JPH07219942A - Automatic kana/kanji conversion system - Google Patents

Abstract

PURPOSE:To accurately convert (substitute) an input instantaneously with the depression of a conversion key without generating any error (misconversion) when an. operator inputs 'HIRAGANA (cursive form of Japanese syllabary) input' or Roman character input' from a keyboard. CONSTITUTION:A character sentence dividing skeleton converts all characters in all sentences written in an initial purpose for a reference repository original into KATAKANA (square form of Japanese syllabary) characters independently of the number of characters and the total number of characters is divided into two parts. At the time of dividing the characters up to the unit' of each character, the dividing work ends. The divided reference repositories are integrated by a KANA character sentence integrating skeleton. The recognition of substitution between a KANA/KANJI (Chinese character) code mixed sentence to be an original for the reference repository and a reference repository preparing sentence obtained by converting the whole original sentence into KATAKANA characters and data preparing work are simultaneously executed.

Description

Detailed Description of the Invention

[0001]

[Industrial applications] Various PCs, workstations, minicomputers, office computers, mainframes, supercomputers, word processing machines, and other devices that are typed in romaji or hiragana from the keyboard can be used in kanji, hiragana, Katakana, numbers, various symbols,
It is a system that can be used for all kinds of computers that need to output or output to a Japanese sentence (text file) containing special characters, or to output on screen or print out.

[0002]

2. Description of the Related Art Conventional techniques and functions have been commercialized as a Japanese word processor by being used in combination with a kana-kanji conversion function and various Japanese word processing functions. Typical examples thereof include ATOK and Ichitaro from Just System Co., Ltd. Japanese word processor products are roughly classified into two types. One is application software for personal computers, for example, classification by software such as <Ichitaro>, <pine>, etc., and further, classification as word processor dedicated machine in which software and hardware are integrated. By classification, <Oasis>, <Bungou>, <Report>, <Panaward>, etc. already exist. The Japanese functions of existing product groups (generation of Japanese with kanji, hiragana, katakana, numbers, various symbols, and special characters) are generally the same, although there are subtle differences. Has been adopted. The outline of the function is as follows. ASCII code and Kanji (JIS 1st level, JIS 2nd level, etc.) code are input from the keyboard by alternately using the front-end processor originally developed by each company and the word processor editing / processing function alternately. It can be converted to kanji by corresponding to the characters that are

[0003]

Therefore, according to the conventional functions and techniques, when converting a large amount of Kana text input from the keyboard into Japanese mixed with Kana-Kanji, there is no single mistake (misconversion) and all You cannot instantly and automatically convert the characters in. Furthermore, there is a limit to the number of characters that can be converted and input at one time. Kana Kanji, which is the present invention
The automatic replacement system has not been realized by anyone yet.
That part (

), And its peripheral and spillover parts
Issues, multiple claims, all of these issues
Is realized.

[0004]

The detailed description of the automatic kana-kanji replacement system in this patent application is not intended to describe what is programmed as software. The detailed description of the present invention is necessary for establishing and acquiring the right to understand and confirm the right, such as discovery / invention of a principle, discovery / invention / development of a new function, which is indispensable for constructing an automatic kana-kanji replacement system. It describes system outline, system detailed design, and outline for programming, requirement definition, detailed design, etc. * After that,
The name of the Kana-Kanji automatic substitution system is used as this system, and when it is expressed as this system,
References should be made to all of the system outline / system detailed design document and the programming outline / detailed design document. This system is installed on a personal computer such as <Ichitaro> or <Matsu> that is generally said, so that it can exhibit the same meaning, function and role as the application software used. The system is built. However, this system realizes a function that is different from the performance that cannot be realized by conventional word processor software and dedicated word processor. When the Kana-Kanji conversion is performed, the performance that cannot be realized by the conventional functions and technologies is "Kanji, Hiragana, Katakana, numbers, etc. that the person who entered the keyboard wants to convert or must convert. One mistake (the wrong conversion) at the moment when you hit the conversion key to the alphabet character
It is possible and can be converted without any restriction on the number of characters that can be converted with one input. " Also, the functions and technologies different from the conventional functions and technologies are word processing function, single sentence registration function, dictionary function, dictionary and Sentences that the operator intends, wants, is ordered, and has a mixture of kanji, hiragana, katakana, and alphabetic characters by using functions such as registration function and example sentence registration function together with learning function and inference function. Convert to However, if the sentence to be converted is a long sentence of 100 characters, 1000 characters or more, for example, it is rarely completed at once, and it is indispensable to fine-tune and confirm each time. But,
In this system, functions such as word registration function, simple sentence registration function, dictionary function, dictionary registration function, example sentence registration function, and learning function, inference function, etc. are not adopted at all.
Developed a replacement system, not a conversion system. The operator inputs <Hiragana> or
If you enter it with the key input>, the moment you hit the conversion key,
Accurate conversion (replacement) without one mistake (misconversion)
Wear. Moreover, this system uses the keyboard to input
1000 sentences, even if the sentence is 100 characters
However, in the replacement ability, even a single character is wrong.
No conversion occurs. Therefore, this system is considered to have a function for text editing, which is composed by freely assembling, copying, moving, and reducing so-called sentences and clauses. Absent. But,
The functions for editing characters, which are formed by rearranging, rearranging, and assembling characters, are well considered. Also, as with so-called word processor software or word processor dedicated machines, it has a function to store characters that have been input as characters, converted to Kana-Kanji characters, and edited as sentences (created document files). Not considered in the first place. However, it is easy to add as a function and there is no problem. Furthermore, man-machine interface functions such as the so-called Windows function have not been considered. However, it is easy to add as a function,
There is no problem. Furthermore, it does not consider advanced editing / printing / publishing / bookbinding functions such as DTP (Desktop Publishing). However, it is easy to add as a function and there is no problem. However, the function to display the created sentences on the screen of the personal computer (computer), the function to print out the characters on a specific printer, and further, from the dedicated word processor, from the personal computer to the engineering workstation, It has a data generation function for text files (also called ASCII files) that can be read by minicomputers, office computers, even mainframes and supercomputers. This system is always developed and equipped in all word processor software and word processor dedicated machines, such as FEP (front end processor), ASCII code and kanji code (JIS first level, JIS second level, etc.). ) Is used together with a kana-kanji conversion function (for example, ATOK installed in Justsystem Ichitaro Co., Ltd.), which can be used together and used together. However, those functions do not have to be ATOK by all means, and all word processing software dedicated to personal computers and even word processing dedicated machines have their own FEP and kana-kanji conversion functions. It is assumed that it can be used for. In addition, this system does not independently develop and equip these functions, because these functions can be effectively used as ready-made products, so FEP and
The Kana-Kanji conversion function is only a provisional form of use in the sense that it saves time and development costs for new development. "Of course, this system is a product
One condition for being sold in the general market is:
Considering that there will be discussions / contracts with companies that have, have, or have the right to use the FEP to be used or the Kana-Kanji conversion function. However, if the FEP and the kana-kanji conversion function, which are the additional development items of this system, are developed, it is a fact that all of the above problems (the part enclosed by "") will be solved. And the solution is that it can be developed in a very short time.
It is a well-known fact. Moreover, the use of FEP and ASCII codes and kanji codes (JIS 1st level, JIS 2nd level, etc.), which are indispensable for the development of kana-kanji conversion function, with no permission does not cause any problems. It is also a fact.

[0005]

[Function] This system is based on the discovery, verification and evaluation of the following principles.

The solution of various problems described in
It is realized by the following actions. This system, for example,
Like an OS such as MS-DOS and UNIX, it exists and operates in a form that resides in the main memory of each computer. The main program created to operate this system works in a form in which processes (execution procedures) and data generated by the processes are always mixed. Moreover, when data is newly generated, not only the data but also the process that generated the data has a function (proliferation function) that is automatically and newly described in the main program. ing.

[0006]

【Example】

 <

[Claim 1] Explanation of the principle of the Kana-Kanji automatic substitution system and its discovery> When humans write sentences, are they really creating existence from nothing? This is the starting point for the idea of developing an automatic kana-kanji replacement system.
When a human writes a sentence, he or she rarely thinks and creates a new one, and most of the time he or she unconsciously recalls and reassembles a sentence that has already been experienced and remembered. It is an idea that there is not. Of course, the idea of what to write is definitely the original of the person who writes the text. But,
The act of writing and the act of actually writing are distinctly different. So, for example, it is described in the Japanese language textbook of the third year of middle school,
If all sentences can be subdivided according to a certain rule *, they can be made into a database, and that database can be made to reside in a computer as a basic repository (a treasure trove of knowledge), the whole sentence can be written in katakana characters. The idea is that even if input, the used Kanji and sending kana patterns can be regularized in a considerable part. For example, if the size of the basic repository is the entire Japanese literature collection, the probability of regularization is 100.
Perhaps it will be close to%. Furthermore, when the corresponding sentence structure does not exist in the basic repository,
If the function of automatically accumulating the sentences as new experience data can be added, the basic repository will be a proliferation type repository, and the probability that it can be regularized should approach 100% without limit. In addition, as a result, the personality of the person who inputs the character data is reflected, and each Kana-Kanji automatic replacement system behaves as if it had its own personality. This is the beginning of the development and realization of the Kana-Kanji automatic substitution system. <

1. A concrete embodiment of an automatic kana-kanji substitution system>. 1, Kana character sentence division skeleton (

(2) Kana character sentence division skeleton) The initial goal of the prototype of the repository is a Japanese language textbook in the third year of middle school. No matter how many characters are written in the textbook, all of them are made into katakana characters, and the total number of the characters is divided into two. At the end, the division work is completed when it can be divided into units of one character. The following description is of a system that can automatically create a repository by entering the entire text of "the textbook of Japanese language in the third year of middle school", which is the initial goal of the prototype of this repository, as a normal work from the keyboard,
References to development procedures. Here, one problem occurs. The fact is that even numbers can be divided, but odd numbers cannot. So I created one formula. * Furthermore, in the Kana-Kanji automatic substitution system, the absolute condition for repository creation is that the text to be a repository must be divided into the total number of characters, and at the end it must be a set of one character. The formula is as follows. For example, suppose you have a sentence that says "Spring is coming, flowers are blooming." When converted to full-text Katakana characters, the total number of characters is "Harugakita, Hanagasaku", 10 characters (dots and circles do not count). In this case, at first, automatic division is performed based on the even-text division skeleton formula. At the moment when it was divided into Harugakita and Hanagasaku, this sentence changes into an odd sentence. Therefore, this time, the odd-numbered sentence division skeleton is automatically selected and further divided automatically. Then, it changes into an even sentence again. So again,
The odd sentence division skeleton is automatically selected,
Automatically split. And at the end, Harugakita and Hanagasaku. The whole sentence of 1
As data, the following formula is established. The basic repository generated in this system by the above odd sentence division skeleton and even sentence division skeleton is a proof of the axiom that "the whole is a part and the part is a whole", and at the same time, n = 1, 1 + 1 = 1
It is also a proof of the theorem, and further recognizing only real numbers without recognizing imaginary numbers, recognizing both <one> 1 and both imaginary numbers and real numbers, and demonstrating that <pair> 1 coexists. is there. * <Mathematical proof of n = 1, 1 + 1 = 1> To prove n = 1, for example, 666 = 1 can be proved. The proof is as follows. However, 1 is 1 <only 1> of <real number>, and i (1) is 1 <pair of 1> of <imaginary number>. Therefore, the number in parentheses is the imaginary number of i. Based on the above formulas and proofs, a basic repository of mixed kanji that is subdivided under certain conditions with certain regularity <A>
Is a repository <a> obtained by converting <A> into only full-text Katakana characters, and a subdivided version of it, which is the same as <A>, with a certain regularity and under certain conditions. If so, repository <A> and repository <a
> Is similar, and it has been proved that it is possible to completely replace a sentence with only full-text Katakana characters and a sentence with mixed Kanji and symbols. <

7. A description of a similar-shaped circumferential value / indexing function>. The axiom "the whole is a part and the part is a whole", the formula n = 1, 1 + 1 = 1, and 1 is the only one that recognizes only imaginary numbers and only real numbers. Then, by the theorem that both imaginary numbers and real numbers are recognized and that <1> of <pair> exists and functions, the formula of “well-known and indexing of similar forms” can be created as follows. Therefore, the above formula of "Awareness / indexing of similar shapes" in □ was first created by the inventor <Akira Sugiyama>.

7. A similar patented peripheral value / indexing function> is declared and recognized as the most important patent application of this system. <Second function of Kana-Kanji automatic replacement system> 1. Kana character sentence integrated skeleton <

5. A kana character sentence integrated skeleton> The kana character sentence divided skeleton divides the basic repository divided under a certain condition with a certain regularity, and this time is integrated by the kana character sentence integrated skeleton. . At this time, the replacement sentence between the sentence containing the Kana-Kanji symbols, which is the original sentence of the basic repository, and the sentence for creating the basic repository in which the sentence is converted into full-text Katakana characters, and the work for data generation are performed at the same time. . Where <

6. The basic principle for kana character / kanji character replacement recognition realized by the kana character / kanji character recognition function> will be described. <Principle 1> Even if there is a difference in the number of data between the original sentence mixed with Kana-Kanji symbols and the full-text Katakana character sentence, the data running time must be the same. A. Hahalurugaki Kita, Hahana Nagasasaku. a. Spring has come, flowers are blooming. * A has 16 characters and a has 8 characters. However, the execution time is the same. <Principle 2> This explanation is <

8. A character ← → note replacement function> will be described. In this system, we have discovered the practical use of the principle of musical notes as a method for recognizing kanji, hiragana, katakana, symbols, etc. in a sentence in which kana-kanji symbols are mixed in full-text katakana characters. That is, 1 byte in a computer is 8 bits, and the number of bits forming a katakana character or alphabetic character is 1 byte of 8 bits. In addition, the number of bits required to convert a kanji into kanji is 16
Bits, 2 bytes. This relationship is exactly the same principle as the relationship between whole notes and eighth notes in musical notes. Whole note has the meaning and function of 1 byte. And the eighth note has the meaning and function of 1 bit. Therefore,
The meaning and function of having eight eighth notes are the meaning and function of the presence of one whole note. It was the discovery and creation of <Akira Sugiyama>, the inventor of this system, that this principle could be applied to make the so-called word processor Kana-Kanji conversion function epoch-making. Moreover, regarding the relationship between whole notes and eighth notes in music, for example, if you engrave a whole note once per second, an eighth note will be eight times. Ticking a whole note twice means to tick an eighth note 16 times. This is exactly the same as the relationship between bits and bytes of a computer. The first principle of this system is that the original sentence mixed with Kana-Kanji symbols and the full sentence Katakana character sentence, even the number of data (number of characters) Even if there is a difference in the number (quantity) of data, the time to run the number of data must be the same. In fact, the relationship between musical notes (sound length), unit beats (sound beat rate), and note value (elapsed time) is established on the assumption that this principle is adhered to. Moreover, this fact is exactly the same principle as the relationship between bytes and bits in a computer: bit length, number and elapsed time. Therefore, this system

6. The function principle of kana / kanji recognition function>, and the following <

8. All the uses and uses derived from the principle of the function in the kana character ← → note replacement function> are recognized as the most important patents of this system, and all of them are the inventors of this system < Sugiyama Akira>'s discovery, which declares and recognizes that it is creation. -Replace the relationship between characters and kanji with the relationship between notes, unit beats, and note values. Due to the above facts, this system is

[Claim 9] Japanese voice ← → note replacement function> and <

[Claim 10] The Japanese voice ← → character replacement function> is also recognized as the most important patent application of this system, and all of these are the discoveries of <Akira Sugiyama> who is the inventor of this system. It declares and recognizes that it is creation. <Principle 3> In this system, the replacement of katakana characters with kanji has already been explained, but how is the recognition of replacement of katakana characters, hiragana characters, and even katakana characters performed? Is it? A. Full-width recognition is performed from katakana characters to hiragana. B. For katakana characters, add one character of katakana to <. > To <. > Half-width, and then <. It is carried out with the recognition that the> is added. The method of> is called a note point in musical terms, and plays a role of halving the note value of the corresponding note. -The following is a sample example of the above explanation. B. C. Le. Gakikita, Ha, Na. Gasasaita. b. Hull came, Hana bloomed. * <Ha> is <ha. Recognized by ＞. Even if the amount of data is different, the elapsed time is the same. Based on <Principle 1> to <Principle 3>,

6. A kana character sentence integrated skeleton, which corresponds to a partial function / principle of kana character / kanji recognition function>, will be described.
The sentence for creating the basic repository "Spring has come, flowers bloom." Has already been converted to full-text Katakana characters, and the Katakana character (sentence) split skeleton allows you to read "Hall gakitahana". Gas "
Is divided into 1 character units. The integrated skeleton adds the recognition of Chinese characters and the like to each character of the Katakana character in the process of returning the divided character to "Hargakita, Hanagasaku."・ First, D0. To A0. The situation of returning to is presented as a sample example. (* The result of the division from A0. To D0. By the division skeleton) * DO. At the time of, the sentence to be divided was divided into one character <
Time> and <place>, and the <time> and <place> are the beginnings of integration. * <Time> and <Place> where the text to be divided is divided into one character is the beginning of integration. ・ D10 line is one character group. -Line C10 is a mixed group of two character groups and one character group.・ Line B10 shall consist of 5 characters.・ Line A10 is a group of all characters. From the D10 line to the A10 line, the character group in each line is in the state of "the part is the whole and the whole is the part" which is the principle of <n = 1>. Because, in each line, if even one of the data that composes the line is missing, the meaning of the sentence on that line cannot be understood. In other words, the data is in a state of being destroyed or not existing. Work from one character group on line 1, D10. Each character of <Hall Gakita Hanagasa>
Perform substitution recognition on the sentence "Spring has come, flowers bloom." To create the basic repository. Then 1, 2,
Be prepared to record history times such as 3, 4, and so on. This record of historical time is continued as long as this repository exists, and is an absolute time in this system. 10C is not recognized because it is not in the basic repository. 20 le is not recognized because it is not in the basic repository. The 30-ga is recognized as the hiragana <ga is>. 40-ki is recognized as the Kanji <Kiki>. 50-ta is recognized as the hiragana <ta-ta>. 60ha is not recognized because it is not in the repository. 70na is not recognized because it is not in the repository.
80-ga has been recognized as <ga> in 30 hours. 90sa is recognized as the kanji <Sakusasa>. A0-ku is recognized as the hiragana <kuku>. 2, the work moves to a mixed group of two character groups and one character group of C10 line. A1 hull is recognized as the kanji <Haruhaharuru>.
A2 moth has been recognized as <ga mo> in 30 hours. A3 Kita is recognized as <Kita Kita> at 40 and 50 historical hours. A4 Hana is recognized as the Chinese character <Hanahahana>.
A5 moth has been recognized as <ga mo> in 30 hours. A6 Saku is recognized as <blooming Saku> in history time 90 and 10. Work shifts to the 5 character group of line 3 and B10. A7 Haruga Kita is recognized as <Haharu Ruga Kiki Ha Haruru Gaki Kitata> in the history time of 11, 30, and 13. The A8 flower is recognized as <flower blooms <Hahana Nagasaku> at history times 14, 30, and 16. 4. The work moves to the whole sentence group of line A10. A9 Haruga Kitahana Nagasaku is recognized as <hahalurugakiki hahalurugakikitatahahananagasaku where the flowers have come to spring> at the time of 17th and 18th. This is the end of the work of the Kana character sentence integrated skeleton to make "Spring has come, flowers bloom." The history time is recorded in 19 steps from 10 to A9, and if the unit of time is 1 second, it is said that the vehicle will run in 19 seconds. However, the real time is only 7 seconds. Furthermore, the actual time is only 7 steps of 30, 40, 50, 90, A0, QA1, and A4. The rest is virtually processed as virtual time. * Definition of tokens * -History time: The recognition of the history of time and place when a certain work was performed.
It is usually said to be experienced. -Validity time: The recognition of the confirmation of the history that no work is done, but that work has been done. I usually say that I have experienced, I have done it.・ Actual time: It is the time when actually working. Usually called actual labor. -Virtual processing: Processing that can be performed without remembering the procedure or collecting data to perform the work. I usually tell you to skip it because you know it. A. Also, the new repository accumulated by the work of creating the basic repository this time is as follows when looking at the program execution status. 10 20 30 ga is <ga mo> 40 ki is <coming kiki> 50 ta is <ta ta> 60 70 80 90 90 sa is <sakusa> A0 Ku is <kuku> A1 hull is <spring haharuru> A2 A3 kita <Kikita came> A4 Hana <Flower Hahana> A5 A6 Saku <Blooming Saku> A7 Hurgakita <Harurugaki Kitata when spring came> A8 Hana Nagasaku <Blooming Hahana Nagasaku> A9 Hargakita Hanasaki <Spring has come and flowers bloom Hahalurugakiki Hahalurugaki Kitatahahana Nagasasaku> B. Looking at the execution status of the Kana character sentence integrated skeleton, it is as follows. D. It is as follows when it is resident in the main memory and is running. 102030 is <ga> 40
Ki <Kiki> 50 ta <ta ta> 60708090 90 sa <sasaki> A0 haku <ku> A1 hull <spring haharuru>
A2A3 Kita <Kita Kita> A4 Hana <Flower Hahana> A5 A6 Saku <Blooming Sasaku> A7 Harga Kita <Spring is coming Haharuru Gaki Haharuru Kitaki> A8 Hanagasaku <Flower Hahana Nagasaku> A9 Halga Kitahana Nagasaku is <Haharurugakiki Haharurugaki Kitahahana Nagasasaku where the flowers come in spring>. * Naturally, the above programs are resident in the main memory. Furthermore, at the end of the main program, in order to confirm that one work has been completed, be sure to check <. Hit ＞.
If you type it, wait for the input of kana characters for the next basic repository creation. E. Here's a test using the first generated base repository: -Enter "Hurga, Saku." And "Spring will bloom."
And generated in a text file.・ Enter "Kita. Hurga," then "Came. Haru ga."
Is output.・ If it is "Sakuhanaga", "Blooming flower" appears on the screen.・ If you hit "Haru, Haru, Harugasaku. Saita, Saita, Harugasaita.", There is no single character error (misconversion) "Spring, spring, spring blooms.
And, instantly fall into the text fill. * To replace Kana-Kanji in the repository, enter the text "Saku. Hanaga," or "Saku. Hanaga," or "SAKU. HANAGA," and the skeleton will be <,>. And <. Since> is not recognized, <,> and <. After the substitution,> is output as “blooming. Flower”, and is generated as if the integrated skeleton recognized and hit it correctly. And at the end of the generated basic repository <. Wait for the> to be typed and the text for creating the next basic repository to be entered from the keyboard. Naturally, when a new Kana character is input, it is mechanically processed by the split skeleton and the integrated skeleton in the same manner, and a basic repository is additionally generated. At the same time, the additional repository is connected after the historical number A9, and is instantly assimilated as one main program and resides on the main memory as a new program. As a result, the main program and the repository are expanded. This system is
Basically, from the function of the Kana character sentence division skeleton,
All operations up to the kana character sentence integrated skeleton are automated, and when the operator inputs sentences from the keyboard at any time, the basic repository automatically accumulates, and on the surface it seems that the repository itself is proliferating. Looks like

[0007]

[Effects of the invention, 1] Invented and discovered in the process of developing this system.

9. A description of Japanese voice ← → note replacement function>,
Describe the effect. By replacing the 48 katakana characters starting with <a> and ending with <n> on the keyboard of the piano, each character of katakana can have a sound. For example, replace <A> with the sound of “do” in the lower 1-octave and <i> with the sound of “Le” in the lower 1-octave, and so on. It is the sound of "La" in the octave. further,
In this system, the katakana characters were replaced by the phonetic notes, and the other note names on the keyboard of the piano were directly replaced by numbers. For example, <A> is the first sound, A is the second sound ... N is the 48th sound, and so on. Replacing the above conditions with 48 Katakana characters beforehand yields the following: * The description format of this table and the creation of this table are

9. In Japanese voice ← → note replacement function>,
<

Claim 14. Recognizing Japanese voice ← → Pitch name substitution table>,
All of them are discoveries of <Akira Sugiyama> who is the inventor of this system, and they declare and recognize that they are creations. Due to the above facts,

9. Japanese voice ← → note replacement function>

8. In combination with the Kana character note replacement function>, in this system, Japanese description can be written on a staff in musical notation format. Furthermore, the Japanese written on the staff is the same as the music when it became music.
-I and CD-ROM technology can be used to enable amazing data compression, and the result means the possibility of ultra-high-speed generation of the basic repository of this system. Furthermore, it became clear that the system, function, and possibility of voice input and voice output would change completely. Normally, if you speak Japanese using Kana characters into a microphone and import it into the Kana-Kanji automatic substitution system,

[9] Japanese voice-> note replacement function> and <

8. A combination of the Kana character note replacement function> enables automatic Kana-Kanji replacement without using a keyboard, and the number of characters that can be converted by one input is "one mistake (misconversion)". It can be converted without any restrictions. ”Voice → Kana characters → Sentence → Voice replacement system can be realized. In this system, this fact

[Claim 13] Voice ← → Kana / kanji mixed sentence replacement function>, all of which are discoveries of <Akira Sugiyama> who is the inventor of this system, and those who declare and recognize creation Is.

[0008]

[Effect of the invention, part 2] Invented, discovered and created in the process of developing this system <

9. Japanese voice ← → note replacement function> is as it is by adding the following data (Table 1) to 48 Katakana characters.

15. English voice ← → note replacement function>, and further <

14. The Japanese voice ← → Pitch name substitution table> is as it is.

16. It is also clear that English voice ← → Pitch name substitution table>, and <Voice ← → Pitch name substitution function> in languages of countries all over the world. This function

11. An alphabetic character ← → Kana character replacing function>, <

12. Alphabet character voice <-> Kana character voice replacement function >>.

16. A list in English phonetic ← → phonetic name replacement. * The description format of this table and everything related to the creation of this table are the discoveries of <Akira Sugiyama>, the inventor of this system, and they declare and recognize that they are creations.

[0009]

EFFECT OF THE INVENTION (3) If the voice input allows the voice to be converted to Japanese with Kana-Kanji mixed, there is no single erroneous conversion, and there is no restriction on the number of inputs to be converted. , It becomes possible to build the following system. At present, a CPU is built in a television currently on sale, and this system is incorporated in the CPU. By incorporating this system in the CPU, we have created a repository that documents various command conditions. Enable audio input / output for the repository. It is clear that the development of these systems cannot be a difficult development by utilizing this system. Then, the telephone function is added to the television. In other words, the following functions are standard equipment as main functions of the TV. CPU (for example, OS) with various command condition document repositories built in this system & this system
-9000 & 80386) + TV + TEL (cordless compatible home PBX) + CD-ROM or a system equipped with CD-I as standard equipment is developed. In addition, the TV will be equipped with the following functions as options. Facsimile, value-added telephone function, still image TV telephone, satellite broadcasting tuner, non-intelligent terminal type TV, PC communication compatible, TEL wiring use twisted cable / home LAN compatible, keyboard (wireless type / wired type), printer, etc. * The contents of various command condition documents are as follows. <Regarding life basic data> Electricity usage / charge Gas usage / charge Water usage / charge Telephone (information) usage / usage / charge <About living maintenance data> Air conditioner usage / power usage Cooler usage / power Usage amount Oil / gas fan heater usage status Washer usage status Clothes dryer usage status Rice cooker usage status Oven usage status Microwave oven usage status Automatic dishwasher usage status Automatic hot water heating / heating system usage status Refrigerator / refrigerator usage status TV & CM viewing status Radio viewing status Video recording viewing status Stereo usage status CD / LD usage status Satellite broadcasting viewing status Wired broadcasting viewing status <Regarding life credit and liability data, Part 1, Bank usage details> Payment status Checking at home Checking status at home Transfer status At-home confirmation Automatic withdrawal status At-home confirmation Reserve status At-home confirmation Debt Status Confirmation at Home <Regarding Living Receivables and Liability Data, Part 2, Financial Usage Details> Payment Status Confirmation at Home Investment Status Confirmation at Home Receipt Status Confirmation at Home Borrowing Status Confirmation at Home <Regarding Living Claims and Liability Data, Part 3, Public Usage Details> Payment Confirmation Receipt Confirmation Confirmation Confirmation Confirmation Confirmation Confirmation <Regarding Living Credit and Debt Data, Part 4, Other Usage Details> Co-op usage / payment confirmation Newspaper usage / payment confirmation NHK usage / payment confirmation Wired usage / payment confirmation <living data collection / storage function Various kinds of life forms that can be created according to> TV program viewership rate <weekly / monthly / annual> Trend performance report TV CF viewership rate <weekly / monthly / annual> Trend performance report Video viewing <weekly / monthly / annual> Trend Record Form Video Recording <Weekly / Monthly / Yearly> Trend Record Form Electricity / Gas / Water Usage Fee <Weekly / Monthly / Yearly> Record Form Electricity Product-specific use trends and frequency data <daily, weekly, monthly year> document phone use of specification & Price <month-year> trends proven form credit drawdown <month-year> Projects form household accounts and statistical / analysis graph

[Brief description of drawings]

FIG. 1 is an overall system diagram of an automatic kana-kanji replacement system.

[Procedure amendment]

[Submission date] December 22, 1994

[Procedure Amendment 2]

[Document name to be corrected] Drawing

[Correction target item name] All drawings

[Correction method] Change

[Correction content]

[Figure 1]

Claims (16)

[Claims] 1. A kana-kanji automatic replacement system 2. A kana character sentence division skeleton. 3. An odd sentence division skeleton 4. An even sentence division skeleton 5. A sentence integrated skeleton 6. A kana / kanji recognition function 7. A similar peripheral value / indexing function 8. A kana character ← → note replacement function [Claim 9] Japanese voice ← → note replacement function [Claim 10] Japanese voice ← → Kana character replacement function 11. An alphabet character ← → Kana character replacement function 12. Alphabet character voice ← → Kana character voice replacement function [Claim 13] Japanese voice ← → Kana / Kanji mixed sentence replacement function [Claim 14] Japanese voice ← → Pitch name substitution table 15. English voice ← → Pitch name substitution function [Claim 16] Japanese & English voice ← → Pitch name substitution table