JPH05216924A - English-japanese mechanical translation device - Google Patents

Abstract

PURPOSE:To improve the translating efficiency by designating an optional paragraph and moving the paragraph to an optional position, thereby easily obtaining the translated sentences without using a text editor. CONSTITUTION:An original sentence inputted by an input means 1 is stored in an original/translated sentence storage means 3 and only the original sentence is displayed on a display means 4. Then the original sentence is translated by a translation executing means 2 and both original and translated sentences are shown on the means 4. Then a user decides whether each paragraph of the translated sentence should be shifted or not. If so, the user designates the relevant paragraph with a paragraph designating means 7 and then designates a paragraph right after the desired position where the first paragraph is put is also designated. Then the button of a mouse serving as the means 7 is clicked to shift the desired paragraph after this paragraph and its position are designated. Furthermore the paragraph may also be designated through a keyboard.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an English-Japanese machine translation device for converting an English sentence into a Japanese sentence.

[0002]

2. Description of the Related Art In recent years, machine translation devices have been widely used as devices for translating one language into another. Current machine translation devices tend to be translation support devices rather than automatic translation devices. That is, the user performs pre-processing and post-processing on the translation process to create a more complete translation. The present invention relates to post-processing of an English-Japanese machine translation device.

In English sentences to be translated, there are often sentences including parallel phrases. At that time, when a modifier such as a prepositional phrase is present after the parallel phrase, ambiguity appears in the range modified by the modifier. For example, in the sentence I saw rings and gloves of leather. "Rings" and "gloves" are parallel phrases.
In this example, the modifier "of leather" is "rings
It's ambiguous whether it's "and gloves" or "gloves" only. That is, as a result of translating this sentence, there are two possibilities as in the following example.

I saw leather rings and gloves.・ I saw a ring and leather gloves. In the case of this example, humans would prefer the latter interpretation because of the unnaturalness of the "leather ring". However, in order for the machine translation device to select one of these two translations, information such as context information and common sense / knowledge is required, which is difficult to realize at the present time. Therefore, according to the conventional technique, one translation is selected by the user determining the destination of the modifier and instructing it to the machine translation device. Here, there are the following three types of methods for the user to indicate the destination of the modifier.

1. A method in which the user specifies the range to which the modifier is applied at the pre-edit stage before the translation process. For example, to set the contact to "gloves", change the original text as follows using an editor.

However, with this method, the user needs to judge by looking at the English sentence, and considering that most of the users of the English-Japanese translation device are Japanese. , Not easy to use.

2. How the system asks the user when ambiguity arises during the translation process. This method is also not easy for the user to use because it requires the user to make a judgment by looking at the English sentence, but it needs to make a judgment by looking at the tree structure that is an internal expression.

3. A method of changing the destination of modifiers using a text editor for the translated Japanese in the post-editing stage. It is easy for users to judge the destination of modifiers by looking at Japanese sentences, but rewriting using a text editor is complicated.

[0009]

As described above, among the three types of methods for determining the destination of a modifier phrase, the method performed at the post-editing stage is the method that allows you to see Japanese sentences. Easy for users to use. However, the work of rewriting the Japanese translation using a text editor is complicated.

Therefore, an object of the present invention is to provide an English-Japanese machine translation device which can easily obtain a translated sentence without using a text editor and can improve the efficiency of translation work.

[0011]

According to the present invention, an input means for inputting an original sentence, a translation executing means for translating an input original sentence to create a translated sentence, and an original sentence for storing the original sentence and the translated sentence in association with each other・ Transliteration storage means, display means for displaying the original text and translated text in association with each other, phrase specifying means for designating any phrase in this translated text, and the designated phrase at any position in this translated text The phrase moving means to be moved is configured.

[0012]

With the above arrangement, the phrase can be moved at the stage of post-editing without using a text editor. For example, in order to make only “gloves” in the modifier “leather” in the sentence “I saw a leather ring and gloves.”, Specify the phrase “leather”. By moving the glove just before, you can change it to the sentence "I saw a ring and leather gloves." In other words, the user can change to the desired sentence in two steps: specifying the phrase "leather" and specifying the insertion position immediately before "glove", so compared to the case using a text editor. , Work easier.

As described above, in carrying out the English-Japanese machine translation, by having the function of moving the clause at the post-editing stage, it is possible to easily change the destination of the modifier to the parallel phrase.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A description will be given below with reference to the drawings. Figure 1
FIG. 3 is a functional block diagram of an English-Japanese machine translation device according to an embodiment of the present invention. Reference numeral 1 is an input means for inputting an original sentence and instructing a change of a translated text for post-editing, 2 is a translation executing means for translating an original sentence input from the input means 1, and 3 is
An original sentence / translated sentence storage unit for storing the original sentence input by the input unit 1 and a translated sentence translated by the translation execution unit 2 in association with each other, 4 is a display unit for displaying the contents of the original sentence / translated sentence storage unit 3, and 5 is a display unit. A post-edit information storage means for storing information for post-edit obtained in the process of the translation execution means 2; 6 denotes a phrase specified by the user based on the information stored in the post-edit information means 5; It is a phrase moving means for moving the phrase to a newly designated position. Reference numeral 7 is a phrase specifying means for inputting a user's phrase specifying instruction to the phrase moving means 6.

FIG. 2 is a circuit block diagram of the English-Japanese machine translation apparatus of the present embodiment. 8 is a keyboard which is an example of the input means 1, 9 is a central processing unit (hereinafter CP) which executes each processing.
U is abbreviated as U), 10 is a cathode ray tube display (hereinafter abbreviated as CRT) which is an example of the display unit 4, 11 is a random access memory (hereinafter abbreviated as RAM), 12 is a read only memory (hereinafter abbreviated as ROM). The RAM 11 or the ROM 12 is provided with the above-mentioned translation execution means 2, the original / translated text storage means 3, the post-edit information storage means 5, the phrase movement means 6 and the like. A mouse 13 is an example of the phrase specifying unit 7. Of course, for example, the original sentence / translated sentence storage means 3 may be provided in a hard disk (not shown).

FIG. 3 is a flow chart showing the flow of processing in the above embodiment. The flow of processing in this embodiment will be described below with reference to this flowchart.

First, in step A, an original sentence is input by the input means 1. The input original sentence is stored in the original sentence / translated sentence storage unit 3. In this state, only the original sentence is displayed on the display means 4 as shown in FIG.

Next, in step B, the input original sentence is translated by the translation executing means 2. Then, the original sentence and the corresponding translated sentence are displayed on the display means 4 as shown in FIG.

Next, in step C, the user determines the necessity of phrase movement for each phrase of the obtained translated sentence, and specifies the phrase if necessary. Here, in the example of FIG. 5, assuming that the modified expression "leather's" must be related only to the phrase "gloves", the phrase specifying means 7 specifies the phrase "leather's" in the translated sentence. .. Figure 6
The state is shown. In addition, it is assumed that the clauses included in the following □ are specified. Then, the phrase immediately after the position where the phrase is to be inserted is specified. FIG. 7 shows the state. In this way, when the desired phrase and the insertion position can be designated, the phrase is moved by clicking the mouse button of the mouse 13 which is the phrase designating means 7. FIG. 8 shows a translated sentence after the phrase movement processing. The phrase may be specified from the keyboard 8.

9 to 12 show states of internal data of the post-edit information storage means in steps B and C.
This internal data is what the translated text is represented by a bidirectional list as shown in FIG. In the bidirectional list, the clause of the translated sentence corresponds to each cell. In addition, each cell points the left and right cells with a pointer to indicate the order of the cells. That is, in FIGS. 9 to 12, C1
.About.C5 correspond to cells, and P1 to P4 correspond to pointers. Therefore, by moving the pointers P1 to P4, it is possible to move the cells C1 to C5, that is, move the phrase. FIG. 10 shows the state of the bidirectional list when the user specifies the phrase to be moved. At this time, the pointers P1 and P2 before and after the "leather" of the moving cell C2 are separated. FIG. 11 shows the state of the bidirectional list when the user specifies the insertion position. In this state, the pointer P3 between the cells C3 and C4 corresponding to the position designated by the user
P3 is disconnected. Cell C2 already separated
"Leather" is the pointer P between these cells C3, C4.
The insertion process is executed by connecting the lines using 3, P1, P2, and P3 (FIG. 12).

[0021]

The present invention has an input means for inputting an original sentence,
A translation execution unit that translates the input original sentence to create a translated sentence, an original sentence / transliteration storage unit that stores the original sentence and the translated sentence in association with each other, and a display unit that displays the original sentence and the translated sentence in association with each other. , The phrase specifying means for specifying an arbitrary phrase in this translation and the phrase moving means for moving the specified phrase to an arbitrary position in this translation are modified. Now you can easily modify the destination of the phrase without using a text editor,
The work efficiency of the translation work using the machine translation device can be improved.

[Brief description of drawings]

FIG. 1 is a functional block diagram of a machine translation device according to an embodiment of the present invention.

FIG. 2 is a circuit block diagram of the machine translation device.

FIG. 3 is a flowchart of the machine translation device.

FIG. 4 is a diagram showing a display unit displaying the original sentence.

FIG. 5 is a view showing a display unit displaying the translation result.

FIG. 6 is a view showing a display unit displaying a result of selecting a phrase to be moved.

FIG. 7 is a diagram showing a display unit displaying a result of selecting a position to insert the same clause.

FIG. 8 is a diagram showing a display unit displaying a result of moving the same clause.

FIG. 9 is a diagram showing a bidirectional list which is an internal representation of a translated sentence in the post-edit information storage means 5.

FIG. 10 is a diagram showing a bidirectional list in which cells corresponding to the clauses to be moved are separated.

FIG. 11 is a diagram showing a bidirectional list in which the cell pointer at the position where the same clause is inserted is separated.

FIG. 12 is a diagram showing a bidirectional list after the movement of the same clause is completed.

[Explanation of symbols]

 1 Input Means 2 Translation Execution Means 3 Original / Translated Text Storage Means 4 Expression Means 5 Post-Edit Information Storage Means 6 Phrase Movement Means 7 Clause Designation Means

Claims (1)

[Claims] 1. An input unit for inputting an original sentence, a translation executing unit for translating an input original sentence to create a translated sentence, and an original sentence / transliteration storage unit for storing the original sentence and the translated sentence in association with each other.
Display means for displaying the original sentence and the translated sentence in association with each other, a phrase specifying means for designating an arbitrary phrase in the translated sentence, and a phrase moving means for moving the designated phrase to an arbitrary position in the translated sentence. An English-Japanese machine translation device characterized by having.