JPH0352102B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a text display proofreading method in a system for automatically translating text in a first language (eg, English) into a second language (eg, Japanese).

In a computer-based automatic translation system, it is essential to check and proofread the computer-generated translation results. Conventionally, general-purpose terminal devices or word processors have been used, but since they do not have functions that fully take into account the characteristics of translation work, there has been a problem that the efficiency of proofreading work is extremely low.

An object of the present invention is to provide a display proofreading method that has functions suitable for a translation system. Specifically, these are (1) a function for simultaneously displaying text in both languages on the same screen, and (2) a function for interlocking scrolling of text in both languages.

To achieve the above objectives, the method according to the invention deals with text separated into phrase elements. IDs are assigned to phrase elements, and text is stored and processed as a series of phrase element data consisting of IDs and character strings. By giving the same ID to corresponding phrase elements in both texts, the correspondence between both texts is maintained.

An embodiment of the present invention will be described below. For convenience of explanation, the first language is English and the second language is Japanese. As shown in FIG. 1, a terminal device 1 according to the present invention includes a processing device 2, a storage device 3, a character generator 4, a CRT 5, and a keyboard 6, and is connected to a central computer 9 via a modem 7 and a communication line 8. Connected. The automatic translation process is executed in the central computer 9, and the English text and the translated Japanese text are sent to the terminal device 1. English text and Japanese text are displayed simultaneously on the terminal device 1, and the user modifies the Japanese text. The modified text is sent to the central computer 9 and stored in a file.

In the automatic translation process in the central computer 9, the English text is parsed and divided into elements (referred to as phrase elements) that have certain syntactic and semantic roles.
A Japanese text is generated by assigning a translation word to each phrase element and arranging the phrase elements in an appropriate order for Japanese. Therefore, an identification symbol (hereinafter abbreviated as ID) is assigned to the phrase element, and a second
Data as illustrated in the figure is transmitted to the terminal device 1. The next byte of the ESC code in the data is a phrase element
It is an ID. Note that the ESC code and statement number are added to the beginning of the statement. Since the sentence number is sandwiched between the ESC codes, it can be distinguished from the phrase element ID. By adding a sentence number, the phrase element ID only needs to be unique within one sentence. Therefore, the phrase element ID can be represented by several bits.

The storage device 3 consists of an English text area 31, a Japanese text area 32, and an area for storing various parameters 33. English text area is 1
Word 9, the Japanese text area, has 17 bits per word and is individually addressed as 1, 2, 3, . . . . 3a shows words in the English text area. The first bit _D1 in the word is 0
The content expressed by the word differs depending on whether the word is a word or a word.
When the first bit _D1 is 0, the second to ninth bits are codes representing characters in the text. When the first bit is 1, the third to ninth bits represent a phrase element ID or sentence number. Whether it is a phrase element ID or a sentence number can be easily determined from the following. That is, when there are two consecutive words in which the first bit _D1 is 1, only the first word is the sentence number. The second bit _D2 of the word representing the phrase element ID is used as a flag indicating whether or not the phrase element has been modified. The third word in the Japanese text area
Shown in Figure b. The content expressed by the word differs depending on the first bit _D1 , as in the case of the English text area. FIG. 4 shows the stored contents of the text area for the data in FIG. 2.

The various parameters stored in the storage device 3 are
Shown in Figure c. English text display start address P1,
The Japanese text display start address P2 indicates the address in the text area of the top portion of square text displayed on the CRT 5. cursor position flag
P3 is 0 when the cursor is on English text,
Set to 1 when on Japanese text. Cursor address P4 indicates the address of the cursor on the text.

The screen of the CRT 5 is divided into an English text display section 51 and a Japanese text display section 52, as shown in FIG. Each, m ₁ (character/line) x n ₁
(line), m ₂ (characters/line) x n ₂ (line) can be displayed. Through the process described below, when one text is scrolled, the other text will also be scrolled.
The corresponding part is always displayed.

As shown in FIG. 6, the keyboard 6 includes character keys 601 as well as various function keys 60.
2 to 617.

The functions of the processing device 2 are: (1) writing the text sent from the central computer 9 into the storage device 3; (2) displaying the text in the storage device 3 on the CRT 5;
(3) Processing that changes the display range as the cursor moves.

These will be explained below.

(1) Processing for writing the text sent from the central computer 9 into the storage device 3 FIG. 7 shows a flowchart of the process for writing the English text sent from the central computer 9 into the English text area of the storage device 3. The same is true for Japanese text. In the figure, K is an index indicating the address of the English text area, and A(K) is the address K of the English text area.
Indicates the word. First, after setting K to an initial value of 1 (101), the transmitted data is set to 1.
Read byte by byte (102). End of data (103)
If not, determine whether it is an ESC code (104), and if it is an ESC code, the next 1 byte is a sentence number or phrase element ID, so read this (105) and set the first bit to 1 and the second bit to is 0, 3rd to 9th bits are sentence numbers or phrase elements
Write the word that is the ID (106). If it is not an ESC code, it is a character code, so write a word in which the first bit is 0 and the second to ninth bits are this character code (107). In either case, after writing to A(K), add 1 to K (108) and proceed to reading the next byte (102).

(2) Processing for displaying text in storage device 3 on CRT 5 Figure 8 shows a flowchart of text display processing. Only the display portion of the English text will be shown in detail, and the Japanese text will be omitted since it is the same as the English text. First, J and
The text display start address P1 and display end address P1+m ₁ ×n ₁ are respectively set in JMAX (202 and 203). It is determined whether the word A(J) at address J is a character code (204), and if it is a character code, this code is sent to the character sounder 4 and displayed on the CRT 5 (206). On the other hand, A(J) is the sentence number or phrase element ID
If so, it is displayed as blank on the CRT screen (215). After displaying A(J), add 1 to J (216), and if J is within the display range (217)
Proceed to process the next word. If J exceeds the display range, the process moves to Japanese text display processing (218).

(3) Processing to change the display range as the cursor moves Some cursor movement keys are shown in Figure 6. Now, a case where the cursor position flag P3 is 0, that is, a case where the cursor is on English text will be explained. The same applies when the cursor position flag is 1. key 605
When is pressed, cursor address P4 is decremented by 1. However, if P40 is (multiple of m ₁ ) + 1, no change is made. When key 606 is pressed, 1 is added to P4. However, if P4 is a multiple of _m1 , no change is made. When key 604 is pressed, P4 is decreased by _m1 . When the key 607 is pressed, m ₁ is added to P4. Also, when the key 608 is pressed, the English text area is scanned in the reverse direction from the current address of P4,
Search for a word that is a phrase element ID. (Address of the phrase element ID that appears second) +1 is set in P4. When the key 609 is pressed, the English text area is scanned forward from the current P4 address to search for the word that is the phrase element ID. (Address of first phrase element ID) +
Set 1 to P4. Furthermore, when the key 610 is pressed, the cursor position flag P3
At the same time, it scans the Japanese text, searches for a phrase element with the same ID as the phrase element pointed to by the cursor address set in P4, and sets the address of its first character to cursor address P4. do.

When the cursor is moved, it becomes necessary to scroll the screen, so the parameters P1 and P2 stored in the storage device 3 are also updated. The flow of this process is shown in FIG. The illustration is for the case where the cursor is over English text. If cursor address P4 is smaller than English text display start address P1 (301), P1
is subtracted by one line (302), and if P4 is greater than the English text display end address P1+m ₁ ×n ₁ (303), P1 is increased by one line (304).

Furthermore, the address Z of the phrase element in the Japanese text with the same ID as the phrase element pointed to by the cursor advice P4 is determined (306), and if Z is smaller than the Japanese text display start address P2 (307), P2
is reduced by one line (308), and if Z is greater than the English text display end address P2 + m ₂ × n ₂ (309),
Increment P2 by one line (310).

According to the present invention, there are the following effects.

(1) Comparison display of text in the first language and text translated into the second language, interlocking scrolling function, dynamic display function of correspondence between texts (phrase element units) by moving the cursor, etc., make translation possible. The results can be calibrated efficiently.

(2) To store the correspondence between the two texts necessary to realize the above function, it is only necessary to add one bit to each word of the refresh memory.

[Brief explanation of drawings]

Fig. 1 is a diagram showing the overall configuration of the machine translation system according to the present invention, Fig. 2 is a diagram showing an example of text sent from the central computer to the terminal device, and Figs. 3 a, b,
c is a diagram showing the configuration of words and various parameters of the storage device, FIG. 4 is a diagram showing an example of the storage contents of the text storage device, FIG. 5 is a diagram showing an example of text display, FIG. 6 is a diagram of the keyboard board, Figure 7 is a flowchart of the process of writing the text sent from the central computer to the storage device, and Figure 8 is the process of writing the text in the storage device.
FIG. 9 is a flowchart of the process of displaying on the CRT. FIG. 9 is a flowchart of the process of updating the contents of the storage device in response to input from the keyboard.

Claims (1)

[Claims] 1 In a machine translation system that translates a first text written in a first natural language into a second text written in a second natural language, the first text corresponds to the first text. Said second
means for dividing and displaying the text on the same screen of a display device in correspondence with the first text on the display screen;
and means for displaying a portion of the other text that corresponds to the changed display portion when the display range of the second one text is changed.