JPH07200588A - Translation device - Google Patents

Abstract

PURPOSE:To provide a translation device capable of analyzing the role of a parenthesis part provided in a sentence at the outside of parentheses. CONSTITUTION:This device is constituted of an input part 4 for inputting the sentence provided with the parenthesis part, a storage part 6 for storing data relating to a dictionary, syntax rules and tree structure conversion rules, a translation module 5 for translating an input sentence inputted from the input part 4 based on the data stored in the storage part 6 and a display part 3 for displaying the input sentence and a translated sentence.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a translation device, and more particularly, to a translation device that requires natural language analysis in machine translation, a sentence understanding system, or the like.

[0002]

2. Description of the Related Art In a general machine translation apparatus, a method of translating a sentence with parentheses into parentheses and outside parentheses has been proposed (Japanese Patent Laid-Open No. 63-136262).
No. 4-283864).

Furthermore, the role of the parentheses outside the parentheses is
In some cases, the system analyzes the relationship between the word immediately before the parenthesis and the central word within the parentheses (Japanese Patent Laid-Open No. 3-1.
27264).

[0004]

However, it is still difficult for the current language processing technology to analyze the role of the parentheses outside the parentheses with a machine, and accuracy cannot be expected. In addition, this method can deal only with the usage (commentary) of supplementing the previous word.

It is an object of the present invention to provide a translation device capable of analyzing the role of parentheses in a sentence outside the parentheses.

[0006]

Means for Solving the Problems The above-mentioned objects are as follows: an input unit for inputting a sentence including parentheses, a storage unit for storing data relating to a dictionary, grammar rules and tree structure conversion rules, and the storage unit. In a translation device including a translation module that translates an input sentence input from the input unit based on stored data, and a display unit that displays the input sentence and the translated sentence, Sending means for detecting the parentheses and sending the inside and outside of the parentheses separately to the translation module;
This is achieved by providing a discrimination means for discriminating the role of the parenthesis portion into a plurality of parenthesis types according to the parenthesis type with respect to the parenthesis external processing.

[0007]

[Operation] The sending means detects the parentheses of the input sentence and sends the inside and outside of the parentheses separately to the translation module, and the distinguishing means changes the role of the parentheses with respect to the external parentheses processing into plural types depending on the type of parentheses. Distinguish.

[0008]

Embodiments of the translation apparatus according to the present invention will be described in detail below.

The translation processing apparatus according to the present embodiment has an input unit for inputting a sentence, a storage unit for storing a dictionary, a grammar rule and a tree structure conversion rule as data, and a data stored in the storage unit. Translation module that translates the sentence input from the input section based on the input section, the display section that displays the original sentence or the translated sentence, and the parentheses of the input sentence are detected and the parentheses and the outside of the parentheses are translated separately. The processing unit to send to the module, the judgment unit to divide the role of parentheses into several types according to the type of parentheses in the processing outside the parentheses and to perform the parsing process and the translation process, and the user instructs the role of the parentheses Definition section,
A footnote display part that turns the annotation part into a footnote etc. at the time of printing, a footnote setting part that indicates the condition to turn to the footnote etc., a parenthesis type designation part that indicates the type of parentheses used for the translated text, open parenthesis, closing parenthesis It is composed of a parenthesis processing designating section for instructing the processing when there is no correspondence and a translation method designating section for instructing the translation method when the translated words inside and outside the footnote brackets are the same.

Here, by providing the definition section, the user can
It is possible to define the type of parenthesis and its role for each.

Further, by providing the footnote display section, when the parentheses are used in the meaning of the annotation and the translated text in the parentheses is too long, it is possible to turn them into a footnote when printing.

Further, by providing a footnote setting section,
It is possible to specify for each user the conditions for turning into footnotes and the position where the annotation part sentence is placed, such as footnote or endnote.

By providing the parenthesis type designation section, since the JIS parentheses and the ASCII parentheses do not correspond 1: 1 when translating the parentheses, the user must designate the parentheses type to be used in advance. Is possible.

By providing the parenthesis processing designating unit, it is possible to designate, for each user, processing of parentheses in which the opening parenthesis and the closing parenthesis do not correspond.

By providing the translation method designating unit, it becomes possible to designate the translation method for each user when the translated words inside and outside the parentheses in the case of an annotation are the same.

In general, parentheses are used in various meanings in natural language.

In order to uniquely determine this, common sense and understanding of the scene are indispensable, and automation is difficult.

Therefore, the system prepares processing for several usages so that the user can specify which usage.

Table 1 shows the types of brackets.

[0020]

[Table 1]

There are various types of brackets as shown in Table 1. Generally, in Japanese, the brackets shown in the upper part of Table 1 are used.

On the other hand, a table of parenthesis types and their roles (role correspondence table, when the original text is in Japanese) as shown in Table 2 is prepared.

[0023]

[Table 2]

As shown in Table 2, parentheses 1 to 4 are highlighted, parentheses 5 are annotations, parentheses 6 are annotations (untranslated), parentheses 7 are replacements, and 8 are substitutions. Parentheses shown are examples, parentheses shown as 9 are nouns, parentheses shown as 10 are unused, and parentheses shown as 11 are citations.
The parentheses shown in 2 indicate proper nouns, and the parentheses shown in 13 indicate noun treatment (no translation).

In the present embodiment, when the user specifies the role, the correspondence table is fixed, and the type of the parenthesis of the original sentence is used by the user in accordance with the role of the parenthesis part distinguished by the translation process. Can be changed, or the correspondence table can be changed by the user.

Table 3 can be considered as an example of the role of the parentheses.

[0027]

[Table 3]

Sentence 1. In, "taihen" means emphasis.

Sentence 2. In, (is the capital of Japan) represents an annotation.

Sentence 3. In, [Yamato] represents a comment (untranslated).

Sentence 4. In, [vegetable] represents replacement.

Sentence 5. In, {banana, mandarin orange, apple} represents an example.

Sentence 6. <For the future of Japan>
Represents noun treatment.

Sentence 7. At

[Thin] means noun treatment (no translation).

Sentence 8. In, "I like her" is a quotation.

Sentence 9. At

[Ogawa] represents a proper noun.

By defining the roles shown in Table 3 in the definition section, the subsequent analysis process or translation process can be changed, and more accurate translation can be performed.

Regarding the annotations and examples that supplementally explain the previous word in the role of parentheses, the length of the translation inside the parentheses becomes too long (in the case of Japanese-English translation, It tends to be longer), and may be difficult to read.

In such a case, it is possible to make it easier to read by turning the annotation section to a special position such as a footnote by the footnote display section.

The condition for turning to the special position can be set by the footnote setting section using the translation parenthesis translation method selection table shown in Table 4, so that the user's preference can be utilized.

[0041]

[Table 4]

In the case of Japanese-to-English translation, since there are more types of parentheses that can be handled by Japanese JIS than ASCII parentheses that can be used in the translated sentence, the parentheses corresponding to the original sentence may not be used.

On the other hand, whether to output the JIS as it is,
Since it is possible to set whether or not different types of parentheses are selected from ASCII by the parenthesis type designation unit, it is possible to save the trouble of post-processing.

Non-corresponding parentheses can be considered in various cases as shown in Table 5 which exemplifies non-corresponding parentheses.

[0045]

[Table 5]

Sentence 10. Shows a case where it is a symbol, and sentence 11. Shows the case where the parentheses are made up of a plurality of sentences, and sentence 12 shows the case of a correspondence error.

Since the usage of this is largely determined by the input text, the accuracy can be improved by the user's specification.

In the case of an annotation, the translations inside and outside the parentheses may be the same.

For example, for "word processor" and "word processor" in Table 6 which exemplify the case where the translated words are the same in the annotations,
If the translation word "Word Processor" is attached to both, the translation text becomes redundant.

[0050]

[Table 6]

Since the user can specify whether the parentheses are not translated or both of them are translated by the translation method designating unit, post-processing can be omitted.

An embodiment of the translation device according to the present invention will be described below.
The details will be described.

FIG. 1 is a flow chart for explaining the processing operation of the translation apparatus according to this embodiment, FIG. 2 is a block diagram of an embodiment of the translation apparatus of the present invention, and FIG. 3 is a translation module block. Figures are shown respectively.

The translation apparatus shown in FIG. 2 has a CPU 1, a main memory 2, a CRT 3, and a keyboard (input section 4).
And a translation module 5 for returning a target language when a source language is given (details are shown in FIG. 3), a dictionary for translation, grammatical rules and tree structure conversion rules. It comprises an external storage unit 6 and a data bus 7.

Here, the input section 4 is provided with a start key for driving the apparatus, a numeric key, a character key, and other keys for executing translation.

The operation of the translation apparatus having the above configuration will be described in detail below.

The source language input from the input unit 4 under the control of the CPU 1 is sent to the translation module 5, and the result is displayed on the display 3.

As shown in FIG. 3, the translation module 5 has a dictionary lookup morpheme analysis unit 8 for dividing the input source language original sentence into Japanese morpheme strings, and an external memory for receiving the Japanese morpheme strings. A syntactic analysis unit 9 for performing syntactic analysis using the grammatical rules stored in the unit 6, and a syntactic analysis result for the syntactic structure of the target language using the tree structure conversion rule stored in the external storage unit 6. It is composed of a conversion unit 10 for converting, and a generation unit 11 for generating a sentence in the target language from the syntactic structure of the target language.

Next, referring to FIG. 1, the processing operation of this embodiment when translating will be described below by taking Japanese-English translation as an example.

In step S1, the translation device is activated and one sentence is read, and then the process proceeds to step S2.

In step S2, if the mode needs to be changed, it is changed, and the process proceeds to step S3.

Here, the mode change means that the user can instruct, 1) the setting of the role depending on the type of parentheses, that is, the rewriting of the table shown in Table 2, and 2) the setting of the translation method of the annotation parentheses. That is, rewriting of the table shown in Table 4 3) Setting of the type of parentheses used for translation, that is, the table of Table 7 showing the table corresponding to parentheses (in the case of Japanese-English translation)
Bull rewriting

[0063]

[Table 7]

4) Setting of the processing method of non-corresponding parentheses, that is, rewriting of the table in Table 8 showing the processing selection table of non-corresponding parentheses

[0065]

[Table 8]

5) Setting of the processing method when the translated text is the same in the annotation, that is, rewriting of the table in Table 9 showing the processing selection table when the translated text is the same in the annotation

[0067]

[Table 9]

It is

Next, in step S3, it is checked whether parentheses are included in the input sentence.
9. If there are parentheses, go to step S4.

In step S4, it is checked whether the parentheses correspond to the open parentheses and the closed parentheses. If they do not correspond, the processing of step S13 (parentheses non-corresponding processing) is performed. The process of (type determination process of parentheses) is performed.

In step S6, the role is judged using the table shown in Table 2 and the subsequent processing is controlled to
Go to step S7.

In step S7, it is checked whether or not the inside of the parentheses needs to be translated differently. If it needs to be translated separately, the process proceeds to step S8. If not, the process proceeds to step S9.

Here, Tables 2 and 3 are those that require separate translation.
“Annotation”, “replacement”, “exemplification”, “noun treatment”, and “quotation”.

On the other hand, those which do not require separate translation are shown in Table 2.
These are “emphasis”, “commentary (untranslated)”, “noun treatment (untranslated)”, and “proper noun” shown in Table 3.

In step S8, translation processing is performed on the inside of the parentheses.

Here, it does not matter if the outside of the parentheses in step S9 is also used as the translation processing algorithm. Also,
There is no need to specify the algorithm.

In step S9, the translation processing outside the parentheses is performed (when there is no parentheses, the normal translation processing is performed).
The translation algorithm in this case need not be specified. If there are no parentheses, steps S10 and S11 may be considered to be plain.

Step S9 for performing translation processing outside parentheses
So, you don't have to specify the translation algorithm,
It is necessary to add the function for processing the parentheses.

In the case of "emphasized", the same translation as in the case without parentheses is performed, and the highlighted word (word with parentheses)
Translate the translated word with the parentheses indicated in Table 7.

In the case of "comment" and "comment (non-translation)",
Remove the parentheses and translate.

The word immediately before the parentheses is attached with an attribute that the parentheses of the comment are included, and is used in step S10 for synthesizing the translation result inside and outside the parentheses.

In the cases of "replacement" and "exemplification", the parentheses are removed and translated.

Similar to the comment, an attribute to that effect is added to the immediately preceding word and used in step S10.

Furthermore, in the case of "replacement", it is possible to add a semantic analysis to determine whether or not the word immediately before the parenthesis is replaced to make a semantic meaning, which contributes to the translation quality.

That is, even if the word in the parenthesis is a multi-word meaning word, the word meaning suitable for the meaning of the sentence outside the bracket is correctly selected.

"Exemplary" is useful for identifying the meaning of a multi-word meaning word and contributes to the translation quality by utilizing the fact that the immediately preceding word and the words in parentheses have the same or similar semantic features.

In the case of "noun treatment" and "noun treatment (non-translation)", the parenthesis is removed and the noun as a dummy is analyzed at the position of the parenthesis.

In the case of "quoting", the parenthesis is removed and the verb as a dummy is analyzed at the position of the parenthesis.

In the case of "proper noun", it contributes to the translation quality by having a proper noun dictionary separately or by searching the dictionary with a proper noun attribute and using it.

In other words, the explanation of sentence 9 in Table 3 is "Ogawa"
, It is confirmed that the former is used as a proper name of a person, and the latter is commonly used as a meaning of "small river", and the ambiguity can be resolved. The parentheses in this case are instructions to the system, and the parentheses themselves need not be translated.

As described above, the type of parentheses may be assigned to the instruction to the system, for example, dependency designation, parallel designation, non-translation designation and the like.

Using the results of steps S10, S8 and S9, the translation results inside and outside the parentheses are combined to create the whole translated sentence. The parentheses used are the tables shown in Table 7.

In the case of "emphasized", no work is performed.

In the case of "comment" and "comment (non-translation)",
The translated sentence inside the parenthesis (original sentence in the case of no translation) is inserted immediately after the translated word of the word immediately before the parenthesis (giving attribute in step S9).

In the case of "replacement" or "exemplification", the translated word inside the parenthesis is inserted immediately after the translated word of the word immediately before the parenthesis (giving attribute in step S9).

In the case of "noun treatment" or "noun treatment (non-translation)", the translation inside the parenthesis is placed at the position where the translation of the noun as a dummy should be placed.

In the case of "quoting", the translation inside the parenthesis is placed at the position where the translation of the verb as a dummy should be placed.

In the case of "emphasized", no work is performed.

When there are no parentheses, no work is performed even when the parentheses are treated as symbols and errors are treated.

Then, in step S11, if there is a comment (which may be extended to an example), the table shown in Table 4 is checked and the translation position at the time of printing is added as an attribute. This is used when printing.

Also, after checking the table shown in Table 9, the processing when the translated words inside and outside the parentheses are the same is determined and the translated sentence is changed.

Then, in step S12, it is determined whether or not there is another sentence. If there is another sentence, the process returns to step S1, and if there is no other sentence, the process ends.

In step S13, when the parentheses are not compatible, when the table 8 is checked and treated as a symbol, the same label as that when a symbol other than the parentheses (#, etc.) is given is given, and the analysis processing is performed. And perform translation processing.

For example, in a system in which a symbol is treated as a noun, the same processing as when a noun comes is performed. In sentence 10 of Table 5, the same processing as "add a punctuation mark before a parenthesis (noun) in the sentence" is performed. In the case of sentence 12 in Table 5 in which the parentheses consist of multiple sentences, first, when the first sentence with the opening parenthesis is processed, only the back of the opening parenthesis is translated and the translated sentence is translated into the parenthesized sentence. Remember. The front part is stored as the original text. And continue reading the sentence.

If the sentence does not include the closing parenthesis (second sentence in sentence 11), it is translated, and the translated sentence is additionally stored as the translated sentence inside the parentheses. The same process is repeated until the sentence containing the closing parenthesis is encountered, and in the sentence with the closing parenthesis (the third sentence in sentence 11), the front of the parentheses is translated, and the translated sentence is additionally stored as a translated sentence inside the parentheses.

The part behind the closing parenthesis is regarded as the outside of the parenthesis together with the part before the opening parenthesis which is remembered, and the role shown in the table of Table 2 is assumed as in the case where the opening parenthesis and the closing parenthesis are included in one sentence. Check and perform the corresponding processing.

Since sentence 11 is shown as an example of "quoting",
Performs processing that regards parentheses as verbs. In the mode for pointing out a mistake in correspondence (in the case of sentence 12 in Table 5), that fact is pointed out.

The control of the above-mentioned operation is all performed by the CPU 1.
Is performed by a control signal from. According to this embodiment,
Parenthesis, which is frequently used in natural language, is processed extremely finely, and processing suitable for each usage is performed, so higher quality translation results can be obtained, and post-processing labor is also reduced. Is possible.

[0109]

INDUSTRIAL APPLICABILITY In the translation device, the parentheses of the input sentence are detected, and the inside and outside of the parentheses are translated separately.
The parentheses are processed according to the type of parentheses, and the parentheses that are frequently used in natural language can be processed in detail. Thus, the translation processing can be appropriately performed, the translation result can be made high quality, and the post-processing labor can be reduced.

[Brief description of drawings]

FIG. 1 shows a flowchart for explaining a processing operation of an embodiment of a translation apparatus according to the present invention.

FIG. 2 shows a block diagram of an embodiment of a translation device according to the present invention.

FIG. 3 shows a block diagram of a translation module.

[Explanation of symbols]

 1 CPU 2 main memory 3 display 4 keyboard 5 translation module 6 external storage unit 7 data bus 8 dictionary lookup morpheme analysis unit 9 syntax analysis unit 10 conversion unit 11 generation unit

Claims (7)

[Claims] 1. An input unit for inputting a sentence including parentheses,
A storage unit that stores data relating to a dictionary, a grammar rule, and a tree structure conversion rule, and a translation module that translates an input sentence input from the input unit based on the data stored in the storage unit. A translating device having a display unit for displaying an input sentence and a translated sentence, and transmitting means for detecting the parentheses of the input sentence and separately transmitting the inside and outside of the braces to the translation module. A translating device comprising: a discriminating means for discriminating the role of the parenthesis portion into a plurality of parenthesis types according to the parenthesis type with respect to the parentheses external processing. 2. The translation device according to claim 1, further comprising defining means for defining the type of parentheses and the role of parentheses for each user. 3. A footnote display means for displaying a footnote of the translated text in parentheses when the translated text in the parentheses is too long when the parentheses are used for the meaning of the annotation. The translation device described in. 4. The condition for setting the footnote to the translated text in parentheses and the condition for setting either the footnote or the endnote are
4. The translation device according to claim 3, further comprising footnote setting means for instructing each of them. 5. The translation device according to claim 1, further comprising a parenthesis type designating unit that allows a user to previously designate the type of parenthesis to be used when translating parentheses. . 6. The method according to claim 1, further comprising: parenthesis processing designation means for designating, for each user, parenthesis processing in which opening parenthesis and closing parenthesis do not correspond to each other. The described translation device. 7. The translation method designating means for designating, for each user, the translation method when the translated words inside and outside the parentheses in the case of an annotation are the same, and the translation method designating means is provided. The translation device described in.