JPH06318229A - Machine translation device - Google Patents

Abstract

PURPOSE:To provide the machine translation device equipped with functions for automatically setting a translation environment corresponding to the kind of a sentence and for defining the set of the translation environment every sentence kinds arbitrarily by a user himself by selecting the sentence kind of a source sentence to be translated. CONSTITUTION:When a source language is inputted, a translation module 5 translates it and outputs a target language. The source language inputted from a keyboard 4 is sent to the translation module 5 under the control of a main CPU 1. The translation module 5 translates the inputted source language into the target language by using a dictionary, grammar rule and tree structure transformation rule or the like stored in a memory 6. The result is temporarily stored in a main memory 2 and displayed on a CRT 3. Since the translation environment corresponding to the sentence kind is automatically set only by selecting the sentence kind, when translating the same kind of sentence, only by selecting the common sentence kind (translation environment set), an equal translation level is maintained at all times.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a machine translation device,
More specifically, it has a function to automatically set the translation environment according to the sentence type by selecting the sentence type of the original sentence to be translated, and the translation environment of the sentence type prepared in advance by the system (translation environment In addition to a set), the present invention relates to a machine translation device having a function that allows the user to define a translation environment (translation environment set) for each sentence type.

[0002]

2. Description of the Related Art Some machine translation systems perform translation in a translation environment specified when translating a document. For example, a "machine translation system" disclosed in Japanese Patent Laid-Open No. 2-224064 discloses a translation environment storage unit for storing a set translation environment in order to perform a translation process using the set translation environment, and the translation. By providing the translation environment resetting means for using the translation environment stored by the environment storage means at the time of translation, the user can set a translation environment that may be different for each field or sentence and There is no need to set to, you can concentrate on the original work. Further, the "translation apparatus" of Japanese Patent Laid-Open No. 63-223962 is unnecessary by respecting a user's prioritized viewpoint of importance and switching a translation mechanism such as a dictionary or a translation rule according to a user's instruction. It is possible to obtain a high-speed translation system and a high-quality translation system according to the user's request by reducing the number of necessary analyzes.

In a machine translation system, a translation environment used for a certain document is stored and the translation environment is reused. For example, Japanese Patent Laid-Open No. 4-19546.
The "language processing device" of Japanese Unexamined Patent Publication No. 0 specifies one or two or more documents out of input sentences, performs language processing on the input sentences, and stores environment data necessary for this language processing. Then, the stored environmental data is designated, and the designated document is stored in association with the environmental data. This allows you to manage environmental data, so when performing language processing in the same or similar environment, the user can easily reuse environmental data by selecting the stored environmental data and the document to process, The burden on the user is reduced.

[0004]

As described above, when the translation is performed by the conventional machine translation apparatus, it has been conventionally required to specify the translation environment before the translation is executed according to the content and purpose of the original sentence to be translated. It was possible. In this case, the translation environment includes a wide range of fields such as field designation, auxiliary verb translation setting, target sentence form (respectal and ordinary), technical term dictionary selection, user registration dictionary selection, etc. Settings such as field designation, technical term dictionary selection, user registration dictionary selection
Since the setting is related to the meaning and content of the original text, it is necessary for the user to set it according to the document. However, the translation environment (designation of style, translation mode, etc.) according to the characteristics of the document of each sentence type is often common to each sentence type, and it is a fixed setting item. If multiple translation environments corresponding to the sentence type are automatically set at once by selecting, the translation environment settings required for translation are
Other than selecting the settings related to the meaning and content of the original sentence, you only need to select the sentence type. This not only reduces the work load of setting the translation environment according to the sentence type, but also eliminates the need for judgment to make appropriate settings for each environment setting according to the main sentence type, and the user is not required to do so. Even if the user is not familiar with the translation work by the machine translation device, it becomes possible to guide the appropriate translation work.

If the user is very familiar with the translation work by the machine translation device, he or she can appropriately determine the setting items of the conventional translation environment according to the content, purpose and use of the original sentence,
You can also set your own translation environment. Therefore, depending on the purpose and use of the original sentence to be translated, the user needs a sentence type different from the translation environment set of the main sentence type prepared in advance by the system, or the existing translation environment setting is handled by the user. It may decide that it is not optimal for the document. Therefore, the user can arbitrarily define a sentence type (translation environment set) and can select the sentence type similarly to the existing sentence type. As a result, the user can always select and set a more appropriate translation environment, and can obtain a higher quality translation result.

[0006] In addition to the translation environment of the sentence type preset by the system, the user can arbitrarily register a set of translation environments, so the translation that can immediately correspond to the specific sentence type handled by the user It becomes possible for the user himself to maintain the environment. Furthermore, since the translation environment preset by the system uses not only the translation environment setting items that can be set by the user but also a dedicated dictionary that the user cannot operate, a dictionary dedicated to this main sentence type is used. If the user sets the translation environment using, the translation environment will be more complete than the new environment set by the user due to the change of the standard translation environment of the system setting that has been possible in the past. .

The present invention has been made in view of such circumstances, and by selecting a sentence type of an original sentence to be translated, a translation environment corresponding to the sentence type is automatically set, and
It is an object of the present invention to provide a machine translation device provided with a function by which a user himself / herself can define a set of translation environments of sentence types.

[0008]

In order to achieve the above object, the present invention provides (1) a machine translation apparatus having a translation means for performing translation processing in a specified translation environment, wherein a plurality of translation environments are provided in a certain sentence. As a specific translation environment corresponding to the seed,
Storage means for storing as a set, display means for searching the set of translation environments stored by the storage means and displaying the contents thereof, and changing means for changing the contents of the translation environment stored as the set, By having a translation means for performing translation processing in the translation environment stored as a set for each sentence type by designating the sentence type, further, (2) a newly set translation environment corresponding to a certain sentence type. It is characterized in that it has an addition / registration means for adding and registering a set to the storage means, and the translation processing is performed in the newly registered translation environment.

[0009]

According to the present invention, a translation environment corresponding to a sentence type is automatically set only by selecting a sentence type. Therefore, when translating a document of the same sentence type, a common sentence type is used. Only by selecting (Translation environment set), a uniform translation level is always maintained. In addition to the translation environment of the sentence type preset in the system, the sentence type (translation environment set) arbitrarily set by the user is added as a selectable sentence type when the sentence type is specified. As the user's sentence type (translation environment set) setting is added, the translation environment desired by the user will be gradually improved.

[0010]

Embodiments will be described below with reference to the drawings. FIG. 1 is a block diagram for explaining an embodiment of a machine translation apparatus according to the present invention, in which 1 is a main CPU (central processing unit), 2 is a main memory, and 3 is a CRT (Cathod).
e Ray tube; 4 is a keyboard, 5 is a translation module, and 6 is a memory. The memory 6 stores a dictionary for translation, grammar rules, tree structure conversion rules, and the like. When the source language is input, the translation module 5
It translates it and outputs the target language.
That is, the source language input from the keyboard 4 is
It is sent to the translation module 5 under the control of the main CPU 1. The translation module 5 translates the input source language into a target language using a dictionary, grammar rules, tree structure conversion rules, etc. stored in the memory 6. The result is temporarily stored in the main memory 2 and is also stored in the CRT 3
Is displayed in.

The machine translation performed by the translation module 5 generally has an analysis level as shown in FIG. When the source language is input in the upper left of the figure, the dictionary lookup of level L1, morphological analysis of level L2, syntactic analysis of level L3, semantic analysis of level L4, context analysis of level L5, intermediate language of level L6, level L7 Context generation, level L
8 meaning generation, level L9 syntax generation, and level L10 morpheme generation and analysis. Machine translation can be roughly divided into the following two, depending on this analysis level. One is
Level L6 source language and target language independent concepts (called intermediate languages) are analyzed, and from there, level L7 context generation, level L8 meaning generation, level L9 syntax generation, level L10 morpheme It is a pivot method that advances the generation and generation to generate the target language. The other is to obtain the internal structure of the source language by performing any of the above-mentioned level L2 morphological analysis, level L3 syntactic analysis, level L4 semantic analysis, and level L5 context analysis to obtain the internal structure of the source language. This is a transfer method in which the target language is generated after converting the obtained internal structure of the source language into the internal structure of the target language at the same level.

Next, each analysis will be described below.Dictionary lookup, morphological analysis Look up the dictionary in the memory 6 shown in FIG.
It is divided into morpheme strings (word strings) and the part of speech for each word
Get grammatical information and translated words such as
Analyze numbers etc.Parsing As described below, the structure of the sentence, such as the dependency between words
(Structural analysis tree) is determined.Semantic analysis It seems that the result of multiple parsing is semantically correct
Determine what is not. Contextual analysis Understand topics and remove omissions and ambiguities.

FIG. 3 is a block diagram of the translation module in FIG. 1, in which 11 is a dictionary lookup morpheme analysis unit, 12 is a syntax analysis unit, 13 is a conversion unit, and 14 is a translation generation unit. The translation module 5 analyzes at least up to the level of the level L3 syntax analysis. First, the read original sentence is divided into each morpheme sequence (word sequence) by the dictionary lookup / morpheme analysis unit 11 using the dictionary of the memory 6, and grammatical information such as a part of speech for each word and a translated word are obtained. , Tense / person / number etc. are analyzed. The syntactic analysis unit 12 determines a structural analysis tree indicating the dependency relation between each word according to the dictionary of the memory 6 and the grammar rules. Converter 1
In 3, the structure of the inputted parse tree for the original sentence is converted into the structure of the parse tree for the translated sentence using the tree structure conversion rule of the memory 6. Then, the translated sentence generation unit 14
Then, an appropriate particle or auxiliary verb is added as a translated sentence in the target language, and the obtained translated sentence is output from the translation module 5.

FIG. 4 is a concrete block diagram of the translation module. In the figure, 15 is a translation CPU (central processing unit), 1
Reference numeral 6 is a translation program, 17a to 17e are buffers, and the other parts having the same functions as those in FIG. 5 to 9 are buffers A shown in FIG. 4 when translating the English sentence "This is a pen." Into Japanese.
3 illustrates the contents of E to E. The operation of the English-Japanese translation will be described below with reference to FIGS. First, the read original text is stored in the buffer A in FIG. 4, as shown in FIG. Translation CP based on translation program 16
Under the control of U15, the dictionary lookup / morphological analysis unit 1 of FIG.
1, the information such as the translated word of each word is obtained using the dictionary of the memory 6 according to the original sentence stored in the buffer A, and is stored in the buffer B of FIG. For example, part-of-speech information, which is a part of each piece of information, is stored as shown in FIG.

Here, although this is a multi-part-of-speech word, it is uniquely determined by the syntax analysis unit 12 in FIG. In the syntactic analysis unit 12, according to the dictionary and grammar rules of the memory 6,
A structural analysis tree showing the dependency relationship between each word is determined as shown in FIG. 7 and stored in the buffer C of FIG. This structural analysis tree is determined as follows. That is, from the grammatical rules of the memory 6, the rules of sentence → main part, predicate main part → noun phrase predicate → verb, noun phrase noun phrase → pronoun noun phrase → article, noun rule can be obtained. This rule represents, for example, that the first rule is "a sentence is composed of a main part and a predicate". Hereinafter, the structural analysis tree is determined according to this rule.

In the conversion unit 13 of FIG. 3, the structure of the syntax analysis tree (see FIG. 7) for the input English sentence is changed by using the tree structure conversion rule of the memory 6 as in the case of the above-described syntax analysis unit 12.
As shown in FIG. 8, the structure is converted into a Japanese sentence.
Then, the obtained result is stored in the buffer D of FIG. The translated sentence generation unit 14 in FIG. 3 attaches an appropriate particle "ha" or auxiliary verb to the obtained Japanese character "kore pen ni wa." Store in buffer E. This Japanese sentence "This is a pen."
It is output from the translation module 5, stored in the main memory 2 and displayed on the CRT 3.

FIG. 10 is a flow chart for explaining the operation of the machine translation device according to the present invention. The operation of automatically setting the translation environment by selecting the sentence type and the process of setting an arbitrary sentence type (translation environment set) will be described by taking an English-Japanese machine translation device as an example. Note that the setting and selectable translation environment illustrated in FIGS. 12 and 13 will be described before the description in order according to each step. FIG. 12 is an example of a sentence type translation environment display screen displayed after selecting a sentence type on the display screen shown in FIG. 11, and displays the contents of the translation environment automatically set by the system.
FIG. 13 is an example of a sentence type setting screen displayed when the user arbitrarily sets a sentence type.

The "auxiliary verb translation setting" allows the user to select either the system setting translation preset by the system or the user changing the translation for each auxiliary verb. The “translated form” is used to set whether the translated word is “is a key” or “is a key”. The "translation mode" (A. uppercase mode) is a mode selected when translating an uppercase English sentence such as a telex form. "B. Automatic division" of "Translation mode" is a mode to set when the length of one sentence of the English sentence to translate is long,
Automatically split and translate long sentences that may fail to parse the syntax. "Translation mode" (C.
Is a mode to set when there are many parenthesized parts such as papers and reference pages, and it outputs the English sentences in parentheses that do not need to be translated as they are without being translated. . The "basic sentence type setting" is for setting whether or not to use the dedicated dictionary corresponding to the main sentence type set in the system in advance. The dedicated dictionary for the sentence type is stored in the memory 6 of FIG. 1, and when the system setting sentence type is selected or the basic sentence type is set, it is preferentially adopted during the translation process. The setting items of the translation environment are merely examples, and any items that can be specified by the user can be incorporated in the system as the translation environment setting items.

The steps will be described below in order. step1 : First, when a sentence type setting is designated, selectable sentence types are displayed. FIG. 11 is an example of a sentence type selection screen displayed when a sentence type is selected before translation is executed. step2 : Next, the sentence type number to be set from the selection screen. Select and enter. At this time, if you want to arbitrarily define and set the sentence type (translation environment set), select No. Select 00. In FIG. 11, the sentence type No. 1 to 5 are set, and the user has already set No. 10 and No. 1
This indicates that the first sentence type (translation environment set) has been set. step3 : The sentence type No. input in step2 Is determined to be 0 or not. step4 : If NO, the selected sentence type No. in the translation environment table shown in FIG. Search for. 14
Is an example showing the contents of a translation environment table in which a set of translation environments for sentence types is stored.

[0020] step5: Then, Bunshu was selected No. Display the contents of the translation environment table. FIG. 12 is an example of a screen showing the contents of the translation environment displayed at this time. The ⊚ mark shown in this example indicates the translation environment preset in the system. step6 : Input to specify whether or not to change the translation environment shown on the display screen. step7 : Judge whether the translation environment needs to be changed. If NO, in step 13, the sentence type N already input in step 2
o. The translation environment set is executed, and the processing for automatically setting the translation environment by selecting the sentence type is completed. That is, if the setting change unnecessary is designated on the display screen shown in FIG.
It means that the settings as shown in are executed. step
If YES in 7, go to step 8. step8 : Switch to the translation environment change mode, and the translation environment settings shown in FIG. 12 can be changed. step9 ： Change the translation environment. Although the display screen at this time is not shown, when the modification is completed, the process proceeds to step 13, the modified translation environment set is executed, and the translation environment setting process by sentence type selection is completed.

Step 10 : On the other hand, when a sentence type (translation environment set) is arbitrarily defined and set, that is, the above step 3
If YES, the mode is switched to the translation environment setting mode. step11 : Translation environment setting, sentence type No. And enter the sentence type name. The display screen at this time is shown in FIG. In this example, the user selects the type name “B company contract”, type number No.
11 is an example of a display screen showing that the item marked with ⊚ is designated to be set. step12 : When the setting is completed, the set translation environment set is added to the translation environment table and registered. step13 : Next, the set translation environment of the set sentence type is executed, and the process of setting the translation environment of the arbitrarily set sentence type ends here.

The operation of the automatic setting and changing of the translation environment by the sentence type selection and the process of setting an arbitrary sentence type (translation environment set) has been described above, but the user newly sets it.
The added translation environment can be selected at any time at the next sentence type selection, and can be deleted.

[0023]

As is apparent from the above description, the present invention has the following effects. (1) The present invention enables the following items. In addition,
"Sentences" here means manuals, patents, letters,
It means the type of purpose of the original text for translating telex, contracts, newspapers, etc. By simply selecting the type of source text to be translated, the translation environment is automatically set according to the type of source text, eliminating the need to individually perform several types of translation environment setting items, which reduces the burden of setting work. It will be reduced. Even if the user cannot determine what to do with the setting contents of the translation environment according to the purpose of the original sentence, selecting the sentence type of the original sentence sets the translation environment according to the sentence type, so that the machine translation device operation or Even a person who is not familiar with the translation work by the machine translation device can more easily perform the appropriate translation work. Sentences defined by the user (translation environment set)
Can be selected at the time of translation as well as the selection of the translation environment set for the main sentence types that the system has set in advance.
By setting, the sentence type can be selected at any time when needed, so that a translation environment closer to the user's usage environment can be easily used. Since it is possible to set a translation environment set that uses a dictionary dedicated to the main sentence types that is set in advance in the system, the translation environment set for the main sentence types in the system settings,
Also, when combined with the user's own translation environment set, it becomes possible for the user to prepare a translation environment that can handle various sentence types. (2) The translation environment corresponding to the main sentence type is automatically set as a set, which reduces the burden of setting work, and
Since there is no need for judgment for proper setting, there is an effect that the difficulty of work for setting the translation environment is significantly reduced. On the other hand, since a set of translation environments can be arbitrarily set, the user can make more appropriate settings according to the usage environment. That is, it is possible to realize a machine translation device that can flexibly cope with whether the user is proficient in translation work by the machine translation device or not. (3) Furthermore, when the sentence type translation environment is combined with other translation environments such as field designation, technical term dictionary selection, user registration dictionary selection, a considerable number of translation environment combinations are possible. A detailed translation environment can be maintained, and translation processing that accurately captures the characteristics of the document to be translated can be easily realized.

[Brief description of drawings]

FIG. 1 is a configuration diagram for explaining an embodiment of a machine translation device according to the present invention.

FIG. 2 is an explanatory diagram of a translation level in machine translation of the present invention.

FIG. 3 is a configuration diagram of a translation module in FIG.

FIG. 4 is a specific block diagram of a translation module in FIG.

FIG. 5 is a diagram showing an example of contents of an original text buffer according to the present invention.

FIG. 6 is a diagram showing an example of the contents of a part of a buffer of a dictionary lookup result according to the present invention.

FIG. 7 is a diagram showing an example of contents of a buffer of a syntax analysis result according to the present invention.

FIG. 8 is a diagram showing an example of contents of a buffer as a result of tree conversion according to the present invention.

FIG. 9 is a diagram showing an example of contents of an output sentence buffer according to the present invention.

FIG. 10 is a flowchart for explaining the operation of the machine translation device according to the present invention.

FIG. 11 is a diagram showing an example of a display selection screen when selecting a sentence type in the present invention.

FIG. 12 is a diagram showing an example of a display screen of contents of a sentence type translation environment according to the present invention.

FIG. 13 is a diagram showing an example of a display setting screen when a new sentence type is set in the present invention.

FIG. 14 is a diagram showing an example of contents of a translation environment table in a storage system of a translation environment according to the present invention.

[Explanation of symbols]

1 ... Main CPU (central processing unit), 2 ... Main memory, 3 ... CRT (Cathode Ray tube), 4 ...
Keyboard, 5 ... Translation module, 6 ... Memory, 11 ...
Dictionary lookup morpheme analysis unit, 12 ... syntax analysis unit, 13 ... conversion unit, 14 ... translation generation unit, 15 ... translation CPU (central processing unit), 16 ... translation program, 17a to 17e ... buffer.

Claims (2)

[Claims] 1. A machine translation apparatus having a translation means for performing translation processing in a specified translation environment, and a storage means for storing a plurality of translation environments as a set as a specific translation environment corresponding to a certain sentence type, A display means for searching the set of translation environments stored by the storage means and displaying the contents thereof, a changing means for changing the contents of the translation environment stored in the set, and a sentence type for each sentence type And a translation unit that performs translation processing in a translation environment stored as a set. 2. A translation processing is performed in the newly registered translation environment, with addition / registration means for adding / registering a newly set translation environment set corresponding to a certain sentence type to the storage means. The machine translation device according to claim 1, wherein: