JPH03141469A - Translation processor - Google Patents

Abstract

PURPOSE:To simplify the operating form of a translation processor by instructing a translator main body to start the translation processing with the sentences inputted up to a time point when the input end is instructed defined as a single translation processing unit when the input end is instructed for each sentence forming an original sentence. CONSTITUTION:A translation processing control means 4 instructs a translator main body 5 to start the translation processing with the original sentence information inputted to the input time point of the control information defined as a single translation processing unit when this information is inputted via an input means 1 to instruct the input end of the original sentence information. In other words, when a sentence punctuation is instructed at input of an original sentence, the translation processing is automatically started with the original sentences inputted up to a time point when the sentence punctuation is instructed defined as a single translation processing unit. Thus the operation of a translation processor is simplified.

Description

DETAILED DESCRIPTION OF THE INVENTION [Structure of the Invention] (Field of Industrial Application) The present invention relates to a translation processing device that can appropriately instruct translation processing for an input original text.

(Prior Art) Recently, various automatic translation processes have been proposed, such as using computer systems to translate English text into Japanese text, or Japanese text into English text. This translation process basically divides the input original text into units of translation processing such as words (phrases), and combines the translated words (phrases) found for each word (phrase) over the entire sentence. The method consists of a process of obtaining a translated text for the input original text.

During this translation process, the input source text is parsed,
According to the analysis results, processing such as eliminating inappropriate translation candidates is performed as appropriate.

By the way, the format for inputting the original text into the translation processing device is as follows:
There are cases in which an original text already created on a computer's storage device is cut out into sentences and subjected to translation processing, and cases in which character strings forming the original text are sequentially input via an input device such as a keyboard.

Conventionally, when inputting original text as a string of characters sequentially from an input device such as a keyboard, generally the "r" character on the keyboard was used.
This is done by inputting character string information (original text) by operating a key, and instructing the break between input sentences by operating a "line feed" key or the like. Then, execution of translation processing for the sentences delimited by the operation of the "line feed" key is instructed by operating a "translation" key provided separately from the "line feed" key.

In other words, for the translation processing device, inputting the original text and translating the inputted original text are completely different and independent processes, so instructions for each process are given using separate keys. ing.

However, there are many people who use such conventional devices (
Users are so absorbed in inputting the original text that they often forget to operate the "translation" key. In this case, an input sentence that should normally be processed as two or three sentences is treated as one sentence and subjected to translation processing, and the translation processing may fail due to reasons such as the syntax analysis result becoming complicated. There are many risks.

(Problem to be Solved by the Invention) In the past, since the human processing of the original text and the translation processing of the original text were completely independent, instructions for the processing were performed using separate keys. It has become. For this reason, not only is the instruction operation complicated and troublesome, but the user often forgets to operate the keys for the translation instruction. As a result, multiple sentences that should normally be translated individually are submitted to the translation process as a single sentence, and the syntax analysis becomes complicated, leading to frequent translation failures. Ta.

The present invention has been developed in consideration of these circumstances, and its purpose is to be able to reliably instruct translation processing when, for example, the original text is input sequentially as a character string, and to An object of the present invention is to provide a translation processing device that can accurately indicate units.

Structure of the Invention F] (Means for Solving the Problems) In view of the fact that the instructions for separating sentences in the original text input operation correspond to the translation instructions, the present invention provides When the end of input is instructed, for example, by operating the "line feed" key, the system instructs the translation device itself to start translation processing by treating the sentences input up to the point in time when the end of input is instructed as one translation processing unit. It is characterized by the following.

Furthermore, when control information for delimiters for the input original text is input, the original text is divided into one translation processing unit at each delimiter position and the translation apparatus is instructed to restart the translation process. It is characterized by:

(Function) According to the present invention, when a sentence break instruction is given when inputting an original text, translation processing is automatically started with the original text up to the point in time when the break instruction is given as one translation processing unit. , it is no longer necessary to separately instruct translation processing each time as in the past, and the operation can be greatly simplified.

Moreover, he forgot the instruction to start translating and translated several consecutive sentences into one.
Problems such as translation processing as a sentence are also eliminated.

In addition, by giving a break position instruction to the input original text, the original text is divided into one translation processing unit at that break position and the translation process is restarted, making it easy to perform reliable translation processing. becomes possible.

(Example) Hereinafter, a translation processing apparatus according to an example of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic configuration diagram of the embodiment device, and FIG. 2 is a diagram showing the basic processing sequence of the input section l in the embodiment device. The input unit 1 includes, for example, a keyboard,
In addition to sequentially inputting character strings by selectively operating the "Multiple Types of R Characters" key, various types of
Configured with function keys.

The key information sequentially input through the input section 1 is determined in the input section 1 as shown in FIG. That is, when key information is input (step a), it is first determined whether the human input information is an instruction for "r line break (translation)" (step b). If the instruction is not a "line break (translation)" instruction, then it is determined whether the instruction is an "end" instruction (step C). After receiving such a determination, the input character information is provided to the original text storage unit 2 and stored in sequence, and is used to form an input original text consisting of the input character string (step d). Further, as described above, the manually inputted character information is displayed on the display device 3 by sequentially forming a character string (step e).

Incidentally, when the instruction to "end" the human input of the original text is given in step C, the input of the original text and the translation process for the input original text are thereby controlled to end. In addition, in step b, if the input of the "r new line" key is detected, a one-sentence break is instructed for the character string input so far, and the translation process is performed using that one sentence as a unit of translation processing. is instructed to start. Instructions for these processes are given by the human resources department 1.
is given to the translation control unit 4 from.

For example, when an original text consisting of English is input and translated into Japanese, the input character string consisting of alphabets is input sequentially and stored in the original text storage section 2, and also displayed on the display device 3, for example. The screen is displayed as shown in FIG. 3(a).

In addition, Figure 3 is Yesterday he saw a girl
with a bagon the right ba
This example shows an example of the English sentence nk of the river, J written manually. In this case, the cursor IO is displayed at the next input character position.

After inputting the original text in this way, from the input section l,
When instructing a line break J and then inputting an English sentence to be translated, the translation control unit 4 that has received the above-mentioned “line break” instruction inputs the characters input up to the point of the above-mentioned instruction and stored in the original text storage unit 2. The column is treated as one translation processing unit (one sentence), and the translation apparatus main body 5 is instructed to start translation processing for the translation processing unit (step f). At the same time, the translation control unit 4 controls the display control unit 6 to format the input character string stored in the original text storage unit 2 and displayed on the display device 3 so as to accurately represent the original text. Have them do it.

This formatting process of the original text detects the position of a space in a human-written character string consisting of, for example, an alphabet, and uses this space to cut out each word that makes up the original text, and also prevents each word from being displayed as a line break inadvertently. This consists of processing such as controlling the displayed line break position so that there are no line breaks. Through such original text formatting processing, the input original text is formatted as shown in FIG. will be moved. By controlling the cursor in this manner, the input original text is displayed with blank lines separated by each sentence forming a predetermined unit of translation processing.

The translation device main body 5, which has been instructed to start the translation process, cuts out the character string stored in the original text storage unit 2 up to the point where the line break is instructed as one sentence forming one unit of the translation process, Translation processing is started for one sentence (original text) in which this delimiter is specified.

As the algorithm for this translation process, it is possible to appropriately employ various algorithms that have been proposed in the past.

- For example, a character string constituting an input sentence is cut out into words (phrases), and the translation dictionary 7 is searched for the translation word (phrase) for that word (phrase).

Then, from among the generally multiple translated words (phrases), inappropriate translated words (phrases) are eliminated according to the syntactic analysis results, etc.
This is done by finding a series of translated words (phrases) consisting of combinations of syntactically correct translated words (phrases) as translation candidates.

When a translation candidate for the input original text is thus obtained in the translation device main body 5 (step h), the translation is managed in correspondence with the obtained input original text and stored in the translation storage unit 8. . The translated text is then given to the display bag W3 under the control of the display control unit 6, and displayed in association with the original text from which the translated text was obtained (step i
).

The display of the translated text in association with the original text can be done, for example, by dividing the display screen of the display device 3 into two left and right, and using one side as the original text display area 3a and the other as the translated text display area 3b, in which the original text and the translated text are written horizontally. It is done by displaying. At this time, display control is performed to detect the line position at which the display of the original text from which the translated text is to be started is detected, and to start displaying the translated text from the same line as the original text.

Through such display control, the original text and its translated text are displayed side by side and correlated with each other, as shown in FIG. 3(e).

FIG. 4 shows an example of control processing in the display control unit 6 for displaying the correspondence between the original text and the translated text.

To briefly explain this process, first, parameters (F, fl, m, n) for display control are initialized (step A). Each time a character is input from the input section 1 (step B), it is determined whether the flag F is 11" or not (step C). If the flag F is 1°, it is determined that the original human power has started. Then, the value of the parameter g is set as display start line data fIn of the sentence, and a sentence number n is set for the sentence input here (step D).

Thereafter, the value of the parameter n specifying the sentence number is incremented, and at the same time, the flag F is reset to '0' (step E).

However, if the flag F is not "11" in step C, and after the processes of steps D and H are performed, it is determined whether the input character is a "line feed" instruction (step F ).

If it is not a "line break" instruction, the parameter m,
The input character is displayed at the display position (R, m) indicated by n (step G), the parameter m is incremented (step H), and the display position of the cursor lO is moved to the next input character display position.

At this time, it is determined whether the parameter m exceeds the maximum number of display characters M in one line (step I).

And if the parameter m exceeds its maximum value M,
The parameter g that specifies the display line is incremented, and the parameter m for the display character position in that line is incremented by "
1" (step J). Then, after determining whether the parameter g of the display line is the last line on the display device 3 (step K), repeat the character input process from step B described above. .

Incidentally, if the value of the parameter g exceeds the final row, the process moves to the next page and the processes from step A described above are repeated.

Through the series of processes shown in steps A to K, the human character strings are sequentially displayed as character strings.

On the other hand, when the "line break" instruction is detected in step F, the display control unit B sets the parameter g of the display line to (
+2) Set L, parameter m and flag F to “1”.
At step L), the next input sentence is processed. At the same time, the input sentence specified by number n is read from the original text storage unit 2. At step M), the translation process for that sentence is executed as described above. (Step N
).

Then, the original text (character string) indicated by sentence number n is formatted according to the number of display columns (step 0). Then, the display start position of the display part where the original text obtained above is displayed is determined from the display start line data J7n, and this un
The translated sentence is displayed starting from the line position indicated by .

By controlling the display position of the translated text in this manner, the translated text obtained by the translation device main body 5 is displayed from the same line position in association with the original text from which the translated text was obtained. Then, by the process shown in step L, a blank line is formed before the display position of the next input original sentence and its translation, and each sentence separated by translation processing unit is individually displayed as one group. It is now possible to do so.

In the example shown in FIG. 3C, the Japanese translation candidate for the input original text shown in English is rYesterday, he saw a girl with a bag on the right bank of the river. ”.

Note that when the translation device body 5 is executing the translation process on the input original text in this way, the next sentence may be inputted via the input unit 1 as, for example, rShe was pretty, J, etc. Needless to say. In this case, by instructing the translation control unit 4 to start translation for that input sentence, when the translation processing of the input original text described above in the translation device body 5 is completed, the translation for the next input sentence can be started. The translation control unit 4 immediately instructs the start of the process.

Therefore, regardless of whether or not a translation result is obtained for the input original text, the user can continuously perform the manual processing while sequentially specifying the breaks in the input original text using "line feed" instructions.

As described above, according to this device, by simply inputting the character strings to be subjected to translation processing in sequence and instructing the "line feed" that indicates the break between the sentences, the line break instruction specifies the break between sentences in the input original text. At the same time, this instruction automatically instructs the start of translation using one sentence cut out as a unit of translation processing.

Therefore, it is no longer necessary to input the original text and instruct the translation process, which are completely different tasks, separately, as in the past, and the burden on the input person (user) can be significantly reduced.

In addition, since the input sentence break and the start of the translation process are specified at the same time by the line feed J instruction, the translation device main body 5
Since the units of sentences to be translated are clearly specified, problems such as the translation process becoming unnecessarily complicated because multiple consecutive sentences are translated as one sentence are caused. Never. Therefore, it is possible to reduce the processing load on the translation device main body 5 itself, and it is also possible to significantly reduce the possibility of translation processing failure.

Therefore, it is possible to improve the efficiency of translation processing overall, and since the user only has to remember to specify "line feed," it is possible to significantly improve the operability. It has great practical effects.

By the way, in the above example, while inputting the original sentence, the sentence break was given by a "line break" instruction, and the translation process for that sentence was instructed to start. , is done as follows.

That is, as shown in FIG. 3(c), a translation candidate is obtained, and if you wish to modify it, for example, as shown in FIG. 3(d), place the cursor IO on the part to be modified and click Operate the "Line feed (translation)" key.

In this case, since it is not the character string input mode, the translation control unit 4 inputs the word (phrase) specified by the cursor
Detects the original sentence (one sentence required as one translation processing unit) containing the sentence, and instructs to re-execute the translation processing for this sentence. At this time, the attributes such as part-of-speech information of the translation of the word (phrase) specified by the cursor 10 are checked, and the modification is selected differently to obtain a translation different from the translation candidate obtained earlier. The translation process is controlled.

With this kind of processing, even if you are in the translation correction mode, just like in the original text input mode described above, "Line break j
This instruction alone will instruct the extraction of a single sentence to be translated and the start of its translation process.

In addition, in this correction mode, if there is a "period" etc. that indicates a sentence break immediately before or after the word (phrase) indicated by the cursor lO, use this "period" as a break and proceed as follows. The input original text processed as one sentence may be divided into two, and each of these divided sentences may be translated as an independent translation processing unit.

If you use this kind of processing sequence together, you can instruct a new line break in the places where you forgot to specify a line break, and re-execute the translation process for each sentence divided by this delimiter specification. It becomes possible to give instructions.

In addition, even if the desired translation cannot be obtained because the translation processing unit of the original text that was translated earlier is long, by re-dividing the translation processing unit, the short sentence that was cut by this can be processed. It becomes possible to perform translation processing for each unit.

At this time, for each sentence divided into two, for example, when the translation process for the previous sentence is finished, the translation process for the next sentence is sequentially instructed to start. Of course, the instruction may be given automatically.

Note that the present invention is not limited to the embodiments described above. Although we have explained here the translation of English text into Japanese text, it goes without saying that the reverse is also possible. Further, the present invention can be similarly applied to cases where the original text is input by voice or input via OCR or the like. Furthermore, although the start of translation is instructed by a "line break," other instruction formats may be used as long as the instruction is to end the input of one sentence of the original text to be translated. In addition, the present invention can be implemented with various modifications without departing from the gist thereof.

[Effects of the Invention] As explained above, according to the present invention, an instruction for separating sentences in the original text and an instruction for translation processing for a single sentence for which the separation is specified can be performed by, for example, operating a "line feed" key. Since these operations can be performed simultaneously, the operation format can be simplified.

Furthermore, it is possible to prevent problems such as forgetting the instruction to start translation and translating a plurality of consecutive sentences as one sentence. Therefore, it is possible to instruct the translation process while reliably performing the human work on the original text, which has a great practical effect, such as reducing the burden on the translation process and its users (operators), and enabling reliable and efficient translation processing. The effect is produced.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a schematic configuration of a translation processing device according to an embodiment of the present invention, FIG. 2 is a diagram showing a basic processing sequence of the embodiment device, and FIG. 3 is a diagram showing the original text and its translation processing in the embodiment device. FIG. 4 is a diagram showing the display format with the translated text, and FIG. 4 is a diagram showing the sequence of display control. ■...Input unit, 2...Original text storage unit, 3...Display device, 4...Translation control unit, 5...Translation device main body, 6...
. . . Display control section, 7. Translation dictionary, and . . . Translation storage section.

Claims (3)

[Claims] (1) An input means for inputting original text information and control information related to the input of this original text information, a translation device main body for translating the input original text information using a translation dictionary and obtaining the translated text, and this translation device main body and a translation processing control means for controlling the operation of the translation processing control means, when control information instructing to finish inputting the original text information is inputted via the inputting means, the translation processing control means A translation processing device, characterized in that the translation processing device instructs the translation device main body to start translation processing using the original text information inputted in the translation processing unit as one translation processing unit. (2) The translation processing device according to claim 1, wherein the control information for instructing the end of inputting the original text information includes information specifying a line break. (3) When control information instructing the division of the original text information is input, the translation processing control means divides the original text information into one translation processing unit at the position divided by this control information, and divides the original text information into the translation processing unit. 2. The translation processing device according to claim 1, wherein the translation processing device instructs the main body to restart translation processing.