JPS6180358A - Translation system - Google Patents

Abstract

PURPOSE:To obtain a correct translated result through a simple specification by specifying at least one out of plural constitutional elements indicating the constitution of an input sentence to the words of the input sentence to translate the input sentence. CONSTITUTION:The titled system is provided with a keyboard 2 for inputting data related to a processor 1, a memory 3 for storing the data related to the processor 1, a display device 4 for displaying a result processed by the processor 1, and a translation module 5 for storing the data related to the translation. An input sentence to be translated is displayed on the display device 4 and at least one out of plural constitutional elements of the input sentence is specified to the words of the input sentence through the keyboard 2 to translate the input sentence.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to an interactive translation system in which a machine and a human cooperate to create a correct translated text.

Background technology General E Il! 1 Machine translation is performed through the process shown in Figure 7. The source text consisting of the input language to be translated needs to be analyzed during the translation process. There are three stages of analysis: morphological analysis, syntactic analysis, and semantic analysis. Morphological analysis refers to looking up a dictionary for machine translation, extracting grammatical information such as part of speech and translation information for each word, and analyzing the person, number, sentence tense, etc. Syntactic analysis is the analysis of the structure of a sentence by examining the dependencies and relationships between words. Semantic analysis is the process of determining which are correct and which are incorrect from multiple syntactic analysis results. Machine translation performs analysis to one of three levels to obtain the internal structure of the input language, then converts it into the same level internal structure of the output language that makes up the translated sentence according to that structure, and then converts it to the same level internal structure of the output language that makes up the translated sentence. is generated. The accuracy of machine translation varies depending on the depth of this analysis level. Those that only perform morphological analysis are unable to translate sentence by sentence, but only by word by word, as typified by so-called electronic translators. Those that perform syntactic analysis output all grammatically correct interpretations, but this results in outputting a wide variety of translation results, which increases the amount of time and effort it takes for humans to determine the correct answer. In principle, a device that performs semantic interpretation can output only one correct translation result, but this requires the machine to memorize an enormous amount of information, which is nearly impossible in reality. The machine translation that is the object of the present invention is defined as at least the level of syntactic analysis. In other words, machine translation is capable of sentence-by-sentence translation, can output all grammatically correct interpretations, and can even perform semantic analysis, but it is not perfect and cannot narrow down the output to a unique one.

FIG. 8 is a block diagram showing the configuration of a typical translation device. A keyboard 2 for manually inputting data related to the processing device 1, a memory 3 for storing data related to the processing device 1, a display device 4 for displaying results processed by the processing device 1, and data related to translation. A translation module 5 in which is stored is provided. The translation module 5 includes an original text buffer 51 that stores input sentences, a dictionary lookup buffer 52 that stores dictionary lookup results, and a composition analysis buffer 53 that stores syntactic analysis results of the input language. It includes a syntax generation buffer 54 for storing morphemes of the output language, a result buffer 55 for storing the results of morpheme generation of the output language, and a table 56 consisting of a dictionary for machine translation, sentence rules, and the like. The wAQ diagram shows that when the input sentence is translated,
rrou+.

This also shows the contents stored in each buffer 51 to 55 in the prior art. FIG. 9(1) shows the contents of the original text buffer 51 and the dictionary lookup button 7a 52. Based on the words entered into the original text buffer 51, the dictionary lookup buffer 52 retrieves grammatical information such as part of speech and translation information corresponding to each word. The parsing buffer 53 shown in FIG. 9(2) parses the input sentence based on part-of-speech information given to each word. Figure 9 (3)
The contents of the syntax generation buffer 54 and the result buffer 55 are shown in FIG. The result of converting the contents of the syntax analysis buffer 53 for the output language is stored in the syntax generation buffer 54, and the result buffer 55 is stored based on the generated syntax.
Translation results are stored in .

FIG. 10 is a diagram showing a display screen of the display device 4 in the prior art. When the input sentence to be translated as shown in FIG. 10 (1) is inputted and the translation key provided on the keyboard 2 is operated, as shown in FIG. 10 (2) to #10. The translation result as shown in (6) is displayed. Each display screen Figure 10 (2) to Figure 10 (
The translated sentences shown in 6) are based on the combination of so-called sentence components such as subjects, verbs, objects, complements, adjective phrases, and adverb phrases, as shown in Table 1 below. This is the result.

(Margin below) These are all grammatically correct interpretations.

Of course, humans can judge that the interpretation shown in Figure 10 (6) is correct.If the semantic analysis is complete, the translated sentence will be as shown in tIS Figure 10 (6). In order to do so, we must know that (a) arrows and flies are not similar, (b) arrows do not have the ability to measure time, and (c) flies do not like arrows. need to be stored in the machine. It is obvious that it is impossible for a machine to memorize all of this knowledge about the real world. Therefore, it can be seen that it is inevitable that there are solutions that should not be compatible at the conventional level of machine translation.

As described above, the conventional machine translation system has a drawback in that, since there are many solutions, the translation key must be pressed many times until the correct answer is obtained. Furthermore, in machine translation, all kinds of possibilities are explored and all solutions are found, so it takes a long time to obtain the correct answer.Therefore, a method is used in which the solutions are output in order from those found.

Therefore, at the time when the first solution is output in FIG. 10(2), it is not known how many remaining solutions exist. Things like the number of remaining homophones in kana-kanji conversion cannot be displayed. For this reason, there is not only the disadvantage that the translation key must be pressed many times, but also the disadvantage that the correct answer cannot be obtained even after pressing the translation key a maximum number of times.

Therefore, the mental burden on the operator is quite large. From this point of view as well, it is desirable to reduce the total number of solutions by at least one through simple operations.

In the above translation results, the problems that the invention aims to solve)
It can be seen that the mistranslations shown in 5) are caused by incorrect recognition of the constituent elements of the sentence. In the example in Figure 10 (2), “flies” is a part of the noun phrase, “1
``ike an arrow'' is an adjective phrase.However, such a mistake would be immediately obvious to someone who has the ability to judge the success or failure of a translation by looking at the translation result. It is.

An object of the present invention is to provide a translation method that eliminates the multi-window nature of translation results in machine translation and allows a human to provide a simple instruction to obtain correct 7 translation results.

Means for Solving the Problems The present invention displays an input sentence to be translated, and specifies at least one component representing the structure of the input sentence with respect to a word of the input sentence=7. - This is a translation method characterized by translating the input sentence.

According to the present invention, an input sentence to be translated is displayed, at least one component representing the structure of the input sentence is designated for a word of the input sentence, and the input sentence is translated. (translation), you can obtain correct translation results with simple operations.

Embodiment FIG. 1 is a block diagram showing the configuration of a translation device that can implement the present invention. Related to the processing device 1 are a keyboard 2 for manually inputting data, a memory 3 for storing data related to the processing device 1, a display device 4 for displaying results processed by the processing device 1, and data related to translation. A translation module 5 in which is stored is provided. The translation module 5 includes an original text buffer 51 that stores input sentences, a dictionary lookup buffer 52 that stores dictionary lookup results, a syntax analysis buffer 53 that stores the syntax analysis results of the input language, and an output 8 that outputs the contents of the syntax analysis buffer 53. Syntax generation buffer 5 that stores the result of conversion into pragmatics
4. A result buffer 55 for storing the results of morpheme generation of the output language, a table 56 consisting of a dictionary for machine translation, grammatical rules, etc., and a syntactic analysis for storing conditions related to sentence components specified by the operator according to the present invention. A check buffer 57 is included.

FIG. 2 is a 70-chart illustrating operation according to the present invention. Further, FIG. 3 shows an example of the display screen of the display device 4 according to the present invention. First step n1
, the next candidate flag is cleared prior to inputting the translation original text. This flag is used to display an error message when no solution is obtained even after syntactic analysis is performed on all combinations of parts of speech, as will be described later.

Next, when the operator operates a key on the keyboard 2, it is determined in step n2 whether the input is the input of 141 words of the input sentence or the input of a translation key for instructing translation.

When tl is input as a word, step n31: the word is displayed on the display means 4 via the processing device 1, and at the same time, the character code is sent from the processing device 1 to the translation module 5. At step n4, the character code input to the translation module 5 is stored in the original text buffer 5'1, and the dictionary of the table 56 is looked up to perform morphological analysis. When the morphological analysis is performed and the process moves to step n5, the analysis result is stored in the dictionary lookup buffer 52. Proceeding to step n6, it is determined whether or not the constituent elements of a sentence have been indicated for the word inputted by the repeater. If no instruction is given, the process returns to step n2, and if an instruction is given, the process proceeds to step n7, where the names of the constituent elements of the specified sentence are displayed. At step n8, the components of the designated sentence are stored in the parsing check buffer 53. The components of this designated sentence become the limiting conditions in step n12, which will be described later. While inputting the original text continues in this way, steps n2 to n8
This processing operation is repeated.

When the input of the original text is completed and the operator issues a translation instruction by operating the translation key in step n2,
Proceed to step n9. FIG. 3(1) shows the original text in which word input has been completed and a state in which the operator has specified "Time" as the subject.

In step n9, the combination of parts of speech is initialized.

Once the combination of parts of speech is initialized, syntactic analysis is performed in step n10 with reference to the parsing rules in table 56.
The analysis result is stored in the syntax analysis buffer 53.

In step 1111, it is determined whether or not the parsing was successful. If the parsing is successful, the process moves to step n12, and it is determined whether the conditions of the parsing check buffer 57 are satisfied. If the condition that "fl ies" is the subject stored in step n8 is not satisfied, the process returns to step nlo and parsing is performed again. When the above condition is satisfied, step n13
, in translation module 5 table 56
Structural transformation is performed while referring to the transformation rules and production rules. Next, in step nl 4, syntax generation is performed and the syntax generation buffer 54 is completed. Then, in step n15, the result buffer 55 is completed, and the result is displayed on the display means 4 in step n16.

In step n17, if the translation result is incorrect, it is determined whether or not the next candidate key has been manually input by an operator in order to generate the next candidate. When the next candidate key is manually entered, the process moves to step nlB and the next candidate flag is set. In this way, in step n10, the parsing is 1
After successfully displaying the translation result, when the repeater requests another translation result, the next candidate flag is set. Step n again from step n18
Moving on to the parsing of lO, the parsing buffer 53 stores the previous combination of parts of speech and the result of parsing, and when the combination of parts of speech is unchanged from the previous one, it has a function to search for a parsing result different from the previous one. be. At this time, if there is no other solution for the same combination of parts of speech, the analysis will fail and the process will move from step nil to step n19.

In step n19, it is determined whether all the parts of speech in the dictionary lookup table 7a 52 have been combined. However, at this stage, all part-of-speech combinations in the dictionary lookup buffer 52 have not been completed, so the process moves to step n20, a new part-of-speech combination is set, and step n20 is entered again.
A parsing of lo is performed. As a result, if the parsing is successful, step nl 1 to step n12
-n16 processing operations are performed. If the translation result is also incorrect, step n17 is performed according to the operator's next candidate instruction.
The process then moves to step n19, and the process returns to step nlO for syntax analysis. At this time, if there is another solution for the same combination of parts of speech, the processing operations of steps nil to n16 are performed. If this result is again erroneous, the process moves from step n17 to step nlB according to an instruction from the operator, and the syntax analysis in step nlO is performed again. The detailed operation after this will be complicated, so a detailed explanation will be omitted, but in general, if there is another solution for the same part-of-speech combination, step n18 = nlO, nl 1 , nl 2 , nl
3, nl 4, nl 5, and n16 are performed, and the solution is displayed and waits for the operator's decision. If no alternative solution exists, steps nl 8 , n
After the processing operations l O and nl 1 are performed, the process moves to step n19, where it is determined whether or not all part-of-speech combinations have been completed. If not completed, step n2
0, a new combination of parts of speech is selected and the process returns to step nlO. If the parsing is successful with the new part-of-speech combination, steps n12 to n16
The processing operation is performed, the solution is displayed, and the operator's decision is awaited.

If it fails, move from step nil to step n19,
A check is made to see if the combination of all parts of speech has been completed.

As shown in Figure 3 (2), when the parsing is successful for a combination of parts of speech and the resulting translated sentence is correct,
Step n17 to step n2 depending on the operator's judgment.
3, it is determined whether or not to input the next sentence. If the next sentence is to be input, the process moves to step n1; if the next sentence is not input, the translation ends.

When the combination of all parts of speech is completed in step n19, the process moves to step n21, and it is determined whether the next candidate flag is set. If the next candidate 7 blowfish is set, the process moves to step 9; if it is not set, the process moves to step n22. in this case
Steps nl O, nl 1, nl 9 (1tn20
If processing operations are performed in this loop, the next candidate flag remains unset. At this time, step n22
An error message will be displayed.

Similar to FIG. 3, FIGS. 4 and 5 show the display screen of the display means 4. Figure 4 (1) and Figure 5 (1) show that the input of the original text to be translated is completed,
The state in which the component names of the sentence designated by the repeater in step n6 above are displayed together with the original sentence is shown. Figure 4 (1) shows the verb "flies", and Figure 4 (2) shows the verb "flies".
"an arrow +" is designated as an adverb phrase.The method of designation is 1. For example, when you perform a key operation to designate the constituent elements of a sentence, the cursor moves word by word, and the word where the cursor is located is displayed. On the other hand, when a preset key indicating the name of a sentence component is operated, the name of the component of the input sentence can be displayed corresponding to the word.Here, when inputting one word, For example, as shown in Figure 4, fl
ies" as a verb, the solution is uniquely determined,
As shown in FIG. 4(2), the correct answer can be obtained with the first candidate. Also#! +5 As shown in the diagram 1ike a
If "n arro-" is designated as an adverbial phrase, the correct answer can be obtained with two candidates as shown in FIG. 5(2) and FIG. 5(3).

In either case, the number of translation candidates can be reduced compared to the prior art, and the correct answer can be obtained quickly. In the present invention, it is sufficient to specify only one word as a component of a sentence at the time of input, but of course the instruction is not limited to one word, and a plurality of words may be specified. For example, specify ``Time'' as the subject and write 1ike an arr.
ow” can also be designated as an adverbial phrase, and in this case, if the total number of translation results indicates a part of speech of one word, the number of uplinks will be reduced.

FIG. 6 shows the contents stored in each buffer 51-55 in the present invention. FIG. 6(1) shows the contents of the original text buffer 51 and dictionary lookup buffer 52.

Based on the words input to the original text buffer 51, the dictionary lookup 7a 52 extracts grammatical information such as part of speech and translation information corresponding to each word. The parsing buffer 53 shown in FIG. 6(2) parses the input sentence based on part-of-speech information given to each word and the constituent elements of the sentence specified according to the present invention. Figure 6 (
3) shows the contents of the syntax generation buffer 54 and result buffer 55.

Based on the parsed content, the syntax generation buffer 54
Then, a syntax in the output language is generated, and the contents of the translated sentence are stored in the result buffer 55 based on the structure.

Effects of the Invention As described above, according to the present invention, it is possible to solve the problem of high yield of translation results in machine translation, and to obtain correct translation results by giving simple instructions from a human.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of the translation device, and FIG. 2 is a block diagram 70-char for explaining the operation according to the present invention.
Figures Aa to 5 are diagrams showing the display screen of the display means 4, Figure 6 is a diagram showing the contents stored in each buffer 51 to 55 in the present invention, and Figure 7 is for explaining the process of machine translation. , FIG. 8 is a block diagram showing the configuration of a translation device of the prior art, and FIG. 9 is a block diagram showing the configuration of each buffer 5 in the prior art.
1 to 56, and FIG. 10 is a diagram showing a display screen of the display device 4 in the prior art. 5...Translation monoyl, 51-55.57...Buffer, 56...Table

Claims (1)

[Claims] Displaying an input sentence to be translated, specifying at least one component representing the structure of the input sentence for a word of the input sentence, and translating the input sentence. A translation method characterized by: