the display interface 103 and presented to the user.

In this embodiment, the delimiter-attached phrase processing unit 112 is activated when a user selects and designates, via the operation input unit 122, to perform redundancy elimination processing for delimiter-attached phrases. Then, the delimited phrase processing unit 112 receives the analysis result from the morphological analysis unit 108 and the translation result from the translation execution unit 109, and receives the translation result of the phrase inside the delimiter of the delimited phrase and the delimited phrase. The translation result of the word immediately preceding the symbol is compared to determine whether the two match.

When the delimiter-attached word/phrase processing unit 112 determines that the translation result of the delimiter-internal word of the delimiter-attached word matches the translation result of the word immediately preceding the delimiter, the translation result memory 111 Delimiters and phrases within the delimiters are deleted from the automatic translation results stored in (that is, phrases with delimiters are not output as automatic translation results). Here, instead of not outputting only the translation results of the words in the delimiters that match the translation results, the reason why the delimiters such as parentheses () are not output as the translation results is that the translation results of the words in the delimiters are not output. This is to avoid an automatic translation result that leaves only delimiters such as parentheses that do not have any words in the delimiters.

Further, in this embodiment, when the delimiter-attached phrase processing unit 112 determines that the translation result of the delimiter-inside delimiter phrase does not match the translation result of the phrase immediately preceding the delimiter-attached phrase, , the delimiters and the translation results of the words in the delimiters are included in the automatic translation results stored in the translation result memory 111 as they are.

In the configuration of the automatic translation apparatus shown in FIG. A configuration of the software processing function unit can be obtained by executing it.

[Processing operation of the automatic translation device of the first embodiment]
In the first embodiment, prior to the start of automatic translation, the user selects a Japanese original text file to be translated from among the Japanese original text files stored in the original text memory 106. is selected and specified. Then, the selected Japanese original text is displayed in the original text display field 1211 of the display screen 121D of the display 121, as shown in FIG.

On the display screen 121D of the display 121, as shown in FIG. 2, input fields 1212 and 1213 are displayed for accepting the selection and specification of translation conditions by the user. An input field 1212 is a field for receiving a selection designation of a foreign language as a second language in which a translation result is to be obtained from the original Japanese text. , French, German, Spanish, Russian, Chinese, and Korean. FIG. 2 shows a case where English is selected and specified.

An input column 1213 is a column for receiving a selection input as to whether or not to perform the above-described redundancy removal processing. If not, enter a check in "No".

In the automatic translation apparatus of this embodiment, when the selection and designation of the translation conditions described above is completed by the user's operation input as described above, a translation start button (icon button) 1214 and a translation stop button (icon button) are displayed. 1215 is displayed on the screen and can be pressed by the user. In the case of this embodiment, selection and designation of translation conditions are input collectively. You can also Of course, after selecting and designating, the selected and designated portion can be previewed, and after the user confirms the preview, the selection and designation can be completed.

Alternatively, instead of inputting a selection as to whether or not to perform the redundancy removal process, the redundancy removal process may be performed by default, and only when the redundancy removal process is not to be performed, an input to that effect may be made. On the contrary, it is also possible to input the fact only when the redundancy elimination process is not performed by default and is to be performed.

When the user presses the translation start button 1214, the automatic translation processing operation in the automatic translation apparatus of the first embodiment is started. Further, when the user presses the translation stop button 1215, the automatic translation processing operation is stopped without being started.

3 and 4 are flowcharts showing an example of the flow of the automatic translation processing operation in the automatic translation apparatus of the first embodiment having the above configuration. Hereinafter, an example of the automatic translation processing operation in the automatic translation apparatus of the first embodiment will be described with reference to the flow charts of FIGS. 3 and 4. FIG. 3 and 4, the control unit 101 performs the functional processing of the translation condition reception unit 107, the morphological analysis unit 108, the translation execution unit 109, and the punctuation mark processing unit 112. , a case where software processing is performed by executing a program stored in the memory 102 will be described.

First, the control unit 101 determines whether or not the user presses the translation start button 1214 through the operation input unit 122 to instruct the start of automatic translation (step S1). When the start instruction is given, the control unit 101 checks the input field 1213 for redundancy removal processing (step S2), and determines whether or not "perform redundancy removal processing" is selected and specified (step S2). S3).

When it is determined in step S3 that "no redundancy removal processing" is selected and designated, the control unit 101 performs morphological analysis on the original Japanese text, executes word-for-word translation (step S4), and outputs the translation result. is written in the translation result memory 111, and the translation result is displayed in the translation result display column 1216 (see FIG. 2) of the display screen 121D of the display 121 (step S5).

Then, the control unit 101 determines whether or not the translation has been completed up to the end of the original Japanese text to be translated (step S6). Repeat the subsequent steps. Further, when it is determined in step S6 that the translation has been completed up to the end of the original Japanese text to be translated, the control section 101 terminates this automatic translation process.

When it is determined in step S3 that "perform redundancy elimination processing" is selected and designated, the control unit 101 performs morphological analysis on the original Japanese text (step S11 in FIG. 4). Next, from the result of the morphological analysis, the control unit 101 determines whether or not a phrase with a delimiter has been detected (step S12).

When it is determined in step S12 that no phrase with a delimiter has been detected, the control unit 101 executes word-for-word translation (step S13), writes the translation result into the translation result memory 111, and The translation result is displayed in the translation result display field 1216 (see FIG. 2) of the display screen 121D (step S17).

Next, the control unit 101 determines whether or not the translation has been completed up to the end of the original Japanese text to be translated (step S18). The processing after S11 is repeated.

Then, when it is determined in step S12 that a phrase with a delimiter has been detected, the control unit 101 translates the original Japanese text word by word, and translates the word within the delimiter and the word immediately before the delimiter. are compared (step S14), and it is determined whether or not the two translation results match (step S15). In step S15, if the translation result of the word within the delimiter and the translation result of the word immediately before the delimiter do not match, the control unit 101 converts the translation result of the word immediately before the delimiter to the parenthesized word. Each translation result is written in the translation result memory 111, and the translation result is displayed in the translation result display column 1216 of the display screen 121D of the display 121 (step S17).

Then, the control unit 101 determines whether or not the translation has been completed to the end of the original Japanese text to be translated (step S18). Repeat the subsequent steps.

If it is determined in step S15 that the translation result of the word within the delimiter matches the translation result of the word immediately preceding the delimiter, the delimiter and the translation result of the word within the delimiter are translated verbatim. It removes it from the translation result (step S16), writes only the translation result of the word immediately before the delimiter into the translation result memory 111, and displays the translation result in the translation result display field 1216 of the display screen 121D of the display 121. (Step S17).

Then, the control unit 101 determines whether or not the translation has been completed to the end of the original Japanese text to be translated (step S18). Repeat the subsequent steps. Further, when it is determined in step S18 that the translation has been completed to the end of the original Japanese text to be translated, the control section 101 terminates this automatic translation process.

[Specific example of automatic translation processing result in the first embodiment]
The operation in the automatic translation process according to the above-described first embodiment when "perform redundancy elimination process" is selected and specified will be further described with a specific example as shown in FIG. FIG. 5 shows an example of processing when the phrases with delimiters included in the original Japanese text shown in FIG. 2 are translated into English. In this example, the delimiters are all parentheses, but it goes without saying that the delimiters are not limited to parentheses, as described above.

The example of FIG. 5(A) is a case where the reading of kanji characters (person's name in the example of the figure) is described as a phrase in parentheses. As shown, when word-for-word translation is performed, the translation result immediately before the parenthesis and the translation result of the phrase inside the parenthesis are both "Tanaka" and match. Therefore, the control unit 101 sets only the parenthesis and the translation result, that is, "Tanaka" from which the parentheses ( ) and Tanaka are removed, that is, from which (Tanaka) is removed, as the automatic translation result.

In the example of FIG. 5(B), the English katakana notation for kanji is described as a parenthesized phrase. In this case, as shown on the right side of FIG. 5(B) , when word-for-word translation is performed, the translation result immediately before the parenthesis and the translation result of the phrase in the parenthesis are both "Patent" and match. Therefore, the control unit 101 sets only the parenthesis and the translation result, that is, "Patent" from which (Patent) is removed, as the automatic translation result.

In the example of FIG. 5(C), the explanation of the immediately preceding word/phrase is written as a parenthesized word/phrase. In this case, as shown on the right side of FIG. 5(C), When translation is performed, the translation result immediately before the parentheses and the translation result of the words in the parentheses usually do not match. Note that the example of FIG. 5(C) is a case where, as an explanation of the immediately preceding word/phrase, the word/phrase for specifying the immediately preceding word/phrase is a word/phrase with parentheses. In this case, the translation result immediately before the parenthesis and the translation result of the phrase in the parenthesis do not match, so the control unit 101 outputs the word-for-word translation with the parentheses as it is as the automatic translation result.

As a type of the example of FIG. 5(C), the origin of the stroke composition of kanji is explained as a note with brackets, or the explanation of traditional Chinese characters without a font is explained as a note with brackets. There is also For example, there are "medicine (easy to wear grass crown)" and "● (no crown on greed)".

As described above, according to the automatic translation apparatus of this embodiment, the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter are compared, and when both match, the delimiter and the delimiter are translated. Since the redundancy of the automatic translation result is removed by removing the word within the symbol from the automatic translation result, as in the conventional method described at the beginning, the types of words with delimiters and each type There is no need to store the corresponding correspondence table in the storage unit, and the redundancy can be removed from the automatic translation result by a simple process.

In addition, in the above-described first embodiment, the user can select and specify whether or not to perform the redundancy removal process according to the purpose of automatic translation. In the case of usage, it is convenient because word-for-word translation can be performed without performing redundancy removal processing.

[Second embodiment]
In the above-described first embodiment, no processing is performed on the translation result when the user selects and designates not to perform the redundancy removal process, and the words of the same translation result are simply replaced with delimiters. It was assumed to be continuous. However, as described above, even if the same translation result is repeated, the relationship between the word immediately before the delimiter and the word with delimiter is "Kanji and its reading", "Kanji and In addition to the corresponding English katakana notation, there are various aspects (classification=attribute) such as English and its phonetic notation (katakana notation). This can be categorized as a difference in the attributes of the delimited word relative to the word immediately preceding the delimiter.

When the same translation results are consecutive words with delimiters in the automatic translation results, the user can see what kind of relationship there was between the words immediately before the delimiters in the original text and the words with delimiters. It is useful to know.

Therefore, in the second embodiment, for the translation result when the user selects and designates not to perform the redundancy removal process, annotations corresponding to the difference in the attribute of the phrase with a delimiter for the word immediately before the delimiter are added. Make it possible to include it in the automatic translation result.

FIG. 6 is a block diagram showing a configuration example of the automatic translation apparatus of the second embodiment, and the same parts as those of the first embodiment in FIG. 1 are denoted by the same reference numerals. In the automatic translation apparatus of the second embodiment, in addition to the above-described processing units of the automatic translation apparatus of the first embodiment, the system bus 100 is connected to the delimiter-attached word/phrase attribute determination unit 113 and the table Information storage unit 114 is connected. A delimiter-attached phrase processing unit 112A is provided instead of the delimiter-attached phrase processing unit 112 of the first embodiment. Otherwise, the configuration is the same as that of the first embodiment.

As shown in FIG. 7, the table information storage unit 114 stores information of a correspondence table between the attributes of delimited words and phrases and examples of annotations corresponding to the respective attributes. More specifically, referring to the example of FIG. 7, the table information storage unit 114 stores an example of a delimiter-attached phrase, its attribute, and an annotation added to the translation result for that attribute in correspondence with each other. ing. In this example, a plurality of annotations (two annotations 1 and 2 in the example shown) are prepared as annotations to be added to the translation result, and as will be described later, the user can select which annotation to use. Selection can be specified.

Specifically, as shown in FIG. The attribute of the phrase with delimiters in "Tanaka" of No. 1 is "Kanji reading" immediately before, and example No. 1 of the phrase with delimiters. The attribute of the phrase with delimiters in “patent” of No. 2 is “English translation of kanji in katakana notation” immediately before. The attribute of the phrase with delimiters in “Patent (patent)” of No. 3 is “Japanese translation of English” immediately before. The attribute of the phrase with delimiters in "Patent" of No. 4 is "Japanese translation of English in katakana notation" immediately before. The attribute of the phrase with delimiter in 5 "Patent" is "English translation of kanji" immediately before.

In the example of FIG. 7, one of the annotations (annotation 1) is a combination of a symbol * and a number, and another annotation (annotation 2) is a translation result of the above attributes. translated into the language of Note that, in this example, the annotation 2 is a simplified translation of the above attributes to avoid lengthy and redundant annotations. Therefore, in the English example of annotation 2 in FIG. 2 to No. 5 are all "translation", but in order to distinguish them from each other, each "translation" is added with a numerical number of the word example.

The above examples of words with delimiters are examples in which the result of word-for-word translation is the same for the word immediately before the delimiter and the word within the delimiter. , the result of word-for-word translation may be a mismatch between the word immediately before the delimiter and the word within the delimiter. However, in this embodiment, it is possible to add an annotation to the translation result only when the result of word-for-word translation is the same for the word immediately before the delimiter and the word with the delimiter.

That is, in this embodiment, as shown in FIG. 7, the table information includes phrase examples No. For example 6 where the word immediately before the delimiter and the word within the delimiter do not match, the attribute is stored in the table information, but the annotation is not defined and stored. This is because, in this embodiment, no annotation is added to a delimiter-attached word/phrase example that does not match the word/phrase immediately before the delimiter and the word/phrase within the delimiter.

Note that FIG. 7 shows only the case of English as annotation 2, but in this embodiment, foreign languages such as English, French, German, Spanish, Russian, Chinese, Korean, etc. translated from Japanese It goes without saying that Annotation 2 may be prepared for all of .

In addition, in the second embodiment, when annotation 1 is selected and specified by the user and the "combination of symbol * and number" of annotation 1 is added to the result of automatic translation, the added annotation 1 When the user clicks on each of them, the attributes corresponding to the added annotation 1 are displayed in a pop-up, for example, by translating the attributes shown in FIG. 7 in the language of the translation result. Similarly, when annotation 2 is selected and specified by the user, when the user clicks on "translation numeric number", the attribute corresponding to "translation numeric number" of the added annotation 2 is displayed. , the translation of the attribute in FIG. 7 in the language of the translation result is configured to be displayed, for example, as a pop-up.

In the second embodiment, when the punctuation-marked word/phrase attribute determination unit 113 receives a selection designation from the user via the operation input unit 122 that the processing of redundancy elimination for the punctuation-marked words/phrases is not performed. , and is activated when an annotation is selected. When the delimiter-attached word/phrase attribute determination unit 113 is activated, when a bracketed word is detected as an analysis result of the morphological analysis unit 108, the attribute of the delimiter-attached word/phrase immediately before the delimiter is determined. judge. Then, the delimiter-attached word/phrase attribute determination unit 113 transfers the attribute information of the determination result to the delimiter-attached word/phrase processing unit 112A.

The delimiter-attached word/phrase processing unit 112A executes the same processing as the delimiter-attached word/phrase processing unit 112 of the first embodiment. When receiving a selection designation not to perform redundancy removal processing for delimited words and phrases, and when it is selected and designated to add annotations, the attribute of the determination result from the delimited word/phrase attribute determination unit 113 Upon receipt of the information, a process of adding annotations selected and specified by the user to the results of automatic translation is executed.

[Processing operation of the automatic translation device of the second embodiment]
Also in the second embodiment, prior to the start of automatic translation, the user selects and designates the file of the original text to be translated from among the files of the original Japanese text stored in the original text memory 106. In addition, it accepts a selection designation of a foreign language to be translated from Japanese, and also accepts a selection input as to whether or not to perform redundancy elimination processing for words with delimiters.

Furthermore, in the second embodiment, when it is selected and specified not to perform the redundancy removal process for the delimited words and phrases, the selection designation as to whether or not to perform the annotation addition process for the delimited words and phrases is accepted, If it is selected and specified to add an annotation to a phrase with a delimiter, a selection specification of which annotation, annotation 1 or annotation 2, is to be added is accepted.

Then, in the case of the second embodiment, when the above-described selection and designation of the translation conditions is completed based on the user's operation input, a translation start button (icon button) is displayed on the screen, and the translation start button can be used. When pressed by the user, the automatic translation processing operation in the automatic translation apparatus of the second embodiment is started.

8 and 9 are flow charts showing an example of the flow of the automatic translation processing operation in the automatic translation apparatus of the second embodiment having the above configuration. An example of the automatic translation processing operation in the automatic translation apparatus of the second embodiment will be described below with reference to the flow charts of FIGS. 8 and 9. FIG. 8 and 9, the processing of each step in the flow charts of FIGS. A case will be described where each function processing of the attribute determination unit 113 is software processing by executing a program stored in the memory 102 .

First, the control unit 101 determines whether or not the user presses the translation start button 1214 through the operation input unit 122 to give an instruction to start automatic translation (step S21). When the start instruction is issued, the control unit 101 checks the input field 1213 for the redundancy removal process (step S22), and determines whether or not "perform redundancy removal process" is selected and designated (step S22). S23).

When it is determined in step S23 that "perform redundancy elimination processing" is selected and designated, the control unit 101 executes the processing shown in the flowchart of FIG. 4, as in the first embodiment. make it Description is omitted here to avoid duplication.

Then, when it is determined in step S23 that "no redundancy removal processing" is selected and designated, the control unit 101 determines whether or not "add annotation" is selected and designated by the user. (Step S24).

When it is determined in step S24 that "no annotation" is selected and specified, the control unit 101 morphologically analyzes the original Japanese text and executes word-for-word translation (step S25). is written in the translation result memory 111, and the translation result is displayed in the translation result display column 1216 (see FIG. 2) of the display screen 121D of the display 121 (step S26).

Then, the control unit 101 determines whether or not the translation has been completed up to the end of the original Japanese text to be translated (step S27). Repeat the subsequent steps. Further, when it is determined in step S27 that the translation has been completed up to the end of the original Japanese text to be translated, the control section 101 terminates this automatic translation process.

When it is determined in step S24 that "add an annotation" is selected and specified, the control unit 101 morphologically analyzes the original Japanese text (step S31 in FIG. 9), and from the result of the morphological analysis, delimiter It is determined whether or not a sign-appended phrase is detected (step S32).

When it is determined in step S32 that no phrase with delimiters has been detected, the control unit 101 executes word-for-word translation (step S33), writes the translation result into the translation result memory 111, and The translation result is displayed in the translation result display field 1216 (see FIG. 2) of the display screen 121D (step S36).

Next, the control unit 101 determines whether or not the translation has been completed up to the end of the original Japanese text to be translated (step S37). The processing after S31 is repeated.

Then, when it is determined in step S32 that a phrase with a delimiter has been detected, the control unit 101 determines the attribute of the phrase with a delimiter (step S34). Then, the control unit 101 sequentially translates the word immediately before the delimiter and the delimiter-attached word word-for-word, and refers to the table information storage unit 114 according to the attribute determined in step S34 to translate the word within the delimiter. An annotation to be added to the result is determined according to the determined attribute and added (step S35). As described above, which of the annotations 1 and 2 is to be added is selected and designated by the user before the start of translation, so the selected annotation is added.

As described above, depending on the attribute of the phrase with delimiters, it may not be necessary to add an annotation.

Next, the control unit 101 writes the translation result of the word immediately before the delimiter, the translation result of the word with the delimiter, and the annotation added as necessary to the translation result memory 111, and The translation result is displayed in the translation result display field 1216 of the display screen 121D (step S36).

Then, the control unit 101 determines whether or not the translation has been completed to the end of the original Japanese text to be translated (step S37). Repeat the subsequent steps. Further, when it is determined in step S37 that the translation has been completed to the end of the original Japanese text to be translated, the control section 101 terminates this automatic translation process.

[Specific example of annotation addition]
FIG. 10 shows a specific example of annotation addition processing that is executed when "redundancy removal processing is not performed" in the automatic translation processing according to the second embodiment. FIG. 10 shows an example of processing when a phrase with a delimiter included in a Japanese original is translated into English. In this example, the delimiters are all parentheses, but it goes without saying that the delimiters are not limited to parentheses, as described above. Further, the comment 1 is added within the delimiter immediately after the term within the delimiter, and the comment 2 is added within the delimiter immediately after the term within the delimiter with [ ].

The example of FIG. 10(A) corresponds to the example No. of phrases with delimiters in FIG. 1, the attribute is the reading of kanji. In this case, as shown on the right side of FIG. It is added immediately after the phrase, and when the user selects and designates annotation 2, [translation 2] is added immediately after the phrase in parentheses.

Also, the example of FIG. 10(B) corresponds to the example No. of phrases with delimiters in FIG. 2, where the attributes are English translations of kanji in katakana notation. In this case, as shown on the right side of FIG. is added immediately after the phrase in parentheses, and when annotation 2 is selected and designated by the user, [translation 2] is added immediately after the phrase in parentheses.

In the second embodiment, although not shown in FIG. 10, when the user clicks "*1" or "*2" of annotation 1 displayed on the display screen, the corresponding delimiter is displayed. Phrase attributes are popped up, for example. Similarly, when the user clicks [translation 1] or [translation 2] of annotation 2 displayed on the display screen, the corresponding delimiter-attached phrase attribute is popped up, for example.

[Effect of Second Embodiment]
According to the automatic translation apparatus of the second embodiment, the same effect as that of the automatic translation apparatus of the first embodiment can be obtained, and in the case of not performing the redundancy removal process, the translation of the word immediately before the delimiter is Even if the result and the translation result of the words in the delimiters are the same, by adding annotations according to the attributes of the words with delimiters to the translation results of the words in the delimiters, There is an effect that it is possible to easily understand how the word with sign is related to the immediately preceding word, which is convenient.

Then, according to the automatic translation apparatus of the second embodiment described above, when the user designates an annotation added to the automatic translation result, detailed information about the attributes that the annotation means is available to the user There is also an effect that it is very convenient because it is presented to.

[Modification of Second Embodiment]
In the description of the second embodiment above, the annotation is added only to the words within the delimiter, but the annotation is added to both the word immediately before the delimiter and the word within the delimiter. By doing so, it is possible to show in more detail the relationship between the phrase with punctuation marks and the phrase immediately preceding it. For example, FIG. 11 shows an example of adding and displaying such an annotation, and shows an example of processing when a phrase with a delimiter included in a Japanese original is translated into English.

The example of FIG. 11(A) corresponds to the delimiter-added word/phrase example No. 1 in FIG. 1, the attribute is the reading of kanji. In this case, as shown on the right side of FIG. [kanji] is added immediately after, and [kana] is added immediately after the translation result "Tanaka" of the reading "tanaka" in the delimiter (in parenthesis). This makes it easy to understand that the words with delimiters in the original Japanese text were read in kanji.

Also, the example of FIG. 11(B) corresponds to the example No. of phrases with delimiters in FIG. 3, the attribute is a Japanese translation of English. In this case, as shown on the right side of FIG. is immediately followed by [English], and [Japanese] is added immediately after the translation result "Patent" of the Japanese translation "Patent" within the delimiter (in parentheses). As a result, it becomes easy to understand that in the original Japanese text, the words with delimiters are Japanese translations of English. Of course, for French translations of English, [French] is added, and for German translations, [German] is added.

[Other embodiments or modifications]
In the above-described second embodiment, the annotation is added to the automatic translation result only when the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter are the same. Even if the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter do not match, an annotation corresponding to the attribute determined by the delimiter-attached word/phrase attribute determination unit 113 may be added. good.

Also, if the words with delimiters are ruby with kanji characters that are difficult to read, the translation result of the words immediately before the delimiters and the translation results of the words within the delimiters will not match. , the ruby translation is unnecessary, so the translation result of the ruby may not be included in the automatic translation result.

In the above-described embodiment, a case has been described in which there is one phrase within the delimiter of a phrase with a delimiter. It can be applied even if there are multiple words within the delimiter, separated by punctuation marks such as . In this case, the translation result of the word immediately before the delimiter and the translation results of each of the multiple words within the delimiter are checked for match or mismatch, and the translation results of the multiple words within the delimiter are checked against each other. match or disagree with each other.

If the words in the delimiter are not delimited by punctuation marks, the entire word not delimited by the punctuation marks is treated as a single word, and the translation result matches the word immediately before the delimiter. , inconsistency is determined in the same manner as in the above-described first embodiment.

Then, if all of the translation results of the word immediately before the delimiter and each of the multiple words within the delimiter match, the translation results of the delimiter and the multiple words within the delimiter will be extracted from the automatic translation results. try to remove it.

Also, if the translation result of the word immediately before the delimiter does not match the translation results of the multiple words within the delimiter, but the translation results of the multiple words within the delimiter match each other, the delimiter is It is left as a translation result, and one matching translation result is left in the delimiter.

Furthermore, if the translation result of the word immediately before the delimiter and the translation results of the multiple words within the delimiter match not all of them, but only a part of the multiple words within the delimiter, the delimiter The symbol is left as the translation result, and the matching translation result within the delimiter is removed.

A specific example will be given for further explanation. For example, in the first embodiment, if "Patent (patent, patent)" is automatically translated without redundancy removal processing, the result of word-for-word translation is "Patent (Patent, Patent)", and the delimiter is All of the translation results of the immediately preceding word and each of the multiple words within the delimiter match. Therefore, when the redundancy removal process is performed, the phrase "Patent" in the delimiter, the "," between the two "Patent", and the delimiter ( ) are removed, that is, (Patent, Patent) is removed. and becomes "Patent". In the second embodiment, when there are a plurality of words in the delimiter, for example, according to annotation 1 in FIG. For example, "Patent(Patent[translation 3], Patent[translation 4])".

Also, for example, for "LCD (Liquid Crystal Display)", when automatic translation (word-for-word translation) is performed without redundancy removal processing, "LCD (Liquid Crystal Display; Liquid Crystal Display)" or " LCD (Liquid Crystal Display; LCD). On the other hand, when the redundancy removal processing of this invention is used, in the former case, the multiple words in the delimiter become "Liquid Crystal Display" with each other, so a match is determined, and duplication is avoided, resulting in a single translation result. ``Liquid Crystal Display'' and the punctuation mark ``;'' are removed from the translation result of the phrase in the delimiter so that the translation result is ``LCD (Liquid Crystal Display)''. In the latter case, a match between the word immediately preceding the delimiter and part of the multiple words within the delimiter is determined, the matched delimiter words "LCD" and ";" are removed, and the translation result is It becomes "LCD (Liquid Crystal Display)".

Furthermore, as another example, when ``Platon (427 BC-347 BC)'' is automatically translated without redundancy removal, ``Platon (B.C.427-B. C.347)”, but after the redundancy removal process, it becomes “Platon (B.C.427-B.C.347)”.

In the above-described embodiment, the case where the first language is Japanese has been described, but the first language is of course not limited to Japanese, and can be any of English, French, German, and Spanish. language, Russian, Chinese, Korean, etc. The second language is of course not limited to English, and may be Japanese, French, German, Spanish, Russian, Chinese, Korean, or the like.

In the above-described embodiment, the existence of words with delimiters is assumed. You can also get the same effect as In this case, another typeface, such as italic, can itself be considered a punctuated phrase.

The automatic translation device and the automatic translation program according to the present invention are not only applied to personal computers, tablet terminals, mobile phone terminals such as smartphones, but also wristwatch type terminals, eyeglass type terminals, car navigation systems, ticket vending machines, guides, etc. It is of course applicable to boards, electronic bulletin boards, electronic menus, digital signage, various display devices, and the like.

101...control unit 106...original text memory 107...translation condition reception unit 108...morphological analysis unit 109...translation execution unit 110...translation dictionary storage unit 111...translation result memory 112, 112A...with delimiter Phrase processing unit 113 Phrase attribute determination unit with delimiter,

Claims (19)

An automatic translation device that translates from a first language to a second language,
detection means for detecting delimited phrases enclosed in delimiters from the original text in the first language;
attribute determination means for determining an attribute of a word within the delimiter with respect to a word immediately preceding the delimiter;
outputting the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter; a translation means for adding and outputting annotations according to the attributes of the words in the delimiters;
Automatic translation device with. 2. The automatic translation apparatus according to claim 1, wherein the translation results of at least the words within the delimiters include words in the original text as translation results. The attributes determined by the attribute determining means are classified in advance into a plurality of attributes,
3. The annotation according to claim 1 or 2, wherein the annotation is added to at least the translation result of the word within the delimiter in a display mode indicating which classification attribute the annotation belongs to, and is output. automatic translation device. a storage unit that stores table information indicating a correspondence relationship between the annotation and each of the plurality of classified attributes;
means for indicating what the corresponding attribute is based on the table information in the storage unit when the annotation displayed attached to the translation result of at least the word within the delimiter is selected; 4. The automatic translation device according to claim 3, comprising: 5. The automatic translation apparatus according to claim 3, wherein the annotation is a translation of each of the plurality of classified attributes into the second language. Claims 3 to 5, wherein the annotation consists of a word meaning the translation of the second language and a number for distinguishing each of the plurality of classified attributes. The automatic translation device according to any one of the above. 7. The automatic translation apparatus according to any one of claims 3 to 6, wherein said annotation comprises a combination of a symbol and a number for distinguishing each of said plurality of classified attributes. The annotation is a translation of each of the plurality of classified attributes into the second language, and/or a word meaning a translation of the second language, and the plurality of classified attributes. A first annotation consisting of a number for distinguishing each of the attributes and a second annotation consisting of a combination of a symbol and a number for distinguishing each of the plurality of classified attributes are selected. is used for purposes of
8. The automatic translation apparatus according to any one of claims 3 to 7, further comprising means for receiving selection of whether to add said first annotation or said second annotation. the first language is Japanese, the second language is a foreign language other than Japanese,
The automatic translation according to any one of claims 1 to 8, wherein the word immediately before the delimiter is a kanji character, and the attribute of the word within the delimiter is the reading of the kanji character. Device. the first language is Japanese, the second language is a foreign language other than Japanese,
10. The method according to any one of claims 1 to 9, wherein the word immediately preceding the delimiter is the foreign language, and the attribute of the word within the delimiter is a Japanese translation of the foreign language. automatic translation device. the first language is Japanese, the second language is a foreign language other than Japanese,
11. The method according to any one of claims 1 to 10, wherein the word immediately preceding the delimiter is a Chinese character, and the attribute of the word within the delimiter is a katakana notation of a foreign language translation of the Chinese character. automatic translation device. the first language is Japanese, the second language is a foreign language other than Japanese,
12. The automated system according to any one of claims 1 to 11, wherein a word immediately preceding the delimiter is a Chinese character, and the attribute of the word within the delimiter is a foreign language translation of the Chinese character. translation device. The attribute determination means also determines the attribute of the word immediately before the delimiter,
The translation means adds an annotation corresponding to the attribute of the word immediately preceding the delimiter determined by the attribute determining means to the translation result of the word immediately preceding the delimiter and outputs the result. The automatic translation device according to any one of claims 1 to 12. A receiving means for receiving an instruction by a user as to whether or not to add an annotation to the phrase with delimiters,
When the receiving means receives a designation to process the words with delimiters, the translation means outputs at least the translation results of the words immediately preceding the delimiters with an annotation added. 14. The automatic translation device according to any one of claims 1 to 13. The method is characterized in that, when receiving a designation not to add an annotation to the phrase with punctuation marks, the phrase immediately before the punctuation mark and the phrase with punctuation marks are translated verbatim in the original text in the first language. 15. The automatic translation device according to claim 14. 16. The automatic translation device according to any one of claims 1 to 15, wherein the translation is performed by using a phrase in a different font from a predetermined font as the punctuation-added phrase. It is characterized by being applied to any of personal computers, tablet terminals, mobile phone terminals, wristwatch terminals, eyeglass terminals, car navigation systems, ticket vending machines, information boards, electronic bulletin boards, electronic menus, digital signage, and display devices. The automatic translation device according to any one of claims 1 to 16, wherein determining whether or not the translation result of the word immediately preceding the delimiter matches the translation result of the word within the delimiter, and if it is determined that they do not match, the translation of the word within the delimiter 18. The automatic translation device according to any one of claims 1 to 17, wherein no annotation is added to the result. To translate from a first language to a second language, a computer comprising:
detection means for detecting delimited phrases enclosed in delimiters from the original text in the first language;
attribute determination means for determining an attribute of a word within the delimiter with respect to the word immediately preceding the delimiter;
outputting the translation result of the word immediately before the delimiter and the translation result of the word within the delimiter; Translation means for adding annotations according to the attributes of words in delimiters and outputting them;
An automatic translation program to function as

Images