JPH08846U - Foreign language conversation support device - Google Patents

Abstract

(57) [Summary] (Correction) [Purpose] To promote mutual understanding by visually displaying pronunciation and conversation in English and Japanese. [Structure] A voice input unit 2 for inputting a foreign voice, a voice recognition unit 2 for recognizing the voice as a series of foreign language character strings,
Operation of a foreigner who is a speaker on the display unit 2 that displays a series of recognized foreign language character strings to a foreigner who is a speaker, the display unit 1 that simultaneously displays to the Japanese who is a listener, and the display unit 2. Is provided with a blank insertion pen input unit 2 for dividing a series of foreign language character strings into some more simplified character strings. The foreign language electronic dictionary 2 and some simplified character strings are re-edited into meaningful foreign language words and displayed on the display unit 1 and the display unit 2. A foreign language syntax processing unit 2 for displaying on the display unit 1 is provided. The voices of Japanese people will also be structured accordingly.

Description

[Detailed description of the device]

[Industrial field of application] The present invention relates to a device that supports conversation between Japanese and foreigners. [Prior art] It is becoming practical to capture the pronunciations of Japanese and foreigners as single notes or continuous single notes, but it has not reached the point where the input voice is regarded as a meaningful sentence. ． [Issues to be solved by the invention] Traditionally, it has been said that Japanese can read and write English, but they are not good at listening and speaking. In other words, even if you have 6 years of junior high school, 8 years of high school, or even 10 years of language education including university, you can read and write some English, but many of you cannot speak English and American. ． I can't speak or listen to foreign languages of this Japanese, but I take advantage of the fact that I am a little confident in reading and writing, and reduce the task of listening to foreign languages to the task of reading foreign languages. Let's make Japanese understand the meaning of foreigners' stories. Also, let's reduce the Japanese speaking of foreign languages to the writing of foreign languages so that the foreigners can understand the pronunciation of Japanese English that only Japanese can understand. This is the present invention. [Means for Solving the Problems] A voice input unit 2 for inputting a foreign language voice of a foreigner, a voice recognition unit 2 for recognizing the voice as a series of foreign language character strings, and a recognized series of foreign languages. On the display unit 2 that displays the character string toward the foreigner who is the speaker, at the same time, the display unit 1 that displays toward the Japanese who is the listener, and on the display unit 2 the foreign character who is the speaker. By a human operation, a series of foreign language character strings is further simplified into a plurality of character strings, a blank insertion pen input unit 2, an electronic dictionary 2 which is a foreign language word database, and a single unit. Some of the purified character strings are re-edited into meaningful foreign words based on the data in the foreign language electronic dictionary 2 and displayed on the display unit 1 and display unit 2. Using the electronic dictionary 2 to display the Japanese translation on the display unit 1 The word syntax processing unit 2, the voice input unit 1 for inputting the Japanese voice, the voice recognition unit 1 for recognizing the voice as a series of character strings, and the operation of the Japanese speaker as the speaker on the display unit 1. The pen input unit 1 for blanking, which divides a series of characters uttered by a Japanese person into several more simplified character strings, the electronic dictionary 1 which is a Japanese word database, It consists of a Japanese syntax processing unit 1 that compiles the converted Japanese character strings into meaningful Japanese words using the electronic dictionary 1, and also adds a translation into a foreign language, and a central control unit that controls the whole. Foreign language conversation support device. [Example] Here, a conversation between a Japanese and an American is given. The American pronounces I want your products while touching the foreign language key 2. Pressing the foreign language key 2 indicates that the non-specific speaker recognition method is used for American speech. The sounded voice signal is input from the voice input unit 2, is recognized by the voice recognition unit 2 as I want your products, and is output to the display unit 2 for foreigners and the display unit 1 for Japanese. Figure 2 is that. The speaker, the American, uses the blank input pen input unit 2 to specify w, y, and p on the display unit 2 as shown in FIG. Then, this character string is converted into a meaningful sentence as shown in FIG. The converted sentence is immediately output to the display unit 1 and the display unit 2. Further, as shown in FIG. 5, only the main part of one sentence, for example, y is specified by the blank input pen input section 2, and when it becomes as shown in FIG. 6, the foreign language database of the electronic dictionary 2 is displayed. Based on, the foreign language syntax processor 2 may automatically create the character strings in FIG. Here, the data structure of the foreign language electronic dictionary 2 is as shown in FIG. This is an index file that uses the entry words of foreign languages as keys. From FIG. 6, the above-described process of automatically generating the character string of FIG. 4 is processed based on the found word data of the electronic dictionary 2. This foreign language syntax processing unit 2 is realized by software, etc., but in order to compensate for it, the blank pen input unit 2 must be manually operated. In other words, one sentence, a series of character strings, is manually divided into several more simplified character strings such as I want or your products by the blank insertion pen input unit 2, The simplified character string is converted by the foreign language syntax processing unit 2 into meaningful words such as I wan t or your products. The blank pen input section here may be a pen computer, a light pen or the like, or a mouse. Also, a mechanism for automatically inserting a blank into a touch with a finger, that is, a portion touched by the finger, may be provided on the display unit. If you are a Japanese who is good at reading English, you can get Figure 4 and get a good sense of what foreigners are talking about, but I think that some words have meaningless meanings. Be done. Therefore, as shown in FIG. 7, the Japanese is displayed below and output to the display unit 1 based on the translated word data of the database of the electronic dictionary 2. In FIG. In order to prevent it, only nouns and verbs such as want and products are displayed as translations. In this way, even a Japanese who has forgotten the meaning of product or has never studied this word can grasp the meaning of the English sentence. The following is an example of a response when a Japanese responds in English. When the owner of this device is a Japanese and that Japanese speaks English, the voice is processed by the specific speaker recognition method. This is judged by the Japanese speaking while touching the foreign language key 1. Being the owner of this device means that a person's voice pattern can be registered in plaintext. It is assumed that the Japanese pronounces from the voice input unit 1 as "We sell our products fen millilion". This pronunciation is recognized by the voice recognition unit 1 as a character string as shown in FIG. The blank is inserted before fen in Fig. 8 because the blank of the speaker's pronunciation is processed as the blank of the display. When two positions are designated on the display unit 1 by the blank insertion pen input unit 1 as shown in FIG. 9, the character string is changed to that shown in FIG. Due to the function of the section 2, it is output as shown in FIG. 11 to the display section 1 for Japanese and the display section 2 for foreigners. Here, since the Japanese touches the foreign language key 1, the function of the foreign language syntax processing unit 2 is automatically performed. Although not impossible to obtain, the blank input pen input unit 1 is used in order to go through FIGS. 8 to 11 in real time. Practically, because Japanese Japanese is not understood by foreigners as English, when the character string in Fig. 8 is displayed, the foreigner in the other party can understand the rough meaning. Is. The blank input pen input section 1 is used as shown in Fig. 9 according to the complexity of the sentence. It should be noted that if the performance of the software of the foreign language syntax processing unit 2 is improved and the cpu processing speed is improved, it is possible to automatically obtain FIGS. The above is the case where the Japanese listens to the English conversation of a foreigner and the Japanese answers in English, but there are also cases where the Japanese cannot speak English at all. In this case, in response to the American I want your products words, all English words are translated into Japanese as shown in FIG. 12 instead of in FIG. 7 on the display unit 1, and as translated words as shown in FIG. The Japanese is not just a line of words, but is displayed by the Japanese syntax processing unit 1 by being corrected to an ordinary Japanese. This ordinary Japanese sentence display is made in different colors or underlined in order to distinguish it from the raw Japanese sentence in FIG. In response to this American word, the Japanese says, while pressing the Japanese key 1, from the voice input unit 1 that it will sell for 10 million yen. Pressing Japanese key 1 here indicates that the specific speaker recognition method is used for Japanese pronunciation. The above pronunciation is understood by the function of the voice recognition unit 1 and is output to the Japanese display unit 1 and foreign display unit 2 as shown in FIG. When a Japanese person specifies "U" in the blank pen input section 1 as shown in FIG. 15, one sentence becomes two more simplified words as shown in FIG. 16, and the data structure shown in FIG. With the function of the Japanese syntax processing unit 1 which uses the database of the Japanese electronic dictionary 1 having the above, it is further simplified as shown in FIG. 17, and the display unit 1 for Japanese and the display unit 2 for foreigners are provided. Is output to. The display unit 2 is displayed as shown in Fig. 18 by the database of the translated words of the Japanese electronic dictionary 1. In the form with an English translation. Furthermore, as shown in FIG. 19, English as a translated word is not a mere word string, but the foreign language syntax processing unit 2 converts it into ordinary English and outputs it to the display unit 2. In order to distinguish this normal English display from the raw English display in FIG. 18, it is output below the raw English display by changing the color or drawing an underline. What is displayed in this way can also be printed by printing section 1 and printing section 2. Therefore, for example, even if the English conversation becomes long, the listener, the Japanese, can listen to the new conversation while looking at the previously spoken English sentence and its translated sentence. This is true even when a foreigner hears a Japanese Japanese conversation or a Japanese Japanese conversation. In these cases, it is possible to save the trouble of collecting the minutes. This device can be realized by connecting two book type computers, or can be realized by providing two display parts in one book type computer, one computer and one display thereof. It is also possible to realize the department by sharing the department with Japanese and foreigners. That is, although FIG. 1 is a functional block diagram of this apparatus, when it is realized as hardware, it may be as shown in FIG. In Fig. 22, each function for Japanese and foreigners is put together. This device can be used not only for conversation, but also for Japanese speakers who speak foreign languages in foreign languages when the display unit 1 for Japanese people is a large screen TV or a screen for projectors. If the display unit 2 for foreigners is a large-screen TV or a screen, it can be used for lectures in foreign languages or Japanese by foreigners who are Japanese. Furthermore, if this device is applied to a foreign language recording tape or listening to a foreign language broadcast, a character string as shown in Fig. 23 will be obtained continuously. In this case, the blank produced by the speaker appears only as a blank in the character string because there is no blank inserted by the foreigner who is the speaker, but the listener is good at reading English. If this is the case, it is expected that this will allow you to read the meaning of the other party. In addition, if the processing performance of the foreign language syntax processing part is improved, it will be displayed as a more readable foreign language. In this case, the method of non-specific speaker recognition will be used, but in foreign language broadcasting, etc., the voice input is expected to be close to the standard pattern, so there is no speaker intervention, but good results are obtained. I can hope. The software of the foreign language syntax processing unit 2 that automatically edits your products into your products described in this embodiment can be realized by the following logic, for example. First, the index file of the foreign language electronic dictionary 2 is index-searched with yourproducts. Next, an index search is performed with yourproduct. Then, do an index search with yourproduct. Then, do an index search with yourprodu. Then perform an index search with yourprod. Then, do an index search with yourpro. Then, do an index search with yourpr. Then, do an index search with yourp. None of these results in an error because READ cannot be performed. This is because there is no headword corresponding to these. Then, do an index search with your. READ is possible. This is because the entry word “your” exists in the electronic dictionary 2. Thus, you can re-edit your products as your products. [Effect of the invention] A Japanese who can read and write English sentences can understand the meaning of the other person correctly by reading the conversation of the foreigner. By displaying the pronunciation of the English sentence, Japanese people can make foreigners understand the meaning of the pronunciation. Japanese people who are not good at English can understand the meaning of the other person by displaying the Japanese translation of the other person's English conversation. Then, by being displayed, foreigners can understand the meaning. This translated word is displayed by the functions of the Japanese and foreign language electronic dictionaries, the Japanese syntax processing section, and the foreign language composition processing section.

[Submission date] December 6, 1993

[Procedure Amendment 2]

[Document name to be amended] Statement

[Compensation target item name] Detailed explanation of the device

[Correction method] Change

[Correction content] [Detailed description of the device]

[Industrial field of application] The present invention relates to a device for supporting conversation between a Japanese and a foreigner.

[Prior Art] It is becoming practical to capture the pronunciations of Japanese and foreigners as a single sound or a continuous single sound, but until the input sound is understood as a meaningful sentence. I haven't arrived.

[ Problems to be Solved by the Invention ] Conventionally, it has been said that Japanese can read and write English, but they cannot listen to and speak English. In other words, even if you have 6 years of junior high school, 8 years of high school, or even 10 years of language education including university, you can read and write some English, but many of you cannot speak English and American. ． Taking advantage of the characteristic of Japanese people that they are not good at speaking and listening to foreign languages, but having a little confidence in reading and writing, we have reduced the task of listening to foreign languages to the task of reading foreign languages. , Let the Japanese understand the meaning of the foreigner's story. Also, let's reduce the Japanese speaking of foreign languages to the writing of foreign languages so that the foreigners can understand the pronunciation of Japanese English that only Japanese can understand. This is the present invention.

[ Means for Solving the Problem] A voice input unit 2 for inputting a foreign language voice of a foreigner, a voice recognition unit 2 for recognizing the voice as a series of foreign language character strings, and a recognized series of voices. The foreign language character string of is displayed to the foreigner who is the speaker, the display unit 1 that is displayed to the Japanese who is the listener at the same time, and the speaker is displayed on the display unit 2. By a foreigner's operation, a series of foreign language character strings is further simplified into a plurality of character strings, the blank pen input section 2 and the electronic dictionary 2 which is a foreign language word database. Then, some of the simplified character strings are re-edited into meaningful foreign words based on the data in the foreign language electronic dictionary 2 and displayed on the display unit 1 and the display unit 2. , Display part 1 with Japanese translation using foreign language electronic dictionary 2 Foreign language syntax processing section 2 to display, voice input section 1 to input Japanese voice, voice recognition section 1 to recognize the voice as a series of character strings, and display section 1 to be a speaker By the operation of the Japanese, a series of character strings uttered by the Japanese can be further simplified into several character strings, such as a blank input pen input section 1 and an electronic dictionary that is a Japanese word database. The book 1 and the simplified Japanese character string are re-edited into meaningful Japanese words by the electronic dictionary 1, and the Japanese syntax processing unit 1 that adds a translation into a foreign language is also controlled. A foreign language conversation support device consisting of a central control unit.

[0005] described [Operation] In this embodiment, foreign language syntax processor 2 software that automatically edit the text to your products that input speech yourproducts, as an example, wear in realization with the following logic. First, the index file of the foreign language electronic dictionary 2 is first index-searched by using the word products . Next, an index search is performed with yourproduct . Then, do an index search with yourproduct. Then, do an index search with yourprodu. Then perform an index search with yourprod. Then, do an index search with yourpro. Then, do an index search with yourpr. Then, do an index search with yourp. None of these results in an error because READ cannot be performed. These is because there is no look out words that correspond. Then perform an index search with "your". READ is possible. This is because the entry word “ your ” exists in the electronic dictionary 2. Thus you can re-edit your products as your products.

[ Examples] Here, a conversation between a Japanese and an American is given. An American pronounces I want your products while touching a foreign language key 2. Pressing the foreign language key 2 indicates that the non-specific speaker recognition method is used for American speech. This sounded voice signal is input from the voice input unit 2, is recognized by the voice recognition unit 2 as I wantyourproducts, and is output to the foreigner display unit 2 and the Japanese display unit 1. Figure 2 is that. The speaker, the American, uses the blank pen input section 2 to specify only the main part of a sentence on the display section 2 as shown in FIG. By the way, based on the foreign language database of the electronic dictionary 2, the function of the foreign language syntax processing unit 2 automatically creates the character strings in FIG. The converted sentence is immediately output to the display unit 1 and the display unit 2. Here, the data structure of the foreign language electronic dictionary 2 is as shown in FIG. This is an index file that uses the headword part of a foreign language as a key. From FIG. 6, the above process of automatically generating the character string of FIG. 4 is processed based on the found word data of the electronic dictionary 2. The foreign language syntax processing unit 2 is realized by software or the like, but in addition to this, the blank pen input unit 2 is manually operated. In other words, a sentence, a series of character strings, is manually divided into several more simplified Iwant or yourproducts-like character strings by the blank input pen input unit 2, and the simplified strings are divided. The character string is converted by the foreign language syntax processing unit 2 into a meaningful word such as I want or your products. The blank pen input section referred to here may be a pen computer, a light pen, or a mouse. If you are a Japanese who is good at reading English, you can get the meaning of a foreigner's story by getting Fig.4, but some of the words have meaningless meanings. Be done. Therefore, according to the data of the translated word part of the database of the electronic dictionary 2, as shown in FIG. In FIG. 7, translations are displayed only for nouns and verbs such as want and product ts in order to prevent the ambiguity of the reader. In this way, even Japanese who have forgotten the meaning of "product" or who have never studied this word can grasp the meaning of the English sentence.

[0007] Next, a response example of when the Japanese to respond in English. When the owner of this device is Japanese and the Japanese speaks English, the voice is processed by the method of specific speaker recognition. This is judged by the Japanese speaking while touching the foreign language key 1. Being the owner of this device means that a person's voice pattern can be registered in plain text. It is assumed that the Japanese pronounces from the voice input unit 1 as "We sell our products ten million yen". This pronunciation is recognized by the voice recognition unit 1 as a character string as shown in FIG. The blank is inserted before ten in Fig. 8 because the blank of the pronunciation of the speaker is processed as the blank of the display. When two positions are designated on the display unit 1 by the blank input pen input unit 1 as shown in FIG. 9, the character string is changed to that shown in FIG. By the function of the section 2, it is output as shown in FIG. 11 to the display section 1 for Japanese and the display section 2 for foreigners. Here, since the Japanese touches the foreign language key 1, the function of the foreign language syntax processing unit 2 is automatically performed. Although not impossible to obtain, the blank input pen input unit 1 is used in order to obtain FIGS. 8 to 11 in real time. Practically, because Japanese Japanese is not understood by foreigners as English, when the character string in Fig. 8 is displayed, the foreigner in the other party can understand the rough meaning. Is. The blank input pen input section 1 is used as shown in Fig. 9 according to the complexity of the sentence. It should be noted that if the performance of the software of the foreign language syntax processing unit 2 is improved and the processing speed of the cpu is improved, FIGS. 8 to 11 will be automatically obtained in real time.

The above is the case where the Japanese listens to the English conversation of a foreigner and the Japanese answers in English, but there are also cases where the Japanese cannot speak English at all. In this case, in response to the American I want your products words, all the English words may be displayed with Japanese translation as shown in FIG. 12 instead of FIG. 7 on the display unit 1. ．

In response to this American word, the Japanese says, while pressing the Japanese key 1, from the voice input section 1 to sell for 10 million yen. Pressing the Japanese key 1 here indicates that the specific speaker recognition method is used for Japanese pronunciation. The above pronunciation is understood by the function of the voice recognition unit 1 and is output to the Japanese display unit 1 and the foreign display unit 2 as shown in FIG. When a Japanese person specifies "U" in the blank pen input section 1 as shown in FIG. 15, one sentence becomes two more simplified words as shown in FIG. 16, and the data structure shown in FIG. With the function of the Japanese syntax processing unit 1 which uses the database of the Japanese electronic dictionary 1 having the above, the display is further simplified as shown in FIG. 17, and the display unit 1 for Japanese and the display unit 2 for foreigners are provided. Is output to. The display unit 2 displays the translated word database of the Japanese electronic dictionary 1 as shown in FIG. In the form with an English translation.

The items thus displayed can be printed by the printing unit 1 and the printing unit 2. Therefore, for example, the Japanese who is the listener even if the English conversation becomes long can listen to the new conversation while looking at the previously spoken English sentence and its translated sentence. This is true even when a foreigner hears a Japanese Japanese conversation or a Japanese Japanese conversation. In these cases, it is possible to save the trouble of collecting the minutes.

The present apparatus can be realized by connecting two book type computers, or can be realized by providing two display portions in one book type computer, and one computer and its It is also possible to realize a single display unit so that Japanese and foreigners can share it. In other words, although FIG. 1 is a functional block diagram of this device, when it is implemented as hardware, it may be as shown in FIG. In FIG. 22, the functions for Japanese and foreigners are put together.

[0012] The present apparatus is not limited to mere conversation, a large-screen TV the Japanese for the display unit 1, there is the stomach, and the projector screen, the Japanese partner, can also be used in foreign language lecture of foreigners. If the display unit 2 for foreigners is a large-screen TV or a screen, it can be used for foreign language or Japanese lectures by Japanese people with foreigners.

Furthermore this device foreign language audio tape, when applied to listening to foreign language broadcasts, would sell string like Figure 23 consecutively. In this case, since there is no insertion of blank by foreigners is a speaker, as a blank of blank string of the calling voice of the speaker, but it only appears, forte thing that the listener is read the English sentence If this is the case, it is expected that this will allow you to read the meaning of the other party. In addition, if the processing performance of the foreign language syntax processing unit 2 is improved, it will be displayed as a more readable foreign language. In this case, the method of non-specific speaker recognition will be used, but in foreign language broadcasting, etc., the voice input is expected to be close to the standard pattern, so there is no speaker intervention, but good results are obtained. I can hope.

[0014] without the use of foreign language syntax processing section 2, of course, in FIG. 2, Americans is a speaker, by the blank so needful pen input unit 2, on the display unit 2, as shown in FIG. 3 W And y and P are specified. Then, this character string is converted into a meaningful sentence as shown in FIG. You can do this .

[0015]  The blank pen input section is touched by the finger, that is, to the place touched by the finger. , A mechanism for automatically inserting blanks may be added to the display.

[0016] Incidentally, from FIG. 12, not further simply be Banora column Japanese is as translation as shown in FIG. 13, the Japanese syntax processor 1, so as to display mended the normal day Text Good . This ordinary Japanese sentence display is made in different colors or underlined in order to distinguish it from the raw Japanese sentence in FIG.

[0017]  In addition, as shown in FIG. 18 to FIG. 19, English as a translated word is simply a verbal sequence. Instead, the foreign language syntax processing unit 2 converts it into ordinary English and outputs it to the display unit 2. You may do so. In order to distinguish this ordinary English text display from the raw English text display of FIG. 18, the color is changed to Alternatively, it is output underlined by underlining.

[0018] The Japanese syntax processing unit 1 is also realized by software or the like.

[0019] Japanese that can read and write [devised effect] English is, by the conversation of foreigners is displayed, by reading the Re its, it is possible to grab the correct opponent of the senses. Also, by displaying the pronunciation of the English sentence, Japanese people can make foreigners understand the meaning even if they have a bad pronunciation. Even Japanese people who are not good at English can understand the meaning of the other person by displaying the English conversation of the other person with the Japanese translation, and the Japanese translation that I speak can be translated into English. Then, by being displayed, foreigners can understand the meaning. This translated word is displayed by the functions of the Japanese and foreign language electronic dictionaries, the Japanese syntax processing unit 1 and the foreign language syntax processing unit 2 .

[Submission date] December 13, 1993

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] 0005

[Correction method] Change

[Correction content]

[0005]

[Action]

The software of the foreign language syntax processing unit 2 for automatically editing the input character string “your products” into the “your products” described in this embodiment can be realized by the following logic, for example. First, the index file of the foreign language electronic dictionary 2 is index-searched with the word products. Then, an index search is performed with yourproduct. Then, do an index search with your product. Then, do an index search with your product. Then do an index search with yourprod. Then, do an index search with yourpro. Then, do an index search with yourpr. Then do an index search with yourp. None of these results in an error because READ cannot be performed. This is because there is no entry word corresponding to these. Then do an index search with your. READ is possible. This is because the entry word “your” exists in the electronic dictionary 2. Thus, your products can be edited back as your products. As another example, the following logic can be used. First, the index file of the foreign language electronic dictionary 2 is index-searched with the word products. Then, do an index search with our products. Next, an index search is performed with urproducts. Then, an index search is performed with rproducts. Then, an index search is performed with products. READ is possible. This is because the entry word “products” exists in the electronic dictionary 2. Thus, your products can be reedited as your products. In this latter example, READ can be performed by performing an index search on the misplaced “your” . Also, in the former example, READ can be performed by performing an index search with the products that have gone wrong . These indicate that the editing of as your products is complete. In more long voice string, longer phrase that went by cutting, so when that can not be READ and indexed over blood for that phrase, the same logic was applied, and the long voice strings, electronic dictionary It can be displayed in an easy-to-understand manner by separating each headword existing in 2 with a blank .

[Submission date] December 23, 1993

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0018

[Correction method] Change

[Correction content]

The Japanese syntax processing unit 1 is also realized by software or the like. A person's pronunciation is not always clear. In the actual situation, the voice character string “ ourproducts ” in this embodiment may be unclear, and the voice recognition device may recognize only “yourproduct” in some cases. <Is an address for unrecognized character strings. > Even in this case, according to the procedure used in the present embodiment, the characters are removed from the end of the character string one by one, and the index search logic is used. After removing it, you can replace it with a blank . > This voice character string can be re-edited into a voice character string that makes use of the word “ our ”, which is easier to understand . In addition, there may be a case in which the voice character string “ yourproducts” is recognized only as “ rproducts” because the beginning of the voice character string is unclear . In this case, the logic of performing an index search by removing each character from the beginning of this character string makes it possible to replace this character string with an easier-to-understand voice character string that uses the word of " products ", that is, " r products " .

[Submission date] December 28, 1993

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] 0015

[Correction method] Change

[Correction content]

The blank insertion pen input unit may be provided with a mechanism for automatically inserting a blank into a position touched by a finger, that is, a portion touched by the finger. People do not always speak sloppyly. On the way, there are things I can use words. For example, sometimes referred to as machine for the conver sa tion with foreign peole interrupted the words a machine for the conversation w ith foreign people in the middle of a conversation. At this time, a blank occurs between the cover and the sation depending on the break of the word . Conver and satio n in such a case or, on the other hand is not in the electronic dictionary, were stuck to each other when the <implemented in software> conversation is in the electronic dictionary, even if interrupted calling tone of the speaker, display of If, as a conversation, you can also use that output in the correct representation of one word.

[Submission date] December 29, 1993

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0017

[Correction method] Change

[Correction content]

Further, from FIG. 18 to FIG. 19, English as a translated word is not a mere word string, but the foreign language syntax processing unit 2 converts it into ordinary English and outputs it to the display unit 2. You may This ordinary English text display is output below the raw English text display in different colors or underlined in order to distinguish it from the raw English text display of FIG. In some words, even if the ending is unknown when the pronunciation of a few letters at the head is determined, the word is uniquely determined or is determined as one of a few words . For example, suppose that the pronunciation at the end of conversation is unknown, so the speech recognizer recognizes only the pronunciation equivalent to conversa. Then, since the word with this pronunciation is only conversation, this word is not available on the screen. It can also be used to display as <NOTE> or <Note> . This can be realized by software by index search of an electronic dictionary .

[Brief description of drawings]

FIG. 1 Functional block diagram

[Figure 2] Display example

[Figure 3] Display example

[Figure 4] Display example

[Figure 5] Display example

[Figure 6] Display example

[Figure 7] Display example

[Figure 8] Display example

[Figure 9] Display example

[Fig. 10] Display example

[Fig. 11] Display example

[Fig. 12] Display example

[Fig. 13] Display example

[Fig. 14] Display example

[Fig. 15] Display example

FIG. 16: Display example

FIG. 17 Display example

FIG. 18 Display example

FIG. 19 Display example

FIG. 20: Data structure of foreign language electronic dictionary

FIG. 21: Data structure of Japanese electronic dictionary

FIG. 22 is an example of a hardware block diagram.

FIG. 23: Display example

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] December 6, 1993

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be amended] Scope of utility model registration request

[Correction method] Change

[Correction content]

[Scope of utility model registration request]

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G10L 3/00 551 C 571 G

Claims (1)

[Scope of utility model registration request] 1. A voice input unit 2 for inputting a foreign language voice of a foreigner, a voice recognition unit 2 for recognizing the voice as a series of foreign language character strings, and a recognized series of foreign language character strings. On the display unit 2 that is displayed to the foreigner who is the speaker, at the same time that is displayed to the Japanese who is the listener, and on the display unit 2 by the operation of the foreigner who is the speaker, a series of A blank pen input section 2 that divides a foreign language character string into several more simplified character strings, an electronic dictionary 2 that is a foreign language word database, and a simplified
Some character strings are re-edited into meaningful foreign words based on the data in the foreign language electronic dictionary 2, displayed on the display unit 1 and the display unit 2, and further, the foreign language electronic dictionary 2 Using the, a foreign language syntax processing unit 2 for displaying a Japanese translation on the display unit 1, a voice input unit 1 for inputting Japanese voice, and a voice for recognizing the voice as a series of character strings. On the recognition unit 1 and the display unit 1, a blank pen for dividing a series of character strings uttered by a Japanese person into a plurality of simplified character strings by an operation of a Japanese person who is a speaker. The input unit 1, the electronic dictionary 1 that is a Japanese word database, and the simplified Japanese character string are edited back into meaningful Japanese words by the electronic dictionary 1, and further translated into foreign languages. Foreign language consisting of a Japanese syntax processing unit 1 and a central control unit that controls the whole Conversation support device.