JP2004310170A - Bilingual word pair learning method and apparatus, and recording medium storing a bilingual word pair learning program - Google Patents

Abstract

A word or a word which is an element of a second natural language sentence corresponding to a word or a word string which is an element of a first natural language sentence necessary for translation processing from a first natural language to a second natural language. Provided is a learning method of a translated word pair that can automatically, accurately, and more extract a translated word pair composed of a word string.
A first natural language sentence and a bilingual second natural language sentence are input (S1), and a word or word sequence that is an element of the first natural language sentence is input to the second natural language sentence. A pair of a bilingual word or word string obtained in association with a word or word string that is an element of the language sentence is generated (S2), and the word or word string that is an element of the second natural language sentence is generated. A pair of a bilingual word or word string obtained in association with a word or word string that is an element of the first natural language sentence is generated (S3), and a pair of (2) generated in (S2) and (S3) The same bilingual word pair is extracted from the set of types of bilingual word pairs (S4).
[Selection diagram] Fig. 1

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a learning method for extracting a bilingual word pair having a bilingual relationship, and more particularly to a learning method for automatically extracting a bilingual word pair from a bilingual natural language sentence.
[0002]
[Prior art]
A method of automatically learning a bilingual word pair from a bilingual sentence pair having a bilingual relationship using statistical information (see Patent Document 1) is known. For example, a word pair having a significantly different language structure such as between Japanese and English is known. In the case of, there is a problem in the extraction accuracy of the translated word pair. Further, when such a language structure is significantly different, a learning method is known (see Non-Patent Document 1) that improves the accuracy of extracting word pairs by using the syntax information of one language.
[0003]
[Patent Document 1]
Japanese Patent Application Laid-Open No. 5-189481, "Peter Phythoe Brown""Translation computer operation method, lexical model generation method, model generation method, translation computer system, lexical model generation computer system, and model generation computer system"
[Non-patent document 1]
Kenji Yamada and Kevin Knight, "A Syntax-based Statistical Translation Model", 39th Annual Meeting of the Association for Licensing for the Promotion of Licensing. 523-530, 2001
[0004]
[Problems to be solved by the invention]
In the above-mentioned conventional bilingual word pair learning method in Patent Literature 1 and Non-Patent Literature 1, when the base language to be associated among the two languages of the bilingual sentence changes, the obtained word pair changes, and a stable word pair cannot be extracted. There's a problem. Further, in the learning method of Non-Patent Document 1, although the extraction accuracy is improved as compared with the method that does not use syntax information, the accuracy is not yet sufficient.
The present invention has been made in view of the above points, and a stable bilingual word is obtained by extracting the same bilingual word pair from bilingual word pairs obtained by using the syntax information of an input bilingual sentence one language at a time. A bilingual word pair learning method that extracts pairs, adds the extracted common bilingual word pairs to the input bilingual sentence, and repeats the learning of the bilingual word pairs to improve the accuracy of extracting the bilingual word pairs. It is an object of the present invention to provide an apparatus and a recording medium in which a learning program for a translated word pair is recorded.
[0005]
[Means for Solving the Problems]
FIG. 1 is a schematic flowchart of a method for learning a bilingual word pair for explaining the present invention.
The present invention relates to an apparatus for learning a word pair having a bilingual relationship by inputting a first natural language sentence and a second natural language sentence which is a translation thereof, and a second natural language sentence which is a translation of the first natural language sentence. (Step 1), and a word or word string as an element of the second natural language sentence is associated with a word or word string as an element of the first natural language sentence to form a set of bilingual word pairs. (Step 2), a word or word string as an element of the first natural language sentence is associated with a word or word string as an element of the second natural language sentence, and a set of bilingual word pairs is obtained ( Step 3) Extract the same bilingual pair from the set of these two types of bilingual word pairs (step 4).
[0006]
The present invention also includes a step of associating a word or word string with a syntax tree obtained by providing a syntax analysis step in the processing of associating a word or word string in the translated word pair acquisition step.
[0007]
In addition, the present invention provides a word or word string that is an element of a first natural language sentence, and a word or a word that is an element of a first natural language sentence of a common word pair obtained by extracting the same bilingual pair. For the word string, a word or word string that is an element of a second natural language sentence, and a word or word that is an element of a second natural language sentence of a common word pair obtained by extracting the same bilingual pair Associating a word string, a word or word string that is an element of a second natural language sentence, and a word that is an element of a second natural language sentence of a common word pair obtained by extracting the same parallel translation pair Alternatively, for a word string, a word or a word string that is an element of a first natural language sentence, and an element of a first natural language sentence of a common word pair obtained by extracting the same bilingual pair. Comprising the step of associating a word or word string.
The present invention also includes a step of using the common word pair obtained in the step of extracting the same bilingual pair, and repeating the method of learning the bilingual word pair until the common word pair no longer changes.
[0008]
FIG. 2 is a schematic configuration diagram of a bilingual word pair learning apparatus according to the present invention.
A bilingual word pair learning apparatus includes: a bilingual sentence reading unit 1 that reads a first natural language sentence and a second natural language sentence that is a parallel translation thereof; A word or a word string which is an element of the second natural language sentence is associated, and a bilingual word pair associated based on the first natural language is stored in the storage device 71. A bilingual word pair acquiring means for associating a word or word sequence which is an element of a first natural language sentence with a certain word or word sequence, and storing a bilingual word pair associated based on a second natural language in the storage device 72 6 and a common bilingual word pair extraction unit 8 that extracts a common bilingual word pair that is the same bilingual word pair from the storage devices 71 and 72 and stores the same in the storage device 9.
[0009]
Further, in order to use the word segmentation result and the syntax analysis result in the bilingual word pair acquisition unit 6 of the present invention, the morphological analysis unit 2 that morphologically analyzes the sentence read by the bilingual sentence reading unit 1 and then performs the syntax analysis. Includes syntax analysis means 3.
Further, in the bilingual word pair acquiring means 6 of the present invention, in order to use the common bilingual word pairs stored in the storage device 9 by the common bilingual word pair extracting means 8, language-specific word extraction for dividing the common bilingual word pairs by language is performed. Means A is included.
The present invention also includes means for repeating the learning of the bilingual word pair until the common bilingual word pair stored in the storage device 9 by the common bilingual word pair extracting means 8 does not change.
[0010]
(Action)
As described above, in the bilingual word pair learning method of the present invention, the first natural language sentence and the second natural language sentence that is a translation thereof are input to the bilingual sentence reading means 1 and the bilingual word pair acquiring means 6 is input. A word or word string that is an element of the second natural language sentence is associated with a word or word string that is an element of the first natural language sentence, and a bilingual word pair that is associated based on the first natural language is stored. The word or word string, which is an element of the first natural language sentence, is associated with the word or word string, which is an element of the second natural language sentence, by the bilingual word pair acquisition means 6, Is stored in the storage device 72, and the common bilingual word pair extraction means 8 extracts a common bilingual word pair that is the same bilingual word pair in the storage devices 71 and 72. , In the storage device 9. As a result, a stable bilingual word pair can be automatically extracted from the first natural language sentence and the second natural language sentence that is a translation thereof.
[0011]
The morphological analysis unit 2 performs morphological analysis of the sentence input by the bilingual sentence reading unit 1, stores the morphological analysis results in the storage devices 41 and 42, and further performs syntactic analysis by the syntactic analysis unit 3, and performs syntactic analysis. The results are stored in the storage devices 51 and 52, and the bilingual word pair acquiring means 6 uses the morphological analysis result and the syntax analysis result to store the bilingual word pair associated with the first natural language in the storage device. 71, and a bilingual word pair associated with the second natural language is stored in the storage device 72. The common bilingual word pair extraction means 8 uses the same word pair in the storage devices 71, 72. A certain bilingual word pair is extracted and stored in the storage device 9. This makes it possible to more accurately extract a common bilingual word pair.
[0012]
Further, the language-specific word extracting means A divides the common bilingual word pairs stored in the storage device 9 by the common bilingual word pair extracting means 8 for each language, and in addition to the bilingual sentence inputted by the bilingual sentence reading means 1, separates the bilingual sentence. The word or word string that is the element of the first natural language sentence and the word or word string that is the element of the second natural language sentence are used in the translated word pair acquisition means 6. As a result, the common bilingual word pair extracting means 8 can extract more common bilingual word pairs more accurately.
[0013]
Further, by means of repeating the learning of the bilingual word pair, the bilingual word pair is learned until the common bilingual word pair stored in the storage device 9 by the common bilingual word pair extracting means 8 does not change. Thereby, more common bilingual word pairs can be extracted more accurately.
Therefore, by executing all of the above methods and using the above means, a translated word pair is automatically obtained, and learning of the translated word pair becomes possible.
Although the above description describes a method of learning a bilingual word pair, the same applies to a learning device of a bilingual word pair and a learning program of a bilingual word pair.
[0014]
BEST MODE FOR CARRYING OUT THE INVENTION
An embodiment of the present invention will be described below with reference to the drawings.
FIG. 3 is a basic block diagram of a bilingual word pair learning apparatus according to an embodiment of the present invention. The bilingual word pair learning apparatus shown in FIG. 1 includes a bilingual sentence reading unit 1, a morphological analysis unit 2, a syntactic analysis unit 3, a bilingual word pair acquiring unit 6, a common bilingual word pair extracting unit 8, a language-specific word extracting unit A, It is composed of storage devices 41, 42, 51, 52, 71, 72 and 9.
[0015]
The first natural language sentence stored in the bilingual corpus and the second natural language sentence corresponding to the first natural language sentence are input (read) to the bilingual sentence reading unit 1.
The morphological analysis unit 2 stores the result of morphological analysis of the input first natural language sentence in the storage device 41, and stores the result of morphological analysis of the input second natural language sentence in the storage device 42.
The syntax analysis unit 3 uses the morphological analysis result stored in the storage device 41 to perform syntax analysis of the first natural language sentence, stores the result in the storage device 51, and stores the result in the storage device 42. The second natural language sentence is analyzed using the morphological analysis result, and the result is stored in the storage device 52.
[0016]
The translation word pair acquisition unit 6 includes a translation word pair acquisition unit 61 based on a first natural language and a translation word pair acquisition unit 62 based on a second natural language.
The bilingual word pair acquiring unit 61 based on the first natural language, for example, associates the syntax analysis result of the second natural language sentence with a word string that is a result of morphological analysis of the first natural language sentence, and performs the bilingual translation. A word pair is extracted and stored in the storage device 71.
The bilingual word pair acquisition unit 62 based on the second natural language, for example, associates the syntax analysis result of the first natural language sentence with a word string that is a result of morphological analysis of the second natural language sentence, and translates the bilingual word. The word pair is extracted and stored in the storage device 72.
The common bilingual word pair extraction unit 8 extracts the same bilingual word pair from the set of bilingual word pairs stored in the storage devices 71 and 72 and stores the same in the storage device 9.
The language-specific word extraction unit A divides a common bilingual word pair stored in the storage device 9 into words of a first natural language and a second natural language, and outputs a bilingual corpus read by the bilingual sentence reading unit 1. Add to parallel text.
[0017]
FIG. 4 is a flowchart for repeating learning according to one embodiment of the present invention.
Hereinafter, an example in which the first natural language is English and the second natural language is Japanese will be described based on this flowchart.
In step 101, a first natural language sentence and a second natural language sentence that is a translation thereof are input to the bilingual sentence reading unit 1. For example, as shown in FIG. 5, the input translation is “The house is somewhere about here”, “Look about”, and the bilingual sentence is “the house is somewhere around here”, “ Look around. "
[0018]
In step 102, the bilingual sentence read in step 101 is morphologically analyzed by the morphological analysis unit 2, and the results are stored in the storage devices 41 and 42. For example, as shown in FIG. 6, it is assumed that the words in the Japanese sentence are “word / house / ha / where / somewhere / this / around / to / is” and “round / to / watch / turn”. In addition, the morphological analysis unit 2 gives a part of speech to each word for syntax analysis. For example, as shown in FIG. 7, it is assumed that a verb is given to "Look", an adverb is given to "about", and a part of speech shown in FIG. 9, as shown in FIG. Also, for example, in Japanese text, as shown in FIG. 8, a noun is given to "hit", a particle is given to "wo", a verb is given to "see", a verb is given to "turn", and a part of speech shown in FIG. Suppose it was done.
In step 103, the syntax analysis unit 3 performs syntax analysis based on the morphological analysis results stored in the storage devices 41 and 42, and stores the results in the storage devices 51 and 52. For example, suppose that the results shown in FIGS. 7 and 9 are obtained for English sentences, and the results shown in FIGS.
[0019]
In Steps 104 and 105, a bilingual word pair based on the first natural language is obtained by using the morphological analysis result and the syntax analysis result stored in the storage devices 41, 42, 51, and 52. A translation word pair based on the first natural language is stored in the storage device 71 by the acquisition unit 61, and a translation word pair acquisition unit 62 based on the second natural language is based on the second natural language. The translated word pair is stored in the storage device 72.
For example, in the above example, as shown in FIG. 11, when the syntax analysis result of the Japanese sentence is associated with the English sentence subjected to the morphological analysis, "the" and "the", "house" and "house", etc. In the case where the results of parsing an English sentence are associated based on the Japanese sentence subjected to morphological analysis, it is assumed that “the” and “the”, “house” and “house” are associated with each other. . Note that a blank column indicates that there is no corresponding item. For example, "?" In the case where the result of syntactic analysis of the Japanese sentence based on the English sentence in FIG. 11 indicates that there is no corresponding English word or word string. means. Further, in this example, the word order of the syntax analysis result is changed and matched so as to match the syntax analysis result with the word string as much as possible without exceeding the intermediate node unit of the syntax analysis result. Here, the intermediate node of the parsing result indicates a grammatical category on the parsing result, such as a verb, a noun, a verb phrase, and a noun phrase. The term “intermediate node unit” refers to a word string belonging to a lower level than an intermediate node (that is, a grammatical category) in the syntax analysis result. For example, in the syntax analysis result shown in FIG. 8, the noun phrase unit is lower than the noun phrase. Since it is a word string belonging to, it refers to "hit" and "wo", and the verb phrase unit refers to "see" and "turn". The range not exceeding the intermediate node unit means that the range is limited to the range in the word string which is the intermediate node unit. For example, in the syntax analysis result shown in FIG. 8, "hit" and "wo" in the range of the noun phrase unit are exchanged, and "hit / wo" of the whole noun phrase unit and "look / turn" of the whole verb / phrase unit are displayed. Can be replaced, but the particle "wo" is a word in the range of noun phrases and the verb "mi" is a word in the range of verb phrases. Cannot be interchanged with the verb "mi". Therefore, when associating the sentence analysis result of the Japanese sentence “per / wo / look / turn” with the English sentence “look about”, as shown in FIG. 11 (upper figure), the noun phrase “per / wo” and the verb phrase “see By changing the word order of “/ turn”, “look” and “look” and “hit” and “about” are associated with each other. These bilingual word pairs are stored in the storage devices 71 and 72.
[0020]
In step 106, the common bilingual word pair extraction unit 8 extracts the same bilingual word pair from the bilingual word pair stored in the storage device 71 and the bilingual word pair stored in the storage device 72. It is determined whether or not all are the same as the common bilingual word pairs stored in. If all are the same (yes), if there is not the same (no), and if yes is selected, the learning of the bilingual word pair ends. , No, the process proceeds to the next step 107. In the above example, common bilingual word pairs extracted by the common bilingual word pair extraction unit 8 from the input sentence shown in FIG. 5 are, for example, “that” and “the”, “house” and “house” as shown in FIG. "house" is extracted. Since nothing has been stored in the storage device 9, no is selected, and the process proceeds to the next step 107.
[0021]
In step 107, a bilingual word pair that does not match the common bilingual word pair stored in the storage device 9 in step 106 is stored in the storage device 9. In the case of the above example, all the translated word pairs extracted in step 106 shown in FIG.
In step 108, the common bilingual word pairs stored in the storage device 9 are classified for each language and added to the input bilingual translation. That is, the common bilingual word pair is stored in the bilingual corpus. In the case of the above example, “the”, “house”, “is”, and “about” are included in the English sentence divided according to the language, and “the” and “house” are included in the Japanese sentence as a translation thereof. ”,“ Wa ”, and“ per ”are added.
[0022]
In the above example, the process proceeds to step 101 again, so that the above example will be further described step by step.
In step 101, both the original bilingual sentence and the bilingual word pair added in step 108 are input to the bilingual sentence reading unit 1.
In step 102, the morphological analysis unit 2 gives the part of speech of the bilingual word pair input in step 101, and stores the results in the storage devices 41 and 42.
In step 103, the parsing unit 3 performs a syntactic analysis on the bilingual word pair input in step 101 using the morphological analysis results stored in the storage devices 41 and 42, and stores the results in the storage devices 51 and 52. .
[0023]
In steps 104 and 105, the first sentence and the morpheme result of the word or word string of the common word pair stored in the storage devices 41, 42, 51, and 52 are used to perform the first analysis. A bilingual word pair based on the first natural language is stored in the storage device 71 by the bilingual word pair obtaining unit 61 based on the natural language, and a bilingual word pair obtaining unit 62 based on the second natural language. Thus, a bilingual word pair based on the second natural language is stored in the storage device 72. In the case of the above example, the common bilingual word pair “hit” and “about” are added to the input sentence (see FIG. 12), so that when the syntax analysis result corresponds to the word string, “hit” and “about” are It is taken into account that it will correspond. For example, when the parsing result of the Japanese sentence “that / house / is / where / ka / this / per // is” is combined with the English sentence “The house is somewhere about here”, the verb is obtained from the parsing result shown in FIG. The phrase “a” is replaced by the adverb phrase “this / periphery / ni”. In the adverb phrase, the adverb “kon” is replaced by the particle “ni”, and the word order is as shown in FIG. As a result, for example, correspondence with English sentences is as shown in FIG. 15, and "this" and "here" are associated with each other. On the other hand, when the parsing result of the English sentence is matched with the Japanese sentence for the same example sentence, the adverb “about” and the adverb “here” in the adverb phrase are switched from the parsing result shown in FIG. 9, as shown in FIG. Word order. As a result, for example, the correspondence with the Japanese text is as shown in FIG. 15, and "here" and "this" are associated with each other. These translated word pairs shown in FIG. 15 are stored in the storage devices 71 and 72.
[0024]
In step 106, when the common bilingual word pair extraction unit 8 extracts the same bilingual word pair from the bilingual word pair stored in the storage device 71 and the bilingual word pair stored in the storage device 72, A common bilingual word pair of "this" and "here" is newly extracted. This does not match the common bilingual word pair of FIG. 12 currently stored in the storage device 9. Therefore, the determination in this step is no and the process proceeds to step 107.
[0025]
In step 107, “this” and “here” that do not match FIG. 12 stored in the storage device 9 are added, and the common bilingual word pairs shown in FIG. 16 are stored in the storage device 9.
In step 108, the common bilingual word pairs stored in the storage device 9 are classified for each language and added to the input bilingual translation.
Thereafter, the process further proceeds to step 101, but when the common bilingual word pair no longer changes, the determination in step 106 becomes yes, and the learning of the bilingual word pair ends.
Although the above example has been described based on the configuration diagram shown in FIG. 3, various changes and applications are possible without being limited to this example within the scope of the claims.
[0026]
The bilingual word pair learning apparatus of the present invention can be composed of a computer having a CPU and a memory, a user terminal, and a machine-readable recording medium such as a CD-ROM, a magnetic disk device, and a semiconductor memory.
The learning program of the translation word pair recorded on the recording medium, or the learning program of the translation word pair transmitted via the communication line is read by the computer, controls the operation of the computer, and the above-described components on the computer and Implement each process.
[0027]
【The invention's effect】
As described above, according to the present invention, a bilingual word in which a first natural language sentence is associated with a second natural language sentence from a first natural language sentence and a bilingual second natural language sentence By extracting a common translated word pair by comparing a pair and a translated word pair in which the second natural language sentence is associated with the first natural language sentence, the automatically translated first natural language and the second translated natural language sentence are automatically extracted. A bilingual word pair of natural language 2 can be extracted.
Further, in the associating step, a translated word pair of the first natural language and the second natural language can be more accurately extracted by using a result obtained by using the morphological analysis means or the syntax analysis means.
In addition, by adding the bilingual word pair once extracted to the input bilingual sentence, more bilingual word pairs of the first natural language and the second natural language can be extracted.
Further, by repeating the learning of the translated word pairs until the extracted translated word pairs no longer change, more and more accurately the translated word pairs of the first natural language and the second natural language can be extracted.
The bilingual word pairs extracted in this way can be used, for example, for constructing or expanding an electronic bilingual dictionary, or can be used as a bilingual dictionary for a machine translation system.
[Brief description of the drawings]
FIG. 1 is a schematic flowchart of a method of learning a translated word pair for explaining the principle of the present invention.
FIG. 2 is a schematic configuration diagram of a bilingual word pair learning apparatus of the present invention.
FIG. 3 is a basic block configuration diagram of a bilingual word pair learning apparatus according to an embodiment of the present invention.
FIG. 4 is a flowchart illustrating repetition of learning according to an embodiment of the present invention.
FIG. 5 is a diagram showing an example of an input bilingual sentence according to an embodiment of the present invention.
FIG. 6 is a diagram showing an example of a morphological analysis result (word cut) of a Japanese sentence according to an embodiment of the present invention.
FIG. 7 is a view showing an example (No. 1) of a result of parsing an English sentence according to an embodiment of the present invention.
FIG. 8 is a view showing an example (part 1) of a result of parsing Japanese sentences according to an embodiment of the present invention.
FIG. 9 is a view showing an example (part 2) of a result of parsing an English sentence according to an embodiment of the present invention.
FIG. 10 is a view showing an example (part 2) of the result of parsing Japanese sentences according to an embodiment of the present invention.
FIG. 11 is a diagram showing an example (part 1) of a word correspondence result according to an embodiment of the present invention.
FIG. 12 is a diagram showing an example (part 1) of common word correspondence according to an embodiment of the present invention.
FIG. 13 is a view showing an example in which the order of words in the results of parsing Japanese sentences according to an embodiment of the present invention is changed.
FIG. 14 is a diagram illustrating an example in which the word order of the result of parsing an English sentence according to an embodiment of the present invention is changed.
FIG. 15 is a view showing an example (part 2) of a word correspondence result according to an embodiment of the present invention.
FIG. 16 is a view showing an example (part 2) of common word correspondence according to an embodiment of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Translation reading part, 2 ... Morphological analysis part, 3 ... Syntax analysis part, 6 ... Translation word pair acquisition part, 8 ... Common translation word pair extraction part, 9 ... Common bilingual word pair, 71... Bilingual word pair storage device associated based on first natural language, 72... Bilingual word pair storage device associated based on second natural language, A.・ Language-specific word extraction unit

Claims (12)

A bilingual word pair learning method for extracting a bilingual word pair by inputting a first natural language sentence and a second natural language sentence that is a translation thereof,
A pair of a word or a word string having a translational relationship obtained by associating a word or a word string that is an element of the first natural language sentence with a word or a word string that is an element of the second natural language sentence is generated. Procedure 1 for storing
Generating a pair of a word or a word string having a bilingual relation obtained by associating a word or a word string as an element of the second natural language sentence with a word or a word string as an element of the first natural language sentence And memorizing step 2
A third step of comparing the pair of words or word strings in the associated bilingual relations generated in step 1 and step 2 to extract and store a common word pair that is the same word or pair of word strings; A method for learning a bilingual word pair, comprising: The method for learning a bilingual word pair according to claim 1,
Procedures 1 and 2 are bilingual word pair learning methods characterized by associating a syntax tree obtained by parsing an input natural language sentence with a word or word string. The method for learning a bilingual word pair according to claim 1 or 2,
In step 1, the word or word string that is an element of the first natural language sentence and the word or word string that is an element of the first natural language sentence of the common word pair extracted in step 3 are A word or word string that is an element of the second natural language sentence and a bilingual relationship obtained by associating a word or word string that is an element of the second natural language sentence of the common word pair extracted in step 3 A procedure for generating a pair of a certain word or word string,
In step 2, the words or word strings that are the elements of the second natural language sentence and the words or word strings that are the elements of the second natural language sentence of the common word pair extracted in step 3 are A word or word string that is an element of the first natural language sentence and a bilingual relationship obtained by associating the word or word string that is an element of the first natural language sentence of the common word pair extracted in step 3 A method for learning a bilingual word pair, comprising a step of generating a pair of a certain word or word string. The learning method of a bilingual word pair according to claim 3,
A learning method of a bilingual word pair, comprising a step of repeating learning until the common word pair stored in step 3 no longer changes. A bilingual word pair learning apparatus for inputting a first natural language sentence and a second natural language sentence which is a translation thereof and extracting a word pair having a bilingual relationship,
A pair of a word or a word string having a translational relationship obtained by associating a word or a word string that is an element of the first natural language sentence with a word or a word string that is an element of the second natural language sentence is generated. First paired translation word pair acquisition means for storing
Generating a pair of a word or a word string having a bilingual relation obtained by associating a word or a word string as an element of the second natural language sentence with a word or a word string as an element of the first natural language sentence Second translation word pair acquisition means for storing
A common word pair that is the same word or pair of word strings is extracted by comparing the pair of pairs of words or word strings having an associated bilingual relation generated by the first and second parallel word pair acquisition means. A bilingual word pair learning device, comprising: The learning device for a bilingual word pair according to claim 5,
The first and second parallel word pair acquiring means associates a syntax tree obtained by parsing an input natural language sentence with a word or a word sequence, and is a bilingual word pair learning device. 7. The learning device for bilingual word pairs according to claim 5,
The first bilingual word pair acquiring means includes a word or a word string which is an element of the first natural language sentence, and an element of a first natural language sentence of a common word pair extracted by the common bilingual word pair extracting means. , The word or word string that is an element of the second natural language sentence, and the element of the second natural language sentence of the common word pair extracted by the common parallel word pair extraction unit Means for generating a pair of words or word strings in a bilingual relationship obtained by associating a word or word string that is,
The second bilingual word pair acquiring means includes a word or word string as an element of the second natural language sentence, and an element of a second natural language sentence of the common word pair extracted by the common bilingual word pair extracting means. , The word or word string that is an element of the first natural language sentence, and the element of the first natural language sentence of the common word pair extracted by the common parallel word pair extraction unit A learning device for bilingual word pairs, comprising means for generating a pair of words or word sequences having a bilingual relation obtained by associating a word or word sequence. The learning device for a bilingual word pair according to claim 7,
A bilingual word pair learning device, comprising: means for repeating learning until the stored common word pair extracted by the common bilingual word pair extracting means does not change. A recording medium storing a bilingual word pair learning program for inputting a first natural language sentence and a bilingual second natural language sentence and extracting a bilingual word pair,
A pair of a word or a word string having a translational relationship obtained by associating a word or a word string that is an element of the first natural language sentence with a word or a word string that is an element of the second natural language sentence is generated. Processing 1 for storing
Generating a pair of a word or a word string having a bilingual relation obtained by associating a word or a word string as an element of the second natural language sentence with a word or a word string as an element of the first natural language sentence Processing 2 for storing
A process 3 for comparing the pair of words or word strings having the corresponding translations generated in the processes 1 and 2 to extract and store a common word pair that is a pair of the same word or word sequence; , A recording medium storing a program for learning a pair of translated words, which causes a computer to execute the process. A recording medium recording a learning program for a translated word pair according to claim 9,
Processing media 1 and 2 are a recording medium storing a bilingual word pair learning program having a process of associating a word or word string with a syntax tree obtained by parsing an input natural language sentence. A recording medium recording a learning program for a translated word pair according to claim 9 or 10,
Processing 1 is performed on a word or word string that is an element of the first natural language sentence and a word or word string that is an element of the first natural language sentence of the common word pair extracted in processing 3 A bilingual translation obtained by associating a word or word string that is an element of the second natural language sentence with a word or word string that is an element of the second natural language sentence of the common word pair extracted in process 3 Having a process of generating a pair of related words or word strings,
Processing 2 is performed on a word or word string that is an element of the second natural language sentence and a word or word string that is an element of the second natural language sentence of the common word pair extracted in processing 3 A bilingual translation obtained by associating a word or word string that is an element of the first natural language sentence with a word or word string that is an element of the first natural language sentence of the common word pair extracted in process 3 A recording medium recording a learning program for a translated word pair having a process of generating a pair of related words or word strings. A recording medium storing a learning program for a translated word pair according to claim 11,
A recording medium in which a learning program for a bilingual word pair having a process of repeating learning until the common word pair extracted and stored in process 3 no longer changes is recorded.

Images