JPH11120176A - Interpreting device, method and medium storing interpreting device control program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To an interpreting device which accurately communicates speech of a speaker to the opposite conversation party who speaks in a different language, translates speech of the opposite conversation party into an own language and communicates it to the speaker. SOLUTION: This translating device is provided with a displaying part 2 that has a screen, an inputting part 3 which inputs speech in an own language of a speaker to the opposite conversation party who speaks in a different language, a parallel translation example retrieving part 54 which retrieves a parallel translation example dictionary that is preliminarily provided and outputs a parallel translation example which is related to the inputted speech to the part 2, an instruction inputting part 4 that indicates the parallel translation example, a voice converting part 51 which converts translation of the indicated parallel translation example into a voice signal, a voice outputting part 7 which performs voice output of translation of the voice signal, a voice inputting part 1 inputs speech of the opposite conversation party in a foreign language by a voice signal, a character converting part which recognizes the speech of an inputted voice signal and converts it into a speech sentence of character codes, a correcting part 56 with which the opposite conversation party can correct the converted speech sentence based on guidance and an automatic interpreting part 52 which automatically interprets the corrected speech sentence into a speech sentence of the own language and outputs it to the part 2.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an interpreting apparatus and method applied to a portable information device for interpreting conversation between speakers using different languages, and a medium storing an interpreting apparatus control program.

[0002]

2. Description of the Related Art Conventionally, as an interpreting apparatus for translating different languages, a translation example sentence prepared in a native language prepared in advance (an example sentence in which a native language and a foreign language are paired) is selected in a menu format, and the selected translation is selected. Interpreters that output multilingual sentences of example sentences by voice are commercially available.

[0003] This translator has the advantage that it is possible to present a correct foreign language sentence to the conversation partner by voice only if the contents of the bilingual example sentence are confirmed in the sentence written in its own language. There is a problem in that it is not possible to transmit utterances in another language from the conversation partner to the speaker. Because, in order to convert this translation device from the bilingual example sentence related to the dialogue partner's utterance to the native language by passing the interpreter to the dialogue partner, it is necessary to predict and prepare the dialogue partner's utterance example. There is a problem that it is difficult to prepare enough examples. Further, there is a problem that it takes time to search for registered translation examples.

Further, a machine translation apparatus has been developed which recognizes input speech by speech, automatically translates the recognition result, and outputs the translation result as speech. However, in a popular portable interpreter, a full-scale automatic translation processing circuit that generates a translated sentence by syntactically or semantically analyzing an input sentence cannot be used because of its large scale. Therefore, in a portable interpreter, erroneous speech recognition and erroneous translation cannot be avoided at a fixed rate for speech recognition and automatic translation. For example, if translation is performed from an erroneous speech recognition result, erroneous translation will occur, which is not practical.

As a prior art for improving the problem of how to deal with such erroneous speech recognition and erroneous translation, according to Japanese Patent Application Laid-Open No. 64-88875, Japanese Patent Application Laid-Open No. 64-88875 discloses language information before automatic translation. The language information after automatic translation is output together with the voice and the display character string. By comparing the language information before and after the translation, the error of the automatic translation is checked, and the dialogue between different languages is effectively performed. An advanced speech translation device has been proposed. Further, according to the description of Japanese Patent Application Laid-Open No. 7-129594, an automatic translation system that asks the operator for confirmation when ambiguity arises in the result of voice recognition and translation, and also outputs a situation explanation to the other party at the same time. Has been proposed.

[0006]

However, Japanese Patent Application Laid-Open No. 64-88875 and Japanese Patent Application Laid-Open No. 7-12959 are known.
In the prior art using the automatic translation disclosed in Japanese Patent Laid-Open No. 4 (1999) -1995, the speech recognition result of the speech information input by the interlocutors can be confirmed and corrected, but ambiguity arises in the automatic translation result of the recognition result. Even if you ask for confirmation at the time, it cannot be confirmed. This is because even if the input speech information is corrected correctly, it cannot be confirmed whether the sentence should be translated automatically, so it is correctly converted to another language and the utterance of the speaker is accurately transmitted to the other party. Not necessarily.
Therefore, there is a problem that it is not practical because communication between the interlocutors cannot be achieved.

The present invention has been made in consideration of the above circumstances. For example, one-way (speaker) translation is performed by searching for a bilingual example sentence, and the other-direction (dialogue) interpretation is performed by automatic translation. By doing so, the speaker's utterance is accurately transmitted to the conversation partner in different languages without using a large-scale translation processing circuit, and the utterance from the conversation partner is converted to the native language and Provide a practical translator that can be transmitted to

[0008]

According to the present invention, there is provided a display unit having a screen, an input unit for inputting an utterance in a speaker's own language to a conversation partner having a different language, and a bilingual translation dictionary provided in advance. A bilingual example sentence search unit that outputs a bilingual example sentence related to the input utterance sentence to the display unit; and a bilingual example sentence is designated by the input unit. A voice conversion unit, a voice output unit for voice output of a translation of the voice signal, a voice input unit for inputting speech in another language of the conversation partner as a voice signal, and recognizing the utterance of the input voice signal. A voice recognition unit that converts the utterance sentence to the character code, a correction unit that can correct the utterance sentence converted to the character code based on the guidance, and automatically converts the corrected utterance sentence to the native language utterance. Automatic translation to translate and output to the display unit A translator device being characterized in that a part.

[0009] In the present invention, the input unit is a pen
It is preferable to be configured with a tablet and an icon. The utterance of the speaker is configured to be input not by voice input but by handwritten character input or a menu format of the input unit, but is input by voice by the voice input unit, and is translated via the voice recognition unit. You may be comprised so that a dictionary may be searched. It is preferable that the display unit is constituted by an LCD (liquid crystal display) panel.

[0010] Preferably, the audio input section is constituted by a microphone and an A / D converter. In particular, it is used by a conversation partner to input an utterance in another language. It is preferable that the audio output unit includes an A / D converter and a speaker. A bilingual example dictionary, a bilingual example sentence search unit, a speech conversion unit, a speech recognition unit, a correction unit, and an automatic translation unit are a CPU, a RO
It is preferable to configure a microcomputer including M, RAM, and I / O ports. The bilingual example dictionary is composed of the ROM therein.

According to the present invention, a bilingual sentence related to an utterance inputted in a native language by a speaker with respect to a dialogue partner in a different language is searched, accurately translated into another language, and transmitted. After converting utterances input in Japanese into utterances, the conversation partner corrects them with guidance and automatically translates them into utterances in their own language and displays them, without using a large-scale translation processing circuit. In addition, a practical translator can be provided that can accurately transmit the utterance of a speaker to a conversation partner of a different language, and can also convert the utterance of the conversation partner into a native language and transmit it to the speaker.

[0012] The automatic translation unit further includes a translation dictionary for translating a different language into a native language, and a dictionary lookup unit for searching for a translation of the native language from words of an utterance sentence in another language using the translation dictionary. The corrected utterance sentence in another language and its translation may be output to the display unit in a corresponding manner. The translation dictionary is an RO that stores the data for automatic translation.
M. The dictionary lookup unit is configured by a ROM storing a dictionary lookup program. According to the above configuration, the utterance of the conversation partner is displayed on the display unit in correspondence with the translated word extracted from the dictionary. If the translation is not appropriate, the speaker can correct the utterance of the conversation partner by re-reading the translation dictionary. Can be understood.

[0013] The speaker may further include a remote instruction unit having a guidance display unit, an operation unit, and a communication unit for remotely confirming and correcting the speech recognition result and controlling the guidance for the conversation partner. According to the above configuration, even if the speaker hands the interpreter to the conversation partner, the speaker can confirm the utterance sentence input by voice and give a guidance for a correction request by remote control.

The bilingual example sentence searching unit searches the bilingual example sentence by referring to the bilingual dictionary to determine a degree of coincidence between a word of an utterance input by a speaker and a characteristic word of the bilingual example sentence. , May be configured to output to the display unit. According to the above configuration, the speaker can easily search and select a bilingual example sentence to be transmitted to the other party.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on embodiments shown in the drawings. Note that the present invention is not limited to this.

FIG. 1 is a block diagram showing a hardware configuration of an interpreter according to the present invention. In FIG. 1, reference numeral 1 denotes an audio input unit including a microphone and an A / D converter.
In particular, it is used by a conversation partner to input an utterance in another language. Reference numeral 2 denotes a display unit composed of an LCD (liquid crystal display) panel or the like. The display unit 2 displays the utterance sentence of the speaker and its parallel translation example sentence, and the utterance sentence of the conversation partner and its translation.

Reference numeral 3 denotes an input unit which is composed of a pen / tablet and an icon, and which retrieves and inputs a bilingual example sentence of a speaker's own language in a menu format. An instruction to switch the language to be input and an instruction to correct the displayed utterance can be given. Reference numeral 4 denotes a remote instruction input unit including a guidance display unit, an operation unit, and a communication unit. The speaker gives the main body device to the conversation partner, and the remote instruction input unit 4 provides guidance for confirmation and correction request of the utterance sentence input by the conversation partner.

Reference numeral 5 denotes a processing unit constituted by a ROM storing a program controlled by a computer.
The processing device 5 recognizes the utterance of the input audio signal and converts the utterance into an utterance of a character code, and automatically translates the corrected utterance into an utterance in its own language and outputs it to the display unit. An automatic translation unit 52, a dictionary lookup unit 53 for searching for a translation of a native language from words of an utterance sentence in another language using a translation dictionary, and a bilingual example dictionary (bilingual example sentence data memory) for searching and relating to the input utterance sentence The translated bilingual example sentence is output to the display unit, the translated bilingual example sentence searching unit 54, the voice converting unit 55 that converts the translation of the designated bilingual example sentence into a voice signal, and corrects the uttered text converted into the character code based on the guidance. And a communication unit 57 for transmitting communication data such as guidance.

Reference numeral 6 denotes a storage device comprising a ROM and a RAM. The storage device 6 stores a voice recognition data memory 6 for storing data obtained by recognizing voice signals and converting them into character codes.
1. Automatic translation data memory 62 storing translation dictionary data, etc .; dictionary lookup data memory 63 storing dictionary-translated translations; bilingual example sentence data memory 64 storing bilingual example dictionary data; A speech conversion data memory 65 for storing the input data, an utterance data memory 66 for temporarily storing utterance data inputted by key input or voice input, and the like. Reference numeral 7 denotes an audio output unit which includes an A / D converter and a speaker and outputs a translation of an audio signal.

Reference numeral 8 denotes a control unit constituted by a CPU of the computer. Reference numeral 9 denotes an external storage medium such as an EEPROM memory card or MD. The external storage medium 9 is configured to store the interpreter control program / data of the present invention controlled by the CPU of the control unit 8.
In particular, to enable language conversation between countries, programs such as bilingual example sentences, speech recognition programs, translation dictionaries, etc.
The data may be stored in an external storage medium and various types of data may be prepared.

Reference numeral 10 denotes a medium reading device for reading programs / data stored in the external storage medium 9. The CPU of the control unit 8 installs the programs / data in the memories of the processing device 5 and the storage device 6. Communication unit 5
7 to receive an interpreter control program / data from a host computer. 1
A bus 1 transfers address data and control data between each device and each unit.

As shown in FIG. 1, according to another aspect of the present invention, there is provided a medium storing a program for controlling an interpreting device by a computer, wherein the control program causes a computer to communicate with a conversation partner having a different language. When the utterance sentence of the speaker's own language is input by the input unit 3, a bilingual example dictionary provided in advance is searched, and a bilingual example sentence related to the input utterance sentence is output to the display unit 2. When the parallel translation example sentence is instructed by the input unit 3, the translation of the instructed parallel translation example sentence is converted into a voice signal, the translation of the voice signal is voice-output by the voice output unit 7, and the voice input unit 1 interacts. When the utterance of the other party's language is input as an audio signal,
Recognizes the utterance of the input voice signal, converts it to a character code utterance, corrects the utterance converted to the character code based on the guidance, and automatically translates and displays the corrected utterance to your native language There is provided a medium 9 storing an interpreter control program which is output to the section 2.

FIG. 2 is a view showing the appearance of the interpreter of the present invention. 2, the same components as those in FIG. 1 are denoted by the same reference numerals. As shown in FIG. 2, the interpreter of the present invention is configured for portable use. Display unit 2 and input unit 3
Are integrally formed, but may be separately formed. Although the remote instruction input unit 4 is connected by wire, the communication unit 57 may be configured to perform wireless communication with the processing device 5.
Further, a telephone function may be added to constitute a multifunctional interpreter.

FIG. 3 is a flow chart showing the procedure of the processing of the interpreter of the present invention. FIG. 4 is a flowchart showing the procedure of the bilingual example sentence search process of this embodiment. Hereinafter, an example in which a speaker A who speaks Japanese speaks English with a conversation partner B who speaks English while possessing the interpreter of the present invention will be described with reference to the flowcharts of FIGS.

First, an input language is set by the input unit 3 (S301). In the present embodiment, the speaker A specifies to input in Japanese, and the conversation partner B specifies to input in English. The result is sent to the voice recognition unit 51. Since the input language can be estimated from the input phoneme sequence in the voice recognition unit 51, the processing in step S301 can be omitted. Next, the speech from the speaker A is inputted as a speech signal by the speech input unit 1 (S302).

Subsequently, the voice signal is sent to a voice recognition unit 51, which performs voice recognition according to the designation of the input language, and converts the utterance from speaker A into a Japanese character code string ( (S303), and display on the display unit 2 (S30).
4). Here, it is confirmed whether or not there is an error in the utterance sentence converted by the speech recognition (S305), and correction is made by the input unit 3 until there is no error (S306).

FIG. 5 is a diagram showing an example of correcting the speech recognition result and the bilingual example sentence according to the present embodiment. As shown in FIG. 5 (a), an example of correction of the speech recognition result is shown in FIG.
May be directly rewritten by the input unit 3. Alternatively, another recognition candidate “correspondence” for the part of the voice recognition result “opened” may be selected. The corrected speech recognition result is stored in the speech recognition data memory 61 as an utterance input. FIG. 5B shows an example of correcting the bilingual example sentence.

It is determined whether or not the input is in the native language of the speaker A (S307). If the input is in the native language of the speaker A, a translation output process by bilingual example sentence search, which is the process of S308, the story If the input is in another language of person B, S311
Then, the process proceeds to a translated sentence generation process by automatic translation.
Since it is determined in S301 that the input language is the native language of speaker A, the utterance information converted into the Japanese character code is sent to the bilingual example sentence search unit 54.

Since the utterance information of the speaker A is the native language, the processing of steps S302 to S307 is not performed, that is,
Instead of inputting the voice, the utterance sentence (Japanese character string) in the native language to be transmitted to the conversation partner is directly input by the input unit 3 in the form of a handwritten character or a menu.
The processing may shift to the processing of No. 8.

FIG. 4 is a flowchart showing a subroutine of the bilingual example sentence search processing of FIG. In FIG. 4, the bilingual example search unit 54 extracts a characteristic word from the sent utterance sentence (Japanese character string) in its own language (S401). In the bilingual example sentence search, the bilingual example data used is a bilingual example dictionary that stores a bilingual example character string and a characteristic sequence representing the characteristics of the example, and a characteristic word dictionary that stores words that are the respective feature elements described in the characteristic sequence. And extracts a matching example sentence by collating the characteristic word extracted from the utterance input with each example sentence (S402).

FIG. 6 is a diagram showing an example of collation between a characteristic word extracted from an utterance input and a bilingual example sentence according to the present embodiment. As shown in FIG. 6A, an utterance input sentence and its characteristic words are specified. For example, if an utterance input “I want you to carry a suitcase” is specified, as shown in FIG. "Please bring me", "I want the porter to carry my luggage" and two English translations of the example sentences are displayed.

In practice, in S401, three characteristic words of "suitcase", "carried", and "desired" are extracted. The extracted characteristic words are not displayed. S40
In 2, the characteristic word “suitcase” is treated as an element of the characteristics “luggage”, “openable”, and “bag” for searching example sentences. Similarly, for the feature word “carry”, an example sentence is searched as an element of the feature “carry”. The feature word "desu" indicates one of the sentence end features (requests) of the example sentence <Re
quest>.

From the utterance input “I want you to carry a suitcase” shown in FIG. 6A, use example 1 “Please bring your luggage” and use example 2 shown in FIG. 6B, and want to have the porter carry your luggage. The two example sentences "I'm sorry" and their English translations are searched and displayed. In the feature sequence (not shown) in FIG. 6B, square brackets “[]” indicate features that are not essential. The vertical bar symbol “|” in the parenthesis symbol “()” indicates the logical sum of the features (any feature). The angle bracket symbol “<
>> indicates a sentence end feature, for example, “<Request>” indicates a sentence end feature “request”.

The bilingual example search unit 54 includes a counter, extracts and counts features including characteristic words extracted from the utterance input from the bilingual example sentence, and includes a characteristic sequence including the characteristic words extracted from the input. The example sentences are sorted in descending order of the feature ratio (S403). When matching the three feature words extracted from the utterance input “I want you to carry a suitcase” with the two parallel translation examples, example 1 “Please bring your luggage”
In this case, "package", "[having] of" ([carrying] | [having]) "" and "<Request>" are counted, and 100% of the essential features in the feature column are satisfied.

In the example 2 "I want the porter to carry the luggage," the element of the feature "porter" of "porter", "(luggage | this)", "carry", and "<Request>" Is not included in the utterance input, so only 75% of the essential features in the feature sequence are satisfied.
Therefore, Example 1 has priority over Example 2.

The translation example sentence candidates are displayed on the display unit in the sorted priority order (S404). In step S404,
All the parallel translation example sentence candidates may be displayed, or only some of them may be displayed in priority order. In this embodiment, FIG.
As shown in (b), a bilingual example sentence of a Japanese sentence and an English sentence is displayed, so that speaker A determines which bilingual example sentence candidate to select from the description of the Japanese sentence.

From the displayed bilingual example sentence candidates, speaker A
Selects one bilingual example sentence that seems to be optimal and instructs it via the input unit (S405). For example, when the translation example sentence shown in FIG. 5B is selected, the variable part of the English sentence may be rewritten, or the variable part of the Japanese sentence may be selected and changed in a menu format, and the English sentence may be changed accordingly. The units may also be rewritten by a method of exchanging. The bilingual example search unit 54 outputs only the selected bilingual example sentence to the display unit and the speech conversion unit (S406).
The process proceeds to S309 shown in FIG.

The voice conversion unit 55 converts only the transferred character code string of the English translation of the translation example sentence into a voice signal and outputs it to the voice output unit (S309). At this time, the speech synthesis method of the speech signal may use a rule synthesis method that is not restricted by the vocabulary, and the output result of the bilingual example search has a vocabulary restriction in advance. An analysis / synthesis method with good sound quality may be used, but if arbitrary editing of bilingual example search or automatic translation output is permitted, it is desirable to adopt a rule synthesis method. The sound output unit outputs the transferred sound signal (S31).
0).

If the conversation using the interpreter does not end here (S311), the language is switched to the input language corresponding to the speaker (S312). The language switching operation may be manually switched by the input unit 3 or the remote instruction input unit 4 each time, or may be automatically switched assuming that the utterance is performed alternately. In this embodiment, the input language is switched from Japanese to English, the process returns to step S301, and the utterance of speaker B is input to the voice input unit.

Steps S302 to S307 are performed in the same manner as in the case of speaker A's speech, except that the input language has been switched from Japanese to English. In step S307, in the case of the input from speaker B in English, the utterance information in which the English sentence is coded is transferred to the automatic translator 52 or the dictionary lookup unit 53. When the automatic translation unit 52 analyzes the English sentence of the utterance to generate a Japanese sentence as a translation result and transfers it to the dictionary lookup unit 53, the automatic translation unit 52 performs a dictionary lookup on the words and idioms included in the utterance (S314). As a result, the translated word is displayed on the display unit (S315).

With the above-described configuration, by using the interpreting method based on the bilingual example sentence search and the interpreting method based on the automatic translation, the advantages of the respective interpreting functions can be utilized. However, if the owner of the device fully understands the operation method of the interpreting device, it is possible to support communication between two persons having different languages.

The content of the utterance transmitted to the conversation partner can be accurately confirmed by the owner of the device in his / her own language. Further, the owner of the device can understand the automatic translation performance of the device in advance. Even if an erroneous translation occurs in the automatic translation, for example, the utterance from the conversation partner can be confirmed again by a dictionary lookup operation.

FIG. 7 is a diagram showing a display example of a translation result and a dictionary lookup result according to this embodiment. As shown in FIG. 7, for example, the input English sentence "We have been doing business togeth"
er for a while. ", the Japanese sentence" We have been trading together for a while "is displayed. If you can almost understand the meaning only by looking at the Japanese language, it will be completed only by the automatic translation process.

When the user wants to check the English sentence and switches to dictionary lookup processing, for example, the input English sentence “We have been d
oing business together for a while. '' and `` We, '' `` doing business, '' `` toge
The translations for "ther" and "for a while" are displayed as "we,""transact,""together," and "a little while." When checking the English sentence, the translated word is also associated and easy to understand.

When the utterance information is transferred to the dictionary lookup unit 53, the dictionary lookup unit 53 analyzes the contents of the utterance from the input English sentence, and assigns the most appropriate translation to each word included in the utterance. In S315, each translated word is displayed on the display unit in a form juxtaposed with the corresponding word. A plurality of translations of each word may be displayed with a priority, or there may be some words to which no translation is given. Also, a translation may be given to a plurality of word sets, such as idioms.

The translation result of the automatic translation unit and the translated word of the dictionary lookup unit may be selectively displayed, or may be simultaneously displayed on the display unit. If the conversation is to be continued, the language is switched at S312 to repeat the conversation. Note that, in this embodiment, an example is given in which the owner of the interpreter speaks Japanese and the other party speaks English, but any language can be set as the language used by both in a similar configuration. Is possible.

[0047]

According to the present invention, a large-scale system can be realized by a configuration in which the translation in one direction (speaker) is performed by searching for a translation example sentence, and the translation in the other direction (dialogue) is performed by automatic translation. Practical interpreter that accurately communicates the utterance of the speaker to the conversation partner in different languages without using a translation processing circuit, and can also translate the utterance of the conversation partner into their own language and transmit it to the speaker An apparatus can be provided.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a hardware configuration of an interpreter of the present invention. .

FIG. 2 is a view showing the appearance of the interpreter of the present invention.

FIG. 3 is a flowchart showing a procedure of a process of the interpreter of the present invention.

FIG. 4 is a flowchart illustrating a procedure of a bilingual example sentence search process according to the embodiment;

FIG. 5 is a diagram showing an example of correcting a speech recognition result and a bilingual example sentence according to the embodiment.

FIG. 6 is a diagram illustrating an example of collation between a feature word extracted from an utterance input and a bilingual example sentence according to the embodiment;

FIG. 7 is a diagram showing a display example of a translation result and a dictionary lookup result according to the embodiment.

[Explanation of symbols]

 Reference Signs List 1 voice input unit 2 display unit 3 input unit 4 remote instruction input unit 5 processing device 51 voice recognition unit 52 automatic translation unit 53 dictionary lookup unit 54 bilingual example sentence search unit 55 voice conversion unit 56 correction unit 57 communication unit 6 storage device 61 Voice recognition data memory 62 Automatic translation data memory 63 Dictionary lookup data memory 64 Parallel translation example sentence data memory 65 Voice conversion data memory 66 Utterance data memory 7 Voice output unit 8 Control unit 9 External storage medium 10 Medium reading device 11 Bus

Claims (9)

[Claims] 1. A display unit having a screen, an input unit for inputting an utterance sentence of a speaker's own language to a conversation partner having a different language, and an input unit associated with the input utterance sentence from a bilingual example dictionary provided in advance. A bilingual example sentence search unit that searches for and outputs a bilingual example sentence to the display unit, a speech conversion unit that indicates a bilingual example sentence by the input unit, and converts a translated sentence of the indicated bilingual example sentence into a voice signal; An audio output unit that outputs a translation of an audio signal, an audio input unit that inputs an utterance in another language of a conversation partner as an audio signal, and recognizes the utterance of the input audio signal and converts the utterance of the input audio signal into an utterance of a character code. A speech recognition unit, a correction unit capable of correcting the utterance sentence converted into the character code based on the guidance, and automatically correcting the corrected utterance sentence into an utterance sentence of a native language and outputting the same to the display unit. It has an automatic translator and Interpreter. 2. The automatic translation unit further includes: a translation dictionary for translating a different language into a native language; and a dictionary lookup unit for searching a translation word of a native language from words of a spoken sentence of another language by the translation dictionary. 2. The interpreting apparatus according to claim 1, further comprising: outputting the corrected utterance sentence in another language and a translation thereof to the display unit. 3. A remote instruction input unit having a guidance display unit, an operation unit, and a communication unit for allowing a speaker to confirm and correct a speech recognition result with respect to a conversation partner and remotely control the guidance. 2. The interpreter according to claim 1, wherein: 4. The bilingual example sentence search unit determines the degree of coincidence between a word of an utterance input by a speaker and a characteristic word of the bilingual example sentence with reference to the bilingual dictionary to determine the bilingual example sentence. The interpreting apparatus according to claim 1, wherein the interpreting apparatus searches and outputs the retrieved information to the display unit. 5. An utterance sentence of a speaker's own language is input to a conversation partner having a different language by an input unit, and a bilingual example dictionary provided in advance is searched to obtain a bilingual example sentence related to the input utterance sentence. Output to a display unit, instruct the bilingual example sentence by the input unit, convert the translated sentence of the instructed bilingual example sentence to a voice signal, output the translated sentence of the voice signal from the voice output unit, and interact with the voice input unit. Input the utterance of the partner's other language as a voice signal, recognize the utterance of the input voice signal, convert it to a character code utterance, and correct the utterance converted to the character code based on guidance, and correct it An interpreting method for automatically translating the uttered sentence into an uttered sentence in a native language and outputting the uttered sentence to the display unit. 6. A translation dictionary for translating a different language into a native language provided in advance, searching for a translation of the native language from words of the utterance sentence of another language, and retrieving the corrected utterance sentence of the other language and its translation. 6. The display according to claim 5, wherein
Interpretation method described. 7. A speaker in which a remote instruction input unit having a guidance display unit, an operation unit, and a communication unit provided in advance enables a speaker to confirm and correct a speech recognition result to a conversation partner and remotely control the guidance. The interpreting method according to claim 5. 8. A bilingual example sentence is searched for by referring to the bilingual dictionary and determining a degree of coincidence between a word of an utterance input by a speaker and a characteristic word of the bilingual example sentence, and is output to a display unit. 6. The interpreting method according to claim 5, wherein: 9. A medium storing a program for controlling an interpreting device by a computer, wherein the control program inputs an utterance sentence of a speaker's native language to a conversation partner having a different language from an input unit. Then, the bilingual example dictionary provided in advance is searched and the bilingual example sentence related to the input utterance sentence is output to the display unit. When the input unit indicates the bilingual example sentence, the instruction is given. When the translated sentence of the bilingual example sentence is converted into a voice signal, the translated sentence of the voice signal is output as a voice by a voice output unit, and when an utterance in another language of the conversation partner is input as a voice signal by the voice input unit, the input voice is input. Recognize the utterance of the signal and convert it to a character code utterance, correct the utterance converted to the character code based on the guidance, and automatically translate the corrected utterance into the native language utterance A medium for storing an interpreter control program, the program being output to the display unit.