JP2003114890A - Translation device, translation method, translation server, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a user with a translation device capable of easily judging the appropriateness of the results of translation of a first language into a second language. SOLUTION: The first language is taken in, the taken-in first language is translated into a second language, and translated sentences are further translated into the first language. Whether the translated words in the translation translated into the first language are present or not in the taken-in first language is judged. When the translated words are found not to be present by this judgement, the words identical to those in the inputted first language are retrieved from the translated words of the second language corresponding to the translated words, the present words are replaced with the retrieved ones to generate translated sentences, and the generated translation is outputted.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a first language to a second language.
A translation device for translating the translated second language into a first language, in particular, a translation that replaces a word used in the original first language as a translated word and outputs the translated word. The present invention relates to a device, a translation method, a translation server and a program.

[0002]

2. Description of the Related Art Conventionally, as disclosed in Japanese Patent Application Laid-Open No. 4-353973, an input means and a display means are provided, and the first means is further provided.
First dictionary means for storing a plurality of entry words of the language and translations of the second language corresponding to the entry words, and a plurality of entry words of the second language and translations of the first language corresponding to the entry words Second dictionary means, first control means for searching the first dictionary means for a translated word of the second language corresponding to the entry word of the first language corresponding to the character input from the input means, and the first control means. When an instruction for reverse conversion is input from the input means and one of the translated words of the second language displayed on the display means is selected by the first control means, the selected second language is set as a headword. Secondly, the second dictionary means is searched for the translated word of the first language corresponding to the entry word and the second word is displayed on the display means.
There was an electronic dictionary device with control means. As a result, when one of the translated words is selected and the reverse conversion is instructed,
The native language for the entry word is displayed.

In a machine translation system that bidirectionally translates a first language and a second language, if the user has no knowledge of the second language, is the result of translating the first language into the second language appropriate? I could not judge whether or not. Therefore, in order to judge the validity of the second language obtained as the translation result, a method of retranslating the second language as the translation result into the first language and presenting the translation result is used.

[0004]

However, in the first prior art, although it can be applied to an electronic dictionary that performs inverse conversion of only words, it is difficult to apply it to a translation device that translates a sentence. there were. On the other hand, in the second conventional translation system, when the second language translated from the first language is translated again to obtain the first language, this includes a lot of ambiguity. Is the first candidate (correct answer)
I was trying. In addition, even if a plurality of translation results are obtained by translating the second language and there is a translated sentence close to the original first language after the second candidate, it may not be presented as the translation result. Furthermore, even if it is a correct interpretation, there is a problem that the translation result is different because the term used in the original first language is not taken into consideration, and it cannot be determined as the correct second language.

The present invention has been made in view of the above circumstances, and when the result of translating the first language into the second language is further translated into the first language, by utilizing the original first language,
An object of the present invention is to provide a translation device, a translation method, a translation server, and a program that present a result closest to the original first language and facilitate the determination of the validity of the second language.

[0006]

In order to solve the above-mentioned problems, the present invention provides an input means for inputting a first language,
First translation processing means for translating the input first language into a second language, second translation processing means for further translating the translated sentence translated by the first translation processing means into the first language, and And referred to when the second translation processing means performs translation processing,
It is determined whether or not the translation dictionary unit including at least the translation rule and the dictionary and the translated word in the translated sentence translated by the second translation processing unit are present in the input first language, and are not present. In this case, a translation word determining unit that searches the translation word of the corresponding second language word in the second language for the same word as the input word in the first language, and generates a replacement translation sentence, and this generation And an output unit for outputting the translated text.

Therefore, according to the present invention, when the first language is input, it is translated into the second language, which is further translated into the first language, and the translation of this translation is used in the original first language. It is possible to present a result that is the closest to the original first language by replacing the word with the existing word, and it is possible to easily judge the validity of the second language. (2) In order to solve the above problem, the present invention is a translation method for translating an input first language into a second language, the method including a step of capturing the first language and a second method of capturing the first language. A step of translating the translated sentence into a language, a step of further translating the translated sentence into a first language, and whether or not a translated word in the translated sentence further translated into the first language exists in the imported first language. If the translated word does not exist in the determining step and the determination, the same word as in the input first language is searched from the translated words of the word in the second language corresponding to the translated word, The translation method is characterized by including a step of generating a replacement translated sentence and a step of outputting the generated translated sentence.

Therefore, according to the present invention, when the first language is input, it is translated into the second language, which is further translated into the first language, and the translation of this translation is used in the original first language. It is possible to present a result that is the closest to the original first language by replacing the word with the existing word, and it is possible to easily judge the validity of the second language. (3) In order to solve the above-mentioned problem, the present invention is a program for translating an input first language into a second language, which has a function of loading the first language and a second function of loading the first language. A function of translating into a language, a function of further translating the translated text into a first language, and whether or not a translated word in the translation translated into the first language exists in the imported first language. If the translated word does not exist by the function for determining whether or not the translated word corresponds to the second
The computer is made to realize the function of searching for the same word as the input word in the first language from the translated words of the language, generating a replacement translated sentence, and the function of outputting the generated translated sentence. This is a translation processing program.

Therefore, according to the present invention, by recording the program as described above, the original first program can be obtained in the same manner as in the above (2).
The result closest to the language can be presented, and the validity of the second language can be easily determined. (4) If the translation server is loaded with the recording medium recording the program described in (3) above, when a translation request is made by the client, the translated sentence translated according to the program is sent to the requesting client. Can be provided. Further, the translation server, when loaded with the recording medium recording the program described in (3) above, receives the request for providing the program recorded in the recording medium from the client, and the requesting client receives the request. The program can be downloaded.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. (First Embodiment) FIG. 1 is a block diagram showing an embodiment of a translation apparatus according to the present invention. This translation device is a device that translates and outputs a first language into a second language.
This translation device includes a first language input means 1 for inputting character data, file data, voice, file data from a network, etc. as a first language, and a first language input from the first language input means 1 (for example, A storage medium 3 in which a program for translating Japanese) into a second language (for example, English) is stored, and a translation dictionary unit 4 for storing grammar / dictionary, etc.
And a data storage unit 5 for storing various data and a read program, and a first language (or a second language) translated into a second language (or a first language) according to a program stored in the data storage unit 5. It is composed of a translation processing main body 2 composed of a CPU and a translation result output means 6 for outputting a translation result.

The first language is input from the first language input means 1. As the input information, the information input by the user from the keyboard 11, the information read and input from the file 12 stored in advance through the application software, and the voice signal input from the voice input device 13 or the CD-ROM 14 are voiced. Information that is input by converting it into digital information that can be processed by the translation processing main unit 2 through the recognition device 15 or information that is captured via a network 16 such as a LAN or the Internet can be considered.

The translation processing unit 2 uses the grammar / dictionary stored in the translation dictionary unit 4 for the first language (for example, Japanese) input by the first language input means 1 to store data in the data storage unit. The translation processing program stored in 5 performs morphological analysis / syntactic analysis, etc., and receives the first translation processing means 21 for translating into a second language (for example, English) and the second language as the translation result, Using the grammar / dictionary stored in the translation dictionary unit 4, the translation processing program stored in the data storage unit 5 performs morphological analysis / syntactic analysis, etc., and translates into the first language Second translation Processing means 22
Then, it is determined whether or not each translated word of the first language obtained by the second translation processing means 22 exists in the first language input by the first language input means 1. Searches for the same word in the input first language from the translations of the word in the second language corresponding to the translation,
The translated word determining unit 23 that replaces this and generates the first language
And comparing the first language generated by the translation determination means 23 with the first language input by the first language input means,
Similarity determination means 2 which makes the most similar sentence the first candidate
4 and.

The data storage unit 5 includes the first language input means 1
The data input from, and various data processed by the translation processing main unit 2 are stored. The translation result output means 6 outputs the translation result output from the translation processing main unit 2. As an output form, a file 61 for storing the translation result, a display device 62 such as a CRT / liquid crystal for display output, a voice output device 63 for voice output, or a network 64 may be considered. A magnetic disk, for example, is used as the storage medium 3, but other than that, for example, a magnetic tape or a CD.
-ROM, DVD-ROM, FD, MO, MD, memory card and the like may be used.

Next, the operation of the above apparatus will be described with reference to FIG. When the operation of the apparatus is started, the translation processing main unit 2 reads the translation processing program from the storage medium 3 and stores it in the data storage unit 5, and the preparation for translation processing is started. In this state, when the first language "J" (for example, Japanese) is input to the first language input means 1, the translation processing main body section 2 takes in the data input from the first language input means 1 (S1). , First language import function).

The first translation processing means 21 uses the grammar / dictionary etc. stored in the translation dictionary unit 4 for the first language “J” and stores the first language “J” (in the data storage unit 5). For example, for Japanese data, the translation processing program stored in the data storage unit 5 performs morphological analysis / syntactic analysis, etc., executes the first translation processing, and executes the second language "E".
(S2, first language translation function). The second language “E” is stored in the data storage unit 5. In the second translation processing means 22, the second language “E” is further analyzed using the grammar / dictionary stored in the translation dictionary unit 4 by the translation processing program stored in the data storage unit 5.
A syntactic analysis and the like are performed, and a second translation process is executed to obtain a plurality of first languages "Jn" (S3, second language translation function).

Here, for example, a word having a plurality of parts of speech in the same headword such as "buy" (verb: buy, etc, noun: shopping, etc) or "find" (find, find) , Etc) past forms of "found" and "fo
If there are words with the same heading but different meanings, such as "und" (established, etc.), each translation will be translated unless there is a grammatical error, and multiple translation results will be obtained. .

First, the initial value of "i" is set to "0" (S
4), the i-th first language is set to "Ji" (S5),
Each translated word of "Ji" is replaced with that used in the first language "J" (S6, translated word replacement function). And the first
It is judged whether or not the language "Ji" is the final translation result (there is no subsequent translation result) (S7). If it is not the final translation result (if there is a translation result next), "i + 1" is set to the next It is set to "i" (S8), and the process returns to S5. If the first language "Ji" is the last translation result (there is no subsequent translation result), "Ji"
The result closest to "J" in "i" is set as "R" (S9,
(Similarity determination function), "R" is output to the translation result output means 6 (S10), and the processing ends.

Here, the similarity determination means 24 determines the similarity based on the translated sentence regenerated by the replacement and the first language "J". As a similarity determination method, for example, the translated sentences of the first languages “J” and “Jn” are subjected to morphological analysis and syntactic analysis, respectively, and respective syntactic trees are created, and the degree of coincidence of each node and the word in each node One method is to make a comparison. This determines the most similar translation result and outputs it to the translation result output means 6 as the first candidate. The output translation result may be one or more.

Next, the translated word replacement function (S6) in FIG.
The operation will be described with reference to FIG. First, the initial value of "j" is set to "0" for the translated word "Ji" which is the i-th result of the plurality of first languages "Jn" obtained in S3 of FIG. 2 (S20). Next, "j" in "Ji"
The second translated word is set to "Fj" (S21), and it is determined whether or not "Fj" is present in "J" (S22, translated word existence determination function). Here, when "Fj" exists in "J", it skips to S30. "Fj"
However, if it does not exist in "J", another translation word candidate "Tn" of the entry word having the same word number as "Fj" is obtained (S2).
3).

Here, the initial value of "k" is set to "0" (S
24), the kth translation word candidate is set to "Tk" (S2)
5). Next, it is determined whether "Tk" is present in "J", and if not, it is determined whether "Tk" is the last translation word candidate (S27). As a result, if “Tk” is not the last translation candidate, “k + 1” is replaced with the next “k”.
(S28), and the process returns to S25. When "Tk" is the last translation word candidate, the process skips to S30. On the other hand, "T
When "k" is present in "J", "Fj" is changed to "T".
k ”(S29).

Then, it is judged whether or not "Fj" is the last translated word (S30), and if it is not the last translated word, "j + 1" is set to "j" (S31), and the process returns to S21. If "Fj" is the last translated word, the process ends. Next, the above-mentioned processing contents will be described with an example. When the device starts working,
The translation processing main unit 2 reads the translation processing program from the storage medium 3, stores it in the data storage unit 5, and then prepares for translation processing. In this state, as the first language "J", "I saw him through the telescope." Is the first language input means 1
Is input to the translation processing main unit 2, the data input from the first language input means 1 is loaded (S1, first language loading function).

The first translation processing means 21 uses the grammar / dictionary stored in the translation dictionary section 4 for the first language "I saw him through the telescope" stored in the data storage section 5, The translation processing program stored in the data storage unit 5 performs morphological analysis / syntactic analysis, etc., executes the first translation processing based on the one shown in FIG. 4, and outputs the second language “E” as “E”. I saw him with telescope. "(S2,
First language translation function). The second language “E” is stored in the data storage unit 5. Here, FIG. 4 shows the correspondence between the word numbers in Japanese (original sentence) that is the first language and the word numbers in the second language that is the translated sentence.

Next, in the second translation processing means 22, the second translation processing means 22
The language "I saw him with telescope." Is further used for morphological analysis, syntactic analysis, etc. by the translation processing program stored in the data storage unit 5 using the grammar and dictionary stored in the translation dictionary unit 4, A second translation process is executed based on what is shown in FIG. 5, and as a plurality of first languages “Jn” shown in FIG. 6, “J0: I saw him holding a telescope. , "J1: I saw him through a telescope." (S3, second language translation function). Here, FIG. 6 shows the correspondence relationship of the translation from the second language to the first language, in which the word number of English as the second language and the word number of Japanese as its translation correspond to each other. There is.

Next, in "Jn", the initial value of "i" is set to "0" (S4), the i-th first language is set to "Ji" (S5), and each translated word of "Ji" is set to the first. It is replaced with the one used in the language "J" (S6). In other words, the 0th first language is "J0: I saw him with a telescope.", And the translated words "J0: I saw him with a telescope." Replace with the one used in "J: I saw him with a telescope." The replacement is as follows.

First, the initial value of "j" is set to "0" (S
20). Next, if the "j" th translation in "J0: I saw him holding a telescope" is "Fj", then "j
"= 0" th translated word becomes "F0" (S21), and "F"
It is determined whether "0" exists in "J" (S2).
2). Here, FIG. 6A and FIG. 4 are compared. Since "F0" in (a) of FIG. 6 is "I am" and exists in "J: I saw him with a telescope."
Skip to 30.

Since "F0" is not the last (there is the next translation), "j + 1" is replaced by the next "j", that is, "j =
1 ”(S31), and the process returns to S21. Similarly, "F1"
Is a "telescope" and is present in "J: I saw him with a telescope.", So skip to S30. "F
"0" is not the last (there is the next translation), so "j +"
1 ”as the next“ j ”, that is,“ j = 2 ”(S3
1) and returns to S21. If you do the same, "F2"
"Has" and does not exist in "J: I saw him through a telescope." Therefore, another translation word candidate "Tn" is obtained with the headword "with" having the same word number 4 as "has F2:" (S23).

The initial value of "k" is set to "0" (S24),
When the kth translation word candidate is set to “Tk” (S25), FIG.
As shown in, each translation of the headword "with""hasT0:","withT1:","T2:to","T"
3: Simultaneously ”and“ T4: de ”. Here, "T
0: ”, but it is present in“ J: I saw him through a telescope. ”(S26). Since it does not exist in "J" here, it is confirmed whether or not it is the last translated word (S27), and the process returns to S25.

Similarly, "with T1:", "T2:
, "And" T3: at the same time ", it is judged whether or not they exist in" J: I saw him through a telescope. " It is confirmed (S27), the process is performed for each of "T1", "T2", and "T3", and it is not the last (the next translation is)
Therefore, "k + 1" is set as the next "k" (S28), and S25
Return to. Similarly, when it is determined whether or not "T4: at" is present in "J: I saw him through a telescope.", This is present at "J". Therefore, "have F2:" is replaced with "T4: with" (S29), and it is determined whether "F2" is the last translation (S30).
Since it is not the last translated word, "j + 1" is set as the next "j" (S31), and the process returns to S21.

Here, when it is determined that a translated word exists in "J: I saw him with a telescope", the translated word is replaced with the translated word. If the translation word of the headword “with” as shown in FIG. 7 does not exist, the original translation word is used as it is, and the translation sentence “Jn” is regenerated. In the same manner as described above, the processes for "F3: see him" and "F4: see" are performed, and the process of FIG.
I saw him through a telescope. The result is obtained.

Furthermore, in S5 of FIG. 2, "J
1: I looked at him through a telescope. Is processed in the same manner as above, and the result "J1: I saw him through a telescope." Is obtained. Then, in response to the generated translation "J0: I saw him through a telescope" and "J1: I saw him through a telescope.", "J: I saw him through a telescope." To get the result "R" closest to. This “R” judgment is the result of morphological analysis / syntactic analysis of “J: I saw him with a telescope” shown in FIG. 8 and “J0: I have seen him with a telescope” shown in FIG. The similarity is determined by the similarity determination means 24 based on the result of the morphological analysis and the syntactic analysis of "I saw him with a telescope."

As the similarity determination method, for example, morphological analysis and syntactic analysis are performed on the translated sentences of the first language “J = I saw him with a telescope” and FIG. There is a method of creating a syntax tree as shown in 9 and comparing the degree of coincidence of each node and the words in the node. This determines the most similar translation result and outputs it to the translation result output means 6 as the first candidate. The output translation result may be one or more.

As a result, "R = I saw him through the telescope."
The result is obtained (S9), and "R" is output to the translation result output means 6 (S10). Therefore, according to the above-described embodiment, when the first language is input, it is used in the input first language when translating into the second language and further translating the result into the first language. By substituting the existing phrase, a sentence closer to the input first language can be generated, the correctness of the second language can be improved, and a sufficient translation result can be obtained.

By determining the degree of similarity, sentences after the second candidate can be output as the first candidate among the re-translated candidates of the first language. (Second Embodiment) The second embodiment will be described below. The description of the same parts as those in the first embodiment will be omitted. The second embodiment shows processing when a plurality of input translated sentences (second languages) for the first language are present. The operation of the device will be described with reference to FIG. When the operation of the apparatus is started, the translation processing main unit 2 reads the translation processing program from the storage medium 3 and stores it in the data storage unit 5 to start preparation for translation.

In this state, when the first language "J" (for example, Japanese) is input to the first language input means 1, the translation processing main unit 2 changes the data input from the first language input means 1. Import (S51, first language import function). The first translation processing means 21 uses the grammar / dictionary stored in the translation dictionary unit 4 for the first language “J”, and the data storage unit 5
The first language “J” (for example, Japanese) data stored in is subjected to morphological analysis / syntactic analysis, etc. by the translation processing program stored in the data storage unit 5 to perform the first translation processing. Execute to obtain the second language "En" (S52,
First language translation function). The plurality of second languages "En"
It is stored in the data storage unit 5.

First, the initial value of "m" is set to "0" (S5
3) Let the second language, which is the m-th translation result, be "Em". In the second translation processing means 22, the second language "Em"
Further, using the grammar / dictionary stored in the translation dictionary unit 4, the translation processing program stored in the data storage unit 5 performs morphological analysis / syntactic analysis, etc., and executes the second translation process. Obtain a plurality of first languages "Jmn" (S5
5, second language translation function). Next, the first language “Jmn” that is the i-th translation result is set to “Ji” (S56), and “J
Each translated word of "i" is replaced with that used in the first language "J" (S58, translated word replacement function).

Then, it is judged whether or not the first language "Ji" is the final translation result (there is no subsequent translation result) (S59), and if not the last translation result (if there is a translation result next). , "I + 1" as the next "i" (S60), S57
Return to. On the other hand, if the first language “Ji” is the final translation result (there is no subsequent translation result), the second language “E”
It is determined whether or not "m" is the final translation result (there is no subsequent translation result) (S61). If it is not the final translation result (if there is a next translation result), "m + 1" is replaced with the next "m". ] (S
62), and returns to S54.

If the second language "Em" is the final translation result (there is no subsequent translation result), the result closest to "J" among "Ji" is set to "R" (S63, similar). Output function, and outputs "R" to the translation result output means 6 (S6).
4) and the process ends. Next, the operation of the translated word replacement function (S58) in FIG. 10 is similar to that in FIG. Therefore, according to the above-described embodiment, when the first language is input, it is used in the input first language when translating into the second language and further translating the result into the first language. It is possible to generate a sentence closer to the input first language by substituting the phrase
The correctness of the language can be improved and sufficient translation results can be obtained.

Even if a plurality of second languages are output as the translation result, the correctness of the second language can be improved and a sufficient translation result can be obtained. (Third Embodiment) FIG. 11 is a block diagram showing an embodiment of a translation server and an information transmission medium according to the present invention. In this embodiment, the server side is shown in FIGS.
Alternatively, the translation processing program that is the application software shown in FIG. 10 is owned, and the client side installs the server-owned translation processing program under the request of service provision, thereby enabling the translation processing by itself, or It is to realize the function improvement of the translation processing.

Server 71 and a plurality of clients 7
When the routers 8 and 79 use a router 76 or a public line, they are connected to a network 74 via a modem 73, respectively. The server 71 is shown in FIG. 2, FIG. 3 and / or FIG.
The database 70 storing the translation processing program shown in 0 and the network 74 using the program data as the information transmission medium 75 under the communication protocol according to the standard of the OSI reference model when receiving the request of the translation processing program from each client side. Control unit 7 for transmitting to
11 is provided.

The information transmission medium 75 is an information medium in which the translation processing program is packetized according to a predetermined communication protocol and transmitted to the client side. On the other hand, each of the clients 78 and 79 is a general PC (personal computer), a mobile terminal or the like, and is packetized along with the translation processing program request information under the destination address and the request source address to the network 74. A transmission control unit 771 for transmission and a translation unit 772 capable of performing the translation processing shown in FIG. 1 by installing the translation processing program provided from the server under a program request are provided.

Therefore, according to such an embodiment,
When the server side is equipped with a translation processing program as one of the service provisions, each client 78, 79 installs the translation processing program from the server so that each client functions as a translation device and the input first It is possible to translate one language. Also,
The client side may request the server side for translation processing and send the translation processing result to the client. In addition, the present invention can be modified in various ways without departing from the scope of the invention.

[0042]

As described above, according to the present invention, when the first language is input, it is translated into the second language, and when the result is translated into the first language, the input first language is input. By substituting the phrase used in, it is possible to generate a sentence closer to the input first language, improve the correctness of the second language, and obtain a sufficient translation result. . Further, according to the present invention, since the server has a translation processing program for providing, if the client installs the translation processing program owned by the server when needed, it becomes possible to provide the function of the translation device. A server and an information transmission medium can be provided.

Further, the present invention can provide a translation server and an information transmission medium which enable a server to perform a translation process and to transmit a processing result to the client upon request from the client.

[Brief description of drawings]

FIG. 1 is a diagram illustrating an embodiment of a translation apparatus according to the present invention.

2 is a flowchart illustrating an example of a processing operation of the translation device shown in FIG. 1 and a program recorded in a recording medium.

FIG. 3 is a flowchart illustrating an example of a translated word replacement function of FIG.

FIG. 4 is a conceptual diagram showing a relationship between a first language and a second language when translating from the first language into the second language.

FIG. 5 is a conceptual diagram showing a result of morphological analysis / syntactic analysis performed on a second language.

FIG. 6 is a conceptual diagram showing the relationship between the second language and the first language when translating from the second language to the first language.

FIG. 7 is a conceptual diagram showing the relationship between headwords and translated words.

FIG. 8 is a conceptual diagram showing a result of morphological analysis / syntactic analysis performed on the first language.

FIG. 9: Morphological analysis / retranslation for the re-translated first language
A conceptual diagram showing a plurality of results of parsing.

10 is a flowchart illustrating a processing operation of the translation apparatus shown in FIG. 1 and another example of a program recorded on a recording medium.

FIG. 11 is a diagram illustrating an embodiment of a translation server and an information transmission medium according to the present invention.

[Explanation of symbols]

1 ... First language input means 2 ... Translation processing main unit 3 ... Recording medium 4 ... Translation dictionary section 5 ... Data storage unit 6 ... Translation result output means 21 ... First translation processing means 22 ... Second translation processing means 23 ... Translation word determining means 24 ... Similarity Determining Means 71 ... server 75 ... Information transmission medium 78, 79 ... Client

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Etsuo Ito             No. 1 Toshiba-cho, Fuchu-shi, Tokyo Toshiba Corporation             Fuchu Office F term (reference) 5B091 AA03 BA03 CA21 CB24 EA14

Claims (15)

[Claims] 1. Input means for inputting a first language, first translation processing means for translating the input first language into a second language, and translated texts translated by the first translation processing means. A second translation processing means for further translating the word into a first language, and determining whether or not a translated word in the translated text translated by the second translation processing means exists in the input first language. If it does not exist, a translation word determining unit that searches the translation word of the corresponding second language word for the translation word for the same word as the input word in the first language and generates a replacement translation And a translation device that outputs the generated translated text. 2. The method according to claim 1, further comprising a similarity determination unit that sets, as a first candidate, a sentence having the highest similarity to the input first language among the generated translated sentences. The described translation device. 3. The translation device according to claim 2, wherein the similarity determination unit makes a determination based on a similarity of a syntax analysis tree. 4. The synonym of the word in the second language, when the translation word determining unit cannot find the same word in the input first language from the translations of the word in the second language. The translation device according to claim 1 or 2, further comprising a unit that replaces a word to generate a translated sentence. 5. A translation method for translating a first language input into a second language, the method comprising: capturing the first language; translating the captured first language into a second language; Further translating the translated sentence into the first language; determining whether the translated word in the translated sentence further translated into the first language exists in the imported first language; and If the translated word does not exist, a step of searching the translated word of the word in the second language corresponding to the translated word for the same word as the input word in the first language, and performing replacement translation generation And a step of outputting the generated translated sentence. 6. The translation method according to claim 5, further comprising the step of setting, as a first candidate, a sentence having the highest similarity to the imported first language among the generated translated sentences. . 7. The translation method according to claim 6, wherein the step of making the one having the highest degree of similarity the first candidate utilizes the degree of similarity of the parse tree. 8. The step of generating the translated sentence, if the same word in the input first language cannot be retrieved from the translated words of the word in the second language, the word in the second language 7. The translation method according to claim 5, further comprising a step of generating a translated sentence by replacing the synonym with. 9. A program for translating a first language input into a second language, the function including the first language, the function that translates the first language that has been input into the second language, and A function for further translating the translated sentence into a first language, a function for determining whether or not a translated word in the translated sentence further translated into the first language exists in the imported first language, and the determination function Therefore, if the translation does not exist,
A function of searching for the same word as in the input first language from the translations of the corresponding second language word of the translation and generating a replacement translation, and outputting the generated translation A translation processing program that causes a computer to realize functions. 10. The translation process according to claim 9, further comprising a function of, of the generated translated sentences, a sentence having the highest similarity to the imported first language as a first candidate. For programs. 11. The translation processing program according to claim 10, wherein the function of using the highest similarity as the first candidate uses the similarity of the parse tree. 12. The function for generating the translated sentence is the second
When the same word in the input first language cannot be retrieved from the translated words of the language, the computer is further provided with a function of generating a translated sentence by replacing it with a synonym of the word in the second language. 10. The method according to claim 9, wherein
Alternatively, the translation processing program according to claim 10. 13. The translation processing program according to any one of claims 9 to 12 is loaded, and when a client requests translation of a first language, the first language is set according to the program. A translation server connected to the network that provides translated translations to the requesting client. 14. The translation processing program according to claim 9 is loaded, and when a request for providing the program is received from a client,
A translation server connected to the network that sends the program to the requesting client. 15. The method according to any one of claims 9 to 12
A server that owns the program for translation processing described in the section and a plurality of clients are connected via a network,
An information transmission medium comprising the translation processing program transmitted from the server to the network under the program provision request of each client.