JPH0332106B2 - - Google Patents

Description

[Detailed description of the invention]

INDUSTRIAL APPLICATION FIELD The present invention relates to an interactive translation device in which a machine and a human cooperate to create a correct translated sentence. Background Art Generally, machine translation is performed through a process as shown in FIG. 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 dependency relationships that indicate subordination between words. Semantic analysis is the process of determining what is correct and what is not from multiple syntactic analysis results. Machine translation performs analysis to any of the 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 It is what generates the output language. 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 will output all grammatically correct interpretations, but this will result in outputting a wide variety of translation results, increasing the amount of time it takes for humans to determine the correct answer. Those who go as far as interpreting the meaning are
Although in principle it is possible to output only one correct translation result, 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, it can translate sentence by sentence, output all grammatically correct interpretations, and even perform semantic analysis.
This refers to machine translation that is not perfect and cannot narrow down the output to be unique. FIG. 7 is a block diagram showing the configuration of a typical translation device. A keyboard 2 for 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 to be stored is provided. The translation module 5 stores the results of converting the contents of an original text buffer 51 for storing input sentences, a dictionary lookup buffer 52 for storing dictionary lookup results, and a composition analysis buffer 53 for storing syntactic analysis results of the input language for the output 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. FIG. 8 shows that when the input sentence is Time flies like an arrow. when translated, each buffer 51-
The contents stored in 55 are shown. Figure 8 1
indicates 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 buffer 52 extracts grammatical information such as part of speech and translation information corresponding to each word. The parsing buffer 53 shown in FIG. 82 parses the input sentence based on part-of-speech information given to each word. FIG. 8 shows the contents of the syntax generation buffer 54 and the result buffer 55. The result of converting the contents of the syntax analysis buffer 53 into the output language is the syntax generation buffer 5.
4, and the translation result is stored in the result buffer 55 based on the generated syntax. FIG. 9 is a diagram showing a display screen of the display device 4 in the prior art. When an input sentence to be translated is inputted as shown in FIG. 91 and a translation key provided on the keyboard 2 is operated, as shown in FIGS. 92 to 96, The translation result will be displayed. The translated sentences shown in Figure 9 2 to Figure 9 6 on each display screen are based on the input sentence as shown in Table 1 below. This is the result of recognition based on the mutual relationships between the constituent elements of a sentence.

[Table] All of these are grammatically correct interpretations.
Of course, humans can judge that the interpretation shown in FIG. 96 is correct. If the semantic analysis is complete, the translated sentence will be as shown in Figure 9.6, but for that, (a) arrows and flies are not similar, and (b) arrows do not have the ability to measure time. (c) It is necessary to make the machine remember the knowledge that flies do not like arrows. 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, machine translation searches for all kinds of possibilities and finds all possible solutions, which increases the time it takes to get the correct answer. For this reason, we have adopted a method of outputting solutions in order, starting with those found.
Therefore, at the time when the first solution is output in FIG. 92, 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 it is impossible to know the maximum number of times the translation key must be pressed to get the correct answer. Therefore, the mental burden on the operator is quite large. From this point of view as well, it is desired to reduce the total number of solutions by even one with a simple operation. Problems to be Solved by the Invention In the above-mentioned translation results, it can be seen that the mistranslations shown in FIGS. 9 2 to 5 are caused by erroneous identification of the constituent elements of sentences. In the example in Figure 9 2, “like an arrow” is “Time flies”
The sentence is erroneously recognized as an adjective. However, such errors can be easily noticed by a person who has the ability to judge the success or failure of a translation by looking at it. An object of the present invention is to provide a translation device that eliminates ambiguity in translation results in machine translation and allows a human to provide a simple instruction to obtain correct translation results. Means for Solving the Problems The present invention provides a translation device that translates a sentence in a first language into a sentence in a second language. an input means capable of specifying a modifying relationship between constituent elements or words in a sentence; and an input means capable of specifying a modifying relationship between constituent elements or words in a sentence; If the component is not specified by the input means, the type of component is determined and translated according to predetermined rules, and the input method is used to translate the component. When a modifying relationship between words is specified, translation is performed using the specified modifying relationship, and for words that are not specified by the input means, the modifying relationship between words is determined according to predetermined rules. a means for translating a sentence in a first language into a sentence in a second language; a means for visually displaying a sentence in the first language inputted by the input means; For words with specified component names, the specified component name is visually displayed near the position where the word is displayed, and for words with a modification relationship between words, related words are displayed. The modification relationship is visually displayed in the vicinity of the displayed position, the translated sentence in the second language is visually displayed by the visual display means, and the constituent elements are displayed in the first language sentence by the input means. For words that are not specified, the component name determined by the translation means is displayed near the position where the word is displayed in a display format different from the component name specified by the input means. However, for words for which a modifying relationship between words has not been designated by the input means, the modifying relation determined by the translation means is designated by the input means in the vicinity of the position where the related word is displayed. This is a translation device characterized by including means for displaying a modification relationship in a display mode different from that of the modified modification relationship. Effects According to the present invention, a sentence in the first language can be input to the input means, and the constituent elements of the words constituting the sentence and the modifying relationships between the words can be specified. The input sentence in the first language is visually displayed on the visual display means. The specified component name and qualifying relationship are also displayed near the position where the corresponding word is displayed. The translation means converts the first language into the first language using the constituent elements and modification relations specified by the input means and the constituent elements and modification relations between words that are not specified and determined according to predetermined rules. Translate into two languages. The translated second language is displayed by the visual display means. The visual display means displays the component name determined by the translation means and the modification relationship between words of the sentence in the first language in a display manner different from the specified component name and modification relationship. Ru. By visually checking such translation results, it is possible to specify more effectively the next time a sentence in the first language is input. Embodiment FIG. 1 is a block diagram showing the configuration of a translation device capable of implementing the present invention. In relation to the processing device 1, there is a keyboard 2 which is an input means for key-inputting data, a memory 3 in which data related to the processing device 1 is stored, and a visual display means for displaying the results processed by the processing device 1. A display device 4 and a translation module 5 in which data related to translation are stored are provided. The translation module 5 includes an original text buffer 5 that stores the input text.
1. Dictionary lookup buffer 52 that stores the result of dictionary lookup, syntactic analysis buffer 53 that stores the syntax analysis result of the input language that is the first language, and conversion of the contents of the syntax analysis buffer 53 to the output language that is the second language. Syntax generation buffer 5 that stores the results
4. A result buffer 55 for storing the results of morpheme generation in the output language, a table 56 comprising a dictionary for machine translation, grammatical rules, etc., and conditions regarding the interrelationships between sentence components specified by the operator according to the present invention. A syntax analysis check buffer 57 is included in which the syntax analysis check buffer 57 is stored. FIG. 2 is a flowchart for explaining the 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, in 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 parsing 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 a word of the input sentence or a translation key for instructing translation. When inputting a word, the process moves to step n3, where 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 input character code of the translation module 5 is stored in the original text buffer 51, 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 operator has instructed the words of the input sentence as to the interrelationships between the constituent elements of the sentence. Step without instructions
Returning to step n2, when instructed, the process moves to step n7, and the interrelationships between the constituent elements of the sentence are displayed according to the instruction. FIG. 31 shows a state in which "Time" and "flies" are specified to have a subject/verb relationship. At step n8, mutual relationship information between constituent elements, which is a modification relationship of the designated sentence, is stored in the syntax analysis check buffer 53. While the input of the original text continues in this manner, the processing operations of steps n2 to n8 are 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, the process moves to step n9. In step n9, the combination of parts of speech is initialized. When the combination of parts of speech is initialized, table 56 is set in step n10.
The parsing is performed with reference to the parsing rules, and the parsing results are stored in the parsing buffer 53. At step n11, it is determined whether the parsing was successful or not. If the parsing is successful, the process moves to step n12, where it is determined whether the conditions of the parsing check buffer 57 are satisfied. step n8
If the stored conditions are not met, the process returns to step n10 and parsing is performed again. When the above conditions are satisfied, the process moves to step n13, where the translation module 5 performs structural conversion while referring to the conversion rules and production rules in the table 56. Next, in step n14, syntax generation is performed,
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. That is, steps n3 to n5 and steps n10 to n15 constitute translation means. Further, step n16 constitutes means for displaying sentences in the second language, component names corresponding to words in the first language, and modification relationships between words. If the operator looks at the display and determines that it would be faster to re-specify the component or the modifying relationship, he or she can input the original text and re-designate the component or the modifying relationship. In step n17, it is determined whether or not the operator has inputted the next candidate key in order to generate the next candidate when the translation result is incorrect. When the next candidate key is input, the process moves to step n18 and the next candidate flag is set. In this way, after the parsing is successful at least once and the translation result is displayed in step n10, when the operator requests another translation result, the next candidate flag is set. Step n18 returns to step n10, where the parsing buffer 53 remembers the previous combination of parts of speech and the result of parsing, and if the combination of parts of speech remains the same as before, it uses a different syntax from the previous one. There is a function to explore analysis results.
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 proceed from step n11 to step n19.
Move to. In step n19, it is determined whether all parts of speech in the dictionary lookup buffer 52 have been combined. However, at this stage, all parts 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 the syntactic analysis of step n10 is performed again.
As a result, if the syntax analysis is successful, processing operations from step n11 to steps n11 to n16 are performed. If the translation result is also incorrect, proceed from step n17 to step n19 according to the operator's next candidate instruction.
The process returns to step n10, parsing the syntax. At this time, if there is another solution for the same combination of parts of speech, the step
Processing operations n11 to n16 are performed. If the result is again an error, the process moves from step n17 to step n18 according to the operator's instruction, and the syntax analysis of step n10 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, n10, n11, n12, n13, n14,
Processing operations are performed in steps n15 and n16, and the solution is displayed to wait for the operator's decision. If there is no other solution, steps n18, n10, and n11 are performed, and in step n19 it is determined whether or not all part-of-speech combinations have been completed. If not completed, a new combination of parts of speech is selected in step n20, and the process returns to step n10. If the parsing is successful with the new part-of-speech combination, processing is performed from step n11 to steps n12 to n16, the solution is displayed, and the operator waits for a decision. If unsuccessful, the process moves from step n11 to step n19, and a check is made to see if all parts of speech have been combined. As shown in Figure 3, when the parsing is successful for the combination of parts of speech and the resulting translated sentence is correct, the operator moves from step n17 to step n23 and starts the next sentence. 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, and if the next sentence is not input, the translation ends. When all parts of speech have been combined in step n19, the process moves to step n21, where it is determined whether the next candidate flag has been set. If the next candidate flag is set, move to step n6,
If it has not been set, proceed to step n22.
In this case, if the processing operation is performed in a loop of steps n10, n11, n190, and n20, the next candidate flag remains unset. At this time, step
An error message is displayed at n22. Similar to FIG. 3, FIG. 4 shows the display screen of the display means 4. As shown in FIG. FIG. 4 1 shows a state in which the interrelationships between the constituent elements of the sentence instructed by the operator in step n6 described above are displayed. FIG. 42 shows the translation result obtained by indicating that "like an arrow" modifies "flies" as an adverb. As an instruction method, for example, first, the constituent elements a and b of the two sentences to be specified are determined by region specification, and
When the related direction c is determined by key operation and a key indicating the interrelationship between the constituent elements of the sentence provided in advance is operated, the interrelationship d between the constituent elements of the inputted sentence is determined corresponding to the said constituent element. can be displayed. Here, a case has been described in which only one mutual relationship between constituent elements of a sentence is specified, but of course, instructions may be given not only for one mutual relationship but also for a plurality of mutual relationships. Also, the fourth
As shown in FIG. 1, one component may be specified for multiple words. That is, a word that is a single word or a combination of multiple words can be specified. As described above, in the present invention, the number of translation candidates can be reduced compared to the conventional technology, and the correct answer can be obtained quickly. FIG. 5 shows the contents stored in each buffer 51-55 in the present invention. FIG. 5 shows the contents of the original text buffer 51 and dictionary lookup buffer 52. Original text Batsuhua 5
Based on the words input in 1, the dictionary lookup buffer 52 extracts grammatical information such as part of speech and translation information corresponding to each word. The parsing buffer 53 shown in FIG. 5 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. FIG. 5 shows the contents of the syntax generation buffer 54 and result buffer 55. Based on the parsed content, a syntax in the output language is generated in the syntax generation buffer 54, and the content of the translated sentence is stored in the result buffer 55 based on the structure. Effects of the Invention As described above, according to the present invention, by specifying component names corresponding to words in a sentence in a first language and mutual modification relationships between words, translation results in machine translation can be improved. It becomes easy to resolve the multiplicity of words and obtain correct translation results into sentences in the second language. Further, according to the present invention, component names or modifying relations between words determined by the translation means and corresponding to words not specified by the input means are translated into the second language by making them correspond to sentences in the first language. can be displayed on the display means together with the translated sentence. Therefore, when translating the same sentence in the first language again, the method for specifying constituent elements or modification relationships can be improved so that the translation can be corrected more quickly and the translated sentence can be obtained. Furthermore, since the constituent elements of a sentence and the modifying relationships between words can be specified in advance, it is easier to obtain a correct translated sentence. Furthermore, according to the present invention, how the translation means determines the constituent elements and modification relationships for sentences in the first language according to predetermined rules is displayed. There are also easily recognized educational benefits. Furthermore, according to the present invention, the component name and the modifying relationship are respectively displayed near the position where the corresponding first language word is displayed and near the position where the related word is displayed, and , is displayed in different display modes depending on whether or not it is specified in the input means. Therefore, it is possible to easily recognize the results determined by the translation means, and it is also easy to improve the method of specifying components.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of the translation device, FIG. 2 is a flowchart for explaining the operation according to the present invention, FIGS. 3 and 4 are diagrams showing the display screen of the display means 4, and FIG. 6 is a diagram showing the contents stored in each buffer 51 to 55 in the present invention, FIG. 6 is a diagram for explaining the process of machine translation, FIG. 7 is a block diagram showing the configuration of a conventional translation device, and FIG. FIG. 8 is a diagram showing the contents stored in each buffer 51 to 56 in the prior art, and FIG. 9 is a diagram showing a display screen by the display device 4 in the prior art. 5... Translation module, 51-55, 57...
Batsuhua, 56...table.

Claims (1)

[Scope of Claims] 1. In a translation device that translates a sentence in a first language into a sentence in a second language, a sentence in the first language is input, and the constituent elements in the sentence of words constituting the sentence in the first language are input. Or an input means that can specify the modifying relationship between words, and a word in a sentence in the first language input by the input means, if the constituent elements are specified by the input means. Translation is performed using specified constituent elements, and when no constituent element is specified by the input means, the type of constituent element is determined and translated according to predetermined rules, and the inter-word modification is performed by the input means. When a relationship is specified, translation is performed using the specified modification relationship, and for words that have not been specified by the input means, the modification relationship between words is determined and translated according to a predetermined rule. means for translating a sentence in a second language into a sentence in a second language; a visual display means for visually displaying a sentence in the first language input by the input means, and for specifying a component name. For words that have been specified, the specified component name will be visually displayed near the position where the word is displayed, and for words that have a modification relationship between words, the name of the specified component will be visually displayed near the position where the related word is displayed. The modification relationship is visually displayed in the vicinity of , the translated sentence in the second language is visually displayed by the visual display means, and the word whose constituent elements are not specified by the input means in the sentence in the first language. , the component name determined by the translation means is displayed near the position where the word is displayed in a display mode different from the component name specified by the input means, and the input means Therefore, for words for which no modification relation between words has been specified, the modification relation determined by the translation means is placed near the position where the related word is displayed, and the modification relation specified by the input means is and means for displaying in different display modes.