JP3533591B2 - Character input device, character input method, and recording medium recording character input control program - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character input device for indirectly inputting Japanese, a character input method, and a recording medium recording a character input control program.

[0002]

2. Description of the Related Art Conventionally, when inputting a Japanese sentence, a kana-key layout or romaji-key layout keyboard is used to generate kana in a one-to-one relationship, and based on this, a kana-kanji conversion dictionary is referenced. I am creating a sentence mixed with kanji.

On the other hand, in recent years, small input terminals such as electronic notebooks and notebook computers are being used, and while miniaturization is required as the outer shape from the viewpoint of portability, input is performed from the viewpoint of ease of input. There are conflicting requirements that the keys should not be too small. For this reason, keyboards with a reduced number of input keys have been devised. For example, those disclosed in JP-A-6-102979, JP-A-6-21699, JP-A-7-28577, and JP-A-8-221170 are known. Multiple characters may be assigned to one input key by reducing the number of input keys, but even in that case, you can specify the kana character to be input by using it in combination with the shift key or increasing the number of keystrokes. There is.

On the other hand, in the technique disclosed in Japanese Unexamined Patent Publication No. Hei 8-211987, a plurality of characters are assigned to one input key,
Among the character combinations corresponding to the operation of the input keys, the registered ones are compared with the entry words in the dictionary file and output.

[0005]

The conventional character input device in which a plurality of characters are assigned to one input key has the following problems. First, in the former group that also uses shift keys, etc., the number of input keys to which letters are assigned is reduced, but when operating the 1 input key, the user may or may not input with the shift key. Since it is necessary to distinguish between and, the input work becomes complicated.

Further, in the latter case of referring to the entry word of the dictionary file, a plurality of character strings are generated from each combination of the characters assigned to the input keys, and these character strings are registered in the dictionary file. It is displayed as a candidate character string, provided that it is present. Therefore, the more words are registered in the dictionary file, the more candidate character strings are displayed, and the more complicated the candidate selection work by the user.

The present invention has been made in view of the above-mentioned problems, and does not increase the input work as in the former group and does not make the selection complicated as in the latter group, and reduces the number of input keys. An object of the present invention is to provide a character input device, a character input method, and a recording medium on which a character input control program is recorded.

[0008]

In order to achieve the above object, the invention according to claim 1 has a plurality of input keys and a Japanese input element based on a combination of vowels and consonants. Key input means that allows duplicate assignment when assigning to each of the input keys, and the Japanese input elements assigned to the input keys are assigned to the input keys when inputting Japanese by referring to the dictionary. Conversion of a character string that can occur in a Japanese input element into a Japanese kana character string, and
And a Japanese conversion control means for the conversion of the second stage to convert the kana character string in kanji mingled string of Japanese, the
Is the Japanese translation control means the result of past translations?
Before corresponding to the finally decided kanji mixed character string
Whether to store the kana character string in the conversion in the first step
A row arrangement storage means and the kana row storage means
Regarding kana character strings, priority in the first conversion
It is configured to include a first conversion unit that increases the degree of conversion and performs the conversion from the next time onward .

In the invention according to claim 1 configured as described above, in the key input means having a plurality of input keys, Japanese input elements based on a combination of vowels and consonants are assigned to each input key. However, duplicate allocation is allowed at that time. In this case, there is no way to select one by using the shift key, etc., and it is duplicated in a pure sense. Therefore, when inputting Japanese with the Japanese input element assigned to the above input key as a premise, there can be multiple Japanese, but at this time, the Japanese conversion control means converts the Japanese input element to a kana character string.
Is finally determined from the past conversion results in the conversion of
Kana character strings corresponding to mixed Kanji character strings are given priority.
One, to convert to the combination of a natural Japanese.

That is, since the Japanese input elements are redundantly assigned to the respective input keys, a plurality of combinations may occur if the Japanese input is performed on the premise of the assigned Japanese input elements. In this case, since the Japanese input elements are based on the combination of vowels and consonants, it must be able to be converted into kana in the first step, and even if such kana can be generated, There are things that make sense and things that do not. Furthermore, because the kana itself in Japanese is an indirect input element for inputting a sentence containing kana-kanji, the Japanese conversion control means refers to the dictionary and the kana Convert to a string and perform conversion in the second stage so that it becomes a character string mixed with Kanji that is appropriate for Japanese . Moreover, at this time
The kana character string is learned in one step .

As described above, when Japanese is input based on a Japanese input element based on a combination of a vowel and a consonant , a Kana character string is input from the Japanese input element and a Kana character string is input from the Kana character string.
A two-step conversion of mixed character strings is performed, and in the first step, it is converted in the past. It should be noted that it is possible to output the converted Japanese one by one, or it is possible to output a plurality of candidates as a group.

As described above, according to the present invention, when the Japanese is input using the key input means in which the Japanese input elements based on the combination of the vowel and the consonant are assigned redundantly,
It is converted into Chinese characters mingled with a string of Japanese over a two-stage
It Therefore, it is possible to provide a character input device capable of reducing the number of input keys and efficiently reducing the number of combination candidates without deteriorating the operability.

Various methods are used for inputting Japanese from the key input means, and one sound is divided and input like a Japanese input element based on a combination of vowels and consonants. As a preferred example, in the invention according to claim 2, in the character input device according to claim 1, a Japanese input element based on a combination of a vowel and a consonant assigned to the key input means is in Roman characters. In addition, the Japanese-language conversion control means is configured to generate a combination of kana by performing romaji-kana conversion in the first-stage conversion.

The key input means assigns Japanese input elements corresponding to vowels and consonants as Roman characters to each input key, and the Japanese conversion control means inputs the input when the key input means is operated. Romaji-kana conversion is performed in the first stage conversion to generate a kana character string based on the Japanese input element of romaji assigned to the key. Therefore, according to the second aspect of the present invention, the conversion can be performed using the Roman alphabet input that is familiar.

The vowels and consonants referred to here are not necessarily limited to the Hebon-type Roman letters, but may be understood as a broad sense in which a single kana is generated by a combination operation of input keys.

As a concrete example of increasing the priority of the conversion in the first step , the invention according to claim 3 is the character input device according to claim 1 or 2, wherein the first conversion means is used.
Is the Japanese input as the conversion with the higher priority.
Of the kana character strings that can be generated by the element, the kana sequence
Second stage with Kana character strings stored in storage means as candidates
The result of the above conversion is preferentially displayed .

In the invention according to claim 3 configured as described above, a sentence that can be generated by a Japanese input element
Kana character string stored in the kana sequence storage means among the character strings
With the result of performing the second stage conversion as a candidate
To display. I want to get the same conversion result in the same sentence
In many cases, the user sees this display and changes in the past.
You can immediately select the converted mixed Kanji character string.
It

As a preferred example for creating a sentence mixed with kana-kanji, the invention according to claim 4 is the character input device according to claim 3, wherein the kana sequence storage means is a kanji sequence of Japanese and its kanji sequence. A kana-kanji conversion file that stores a kana sequence corresponding to reading is searchable, and the kana sequence is stored.

In the invention according to claim 4 configured as described above, the kana sequence storage means has a kana-kanji conversion file, and stores the kanji sequence of Japanese and the kana sequence corresponding to the reading. Therefore, a character string that is natural in Japanese is converted as it is to a character string that is mixed in Kanji that is natural in Japanese. Therefore, according to the invention of claim 4, the kana-kanji arrangement is also stored based on the kana sequence, so that efficient kana-kanji conversion can be performed.

The Japanese conversion control means need not be limited to such conversion , and various techniques for improving the efficiency of this conversion can be adopted. As an example thereof, the invention according to claim 5 is the character input device according to any one of claims 1 to 4, wherein the Japanese conversion control means mixes the converted kanji characters at the conversion in the second step. letter
And the candidate generating means for generating a column as Japanese candidate, the strange
Display means for displaying the replaced Japanese candidates and one candidate selected from the Japanese candidates displayed by the displaying means are used.
Selection control means for selecting based on the user's operation , and the selection result by the selection control means is given priority during the next conversion in the second step.
And learning control means to be converted .

In the invention according to claim 5 configured as described above, the candidate generating means performs conversion as described above.
When the Japanese candidates are generated, the display unit displays the converted Japanese language candidates. When the operator selects one of the Japanese candidates displayed by the selection control means, the learning control means prioritizes the selection result by the selection control means at the next conversion . For this reason, then enter the same Japanese
If so , not only kana character strings but also kanji characters
Also in the row, the previously selected kanji-blended character string is preferentially targeted. Therefore, according to the invention of claim 5, the result selected from a plurality of combinations is learned, so that the conversion becomes accurate.

The priority target is a target to be preferentially selected in the next selection, and includes various modes.
For example, a mode is included in which the conversion result for the time being is displayed, or a head of the conversion result is displayed when a plurality of conversion results are displayed. In addition, as a specific method for giving priority to such a method, a method of writing the selection result in a learning file and referring to the learning file before referring to the dictionary file, or setting a flag in the dictionary file Various methods such as a method of writing can be adopted. Also, the learning stage
Since it is possible to correspond to the stages of conversion , it is possible to learn at each stage if the conversion is multistage.

[0023] The Japanese conversion control unit, and the candidate generation unit for generating a converted Japanese candidates, a means and display means for selectably displaying the converted Japanese candidates, the display unit, the Japanese It is also preferable that the word candidates are displayed in different modes for each combination of kana that can occur from the Japanese input element. In this way, a desired character string can be found early from the Japanese candidates displayed as candidates.

The conversion of Japanese does not have to be performed in units of one clause represented by a Kanji dictionary. As an example, the invention according to claim 7 is defined in any one of claims 1 to 5. In the character input device described above, the Japanese-language conversion control means stores the grammatical information storage means for storing grammatical information of words in the dictionary, and the kana character string obtained by the first-stage conversion. Based on the grammatical information, bunsetsu segmentation and writing means that divides into bunsetsu units,
A connection verification means for performing connection verification between the segmented clauses based on the stored grammatical information is provided, and by performing the verification by the connection verification means, the Japanese character string mixed with kanji is converted. The point is that the method is used.

In the invention according to claim 7 configured as described above, the combinations are classified from the combinations that can occur in the Japanese input elements assigned to the input keys when the key input means is operated. This segmentation itself also contributes to the conversion into natural kana character strings. In other words, if the connection test between clauses is performed based on the grammatical information for the clauses that can occur, the combinations that are allowed will be converted , and the conversion that cannot be performed when inputting in clause units is realized. You can do it .

Therefore, according to the invention of claim 7,
Since the conversion is accurate by the processing of bunsetsu, and the one that remains in the connection test between multiple bunsetsu becomes the target of conversion ,
It's easier to use because it can be converted more like Japanese and can be converted after inputting multiple clauses.

[0027] grammar information that can be used in such a connection test are those that can be interpreted in a broad sense, but it may also be a part of speech of the information that may be connected.

[0028]

[0029]

As described above, the key input means having a plurality of input keys can have various concrete configurations. As an example thereof, the invention according to claim 8 is the character input device according to any one of claims 1 to 7 , wherein the key input means is the plurality of input keys, and The means is configured by using some of the keys.

In the invention according to claim 8 configured as described above, since there are many input devices having a plurality of input keys, a predetermined number of input keys in such input devices are selectively utilized and selected. Operate the entered input key to input the Japanese input element. Therefore, according to the invention of claim 8 , it is possible to input Japanese by using various input devices which are not limited to the key input means in the narrow sense.

As another example, the invention according to claim 9 is the character input device according to any one of claims 1 to 7 , wherein the key input means includes at least the Japanese input element of at least a character. Display means capable of displaying in form,
A part specifying unit that is provided integrally with the display unit and specifies the contacted part by touching a specific part on the display by the display unit, and the character is displayed on the part specified by the part specifying unit. It is characterized by comprising an input element input means for inputting a Japanese input element corresponding to the character when being displayed. According to this configuration, the character is input by touching the portion displayed in the character form by the display means. Therefore, even a user who is not familiar with keyboard operation can easily create a sentence.

The invention according to claim 10 is the following:
9. The character input device according to claim 9 , wherein the display means includes a transparent pressure-sensitive film that covers at least an area in which the characters are displayed, and the site specifying means is provided with respect to the transparent pressure-sensitive film. The gist is that it is a means for identifying the contacted portion. That is, a touch display is provided that covers the display device with a transparent pressure-sensitive film to input the display content of the touched portion.

In the invention according to claim 10 configured as described above, the key input means is a touch display covered with a transparent pressure-sensitive film on the display device, and the Japanese input element is displayed on the display device. If you touch the Japanese input element you want to input, the display content of the touched part will be input. In this case, the display on the display device may be changed appropriately.

Therefore, according to the tenth aspect of the present invention, by using the touch display, it is possible to change the allocation of the Japanese input elements depending on the situation, and it is possible to improve the input efficiency.

As another example, the invention according to claim 11 is the character input device according to any one of claims 1 to 7 , wherein a key for displaying an array of input keys to which the Japanese input elements are assigned. Arrangement display means is provided, and the key input means is means for inputting a Japanese input element assigned to the designated key when a specific input key is designated on the display by the display means. And A twelfth aspect of the invention is summarized in that the key input means is configured to instruct the specific input key by using a pointing device. For example, the key input means may be configured as a software keyboard including a display device and a pointing device.

Claims 11 and 1 configured as described above
In the invention according to 2 , the input key is not a physical operation key, but is realized by a so-called "software keyboard" that displays the key layout on the screen of the display device and instructs the key layout to be displayed by the pointing device. become.

Therefore, according to the eleventh and twelfth aspects of the present invention, by using the software keyboard, it is easy to change the arrangement of the input keys according to preference, and change according to the language. , It can be displayed only during the input operation to effectively utilize the narrow display area.

As another example, the invention according to claim 13 is provided with a keyboard in which a plurality of input keys to which Japanese input elements based on a combination of vowels and consonants are assigned are arranged, and arranged on the keyboard. A character input device for inputting a character string mixed with Japanese Kanji characters based on the operation of the input key, wherein the Japanese input element is allowed to be duplicately assigned to one key in the input key. A kana character string conversion means for allocating and converting a character string generated from the redundantly allocated Japanese input elements into a Japanese kana character string with reference to a dictionary and converted by the kana character string conversion means. The kana-kanji conversion means for converting a kana-character string into a character string mixed with Japanese kanji by referring to a dictionary is further provided. The kana character string corresponding to the kanji mixed character string finally determined from the candidate display means shown and the result of the conversion performed in the past is stored, and the stored kana character string is stored next time. The gist is that the means for preferentially making the candidate in the conversion is provided.

According to a thirteenth aspect of the invention configured as described above, the character input device according to any one of the first to seventh aspects is a dedicated keyboard in which the plurality of input keys are arranged. Operate the input keys to input Japanese input elements. Therefore, according to the thirteenth aspect of the present invention, by using the dedicated keyboard, the candidate character strings can be learned, so to speak, and efficient Japanese input can be realized universally.

As described above, the method of converting into natural Japanese corresponding to the input operation performed by the key input means in which the Japanese input elements are redundantly assigned to the plurality of input keys is a substantial device. It is not necessary to be limited to the above, and it can be easily understood that this method also functions. For this reason,
The invention according to claim 14 is a character input method for detecting an operation of an input key of a key input means to input Japanese.
A Japanese input element is input from a keyboard that has a plurality of input keys and allows duplicate assignment when assigning a Japanese input element based on a combination of vowels and consonants to each input key, and assigns it to the input key. When inputting Japanese by referring to the dictionary assuming the Japanese input element, the character string that can occur in the Japanese input element assigned to the same input key is converted into the Japanese Kana character string. The conversion and the second stage conversion of converting the Japanese kana character string into a Japanese kanji mixed character string are performed, and the kanji character mixture finally determined from the result of the conversion performed in the past. Storing the kana character string in the first-step conversion corresponding to the character string, and increasing the priority of the stored kana character string in the first-step conversion,
It is configured to perform the conversion after the next time.

That is, there is no difference in that the method is not limited to a substantial device and is effective as a method.

The idea of the invention of inputting characters by the above method includes various modes. That is, it can be changed as appropriate, such as being realized by hardware or software. If the example embodying the idea of the invention is software for character input processing, it must be said that it naturally exists and is used on a software recording medium recording such software. As an example thereof, the invention according to claim 15 is a medium in which a character input control program for detecting an operation of an input key of a key input means by a computer and inputting Japanese is provided with a plurality of input keys. A function of inputting a Japanese input element from a keyboard that allows overlapping assignment when assigning a Japanese input element based on a combination of vowels and consonants to each input key, and a Japanese input element assigned to the input key The first step of converting a character string that can occur in a Japanese input element assigned to the same input key into a Japanese kana character string when inputting Japanese with reference to a dictionary, and the conversion. The final decision is made from the function of performing the second-stage conversion that converts the converted Kana character string into a Japanese Kanji mixed character string and the result of the conversion performed in the past. A function of storing the kana character string in the conversion in the first step corresponding to a mixed kanji character string and a high priority in the conversion in the first step of the stored kana character string It is a medium in which a program for causing a computer to realize the function of performing the conversion of is recorded.

Therefore, according to the fourteenth aspect of the present invention, it is possible to provide a character input method that achieves the same effect. According to the fifteenth aspect of the present invention, it is possible to provide a medium in which the character input control program is recorded.

Of course, the recording medium may be a magnetic recording medium or a magneto-optical recording medium, and any software recording medium developed in the future can be considered in exactly the same manner. In addition, the duplication stage of the primary duplication product, the secondary duplication product, and the like is absolutely the same. In addition, the present invention is used even when a communication line is used as a supply method, and the same applies to a case where the present invention is written in a semiconductor chip.

Further, even when a part is software and a part is realized by hardware, there is no difference in the idea of the invention, and it is necessary to store a part on a software recording medium. It may be in a form that can be appropriately read according to the above.

[0047]

Here, the grammatical information of a word means information about the nature and usage of the word, for example, distinction between an independent word and an annex,
It includes information such as distinctive and indivisual terms for independent words, part-of-speech information for each word, and information about whether or not to connect words or phrases.

The adjunct word means a word that forms a clause only after being attached to an independent word, for example, an auxiliary verb or a particle. Of course, it is acceptable to include inflection endings and suffixes. The vowels have five tones of "aiueo" in Japanese, and the consonants have tones other than vowels.

[0050]

[0051]

[0052]

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

The character input method of the present invention refers to a dictionary that stores a word and grammatical information of the word based on the operation of an input key in which a part of the 26 letters of the alphabet is assigned as a Roman character for Japanese input. to create a Japanese character string while, a character input method for inputting the character string, assign duplicate two or more of the Roman alphabet in 1 key, and
The consonants that make up the adjuncts "ga", "da", and "ha"
Based on the operation of the input keys respectively assigned to different keys, a step of converting into kana character strings allowed by the combination of Roman letters assigned to the respective input keys, and the converted kana character strings into the dictionary. Referring to gist that example Bei and a step of converting the candidate character string of kanji and kana while.

According to the character input method of the present invention, when the same input key is operated, a plurality of kana character strings which are the subordinate words "ga", "da" and "ha" are not generated. Therefore, the number of candidate character strings can be reduced by converting the narrowed kana character string into a kana-kana mixed character string.

The recording medium recording the computer program of the present invention stores a word and grammatical information of the word based on the operation of an input key in which a part of 26 letters of the alphabet is assigned as a Roman character for Japanese input. A recording medium in which a Japanese character string is created with reference to a dictionary and a computer program for inputting the character string is recorded, and two or more Roman characters are assigned to one key in duplicate.
And compose each subordinate word such as "ga""da""ha"
Converting consonants into kana character strings permitted by the combination of Roman letters assigned to the respective input keys, based on the operation of the input keys assigned to the different keys ;
The converted kana character string, and summarized in that were readable recording a computer program for executing a step of converting the candidate character string of kanji and kana in the computer with reference to said dictionary.

Even with such a configuration, when the computer reads and executes the program recorded in the recording medium, in the process of inputting the Japanese character string, the subordinate characters "ga", "da" and "ha" are added. Multiple kana character strings that are words are not generated. Therefore, the number of candidate character strings can be reduced.

[0063]

[0064]

In the character input device according to the second aspect of the present invention, if the key input means is a means in which one of a plurality of input keys is assigned a dakuten as a Japanese input element, it is possible to easily input a kana having a dakuon. can do. In this case, the input key may be assigned a semi-voiced point.

Further, if at least a consonant representing a dull sound is excluded and assigned to the input keys, it is possible to realize character input with a smaller number of input keys. Here, the "consonant representing a dull sound" refers to a consonant that, when combined with a vowel, produces a reading of a kana that has a dull sound, such as "gagugugogo" in "ga" or "zajizuzezo" in a row. For example, "B", "D", "G" and "Z" correspond to this.

[0067] The key input means, voiced sound mark is, of the Roman
If at least one key is assigned in duplicate, it is not necessary to provide an independent input key for assigning a dakuten, and the number of keys can be further reduced. In addition, if the key input means is a means for assigning the vowel and the dakuten of the Roman alphabet to the different input keys, a plurality of kana character strings will not be generated based on both the dakuten and the vowel. Therefore, it is possible to prevent an increase in the number of candidates for a character string containing kanji and kana.

[0068] A consonant representing the Kiyone, from a continuous combination of the corresponding consonant and a voiced sound mark that dullness exists, includes a consonant specifying means for specifying the consonant representing the dullness, the Japanese conversion control unit, the consonant specifying means the combination of a specified consonants and vowels and Romaji kana conversion by 1 kana sentence
It is also desirable to use it as a means for converting into a character string . By adopting such a configuration, it is possible to realize an efficient dictionary search without converting a kana having a voiced sound into a kana character string having a clear sound while inputting Roman characters.

Here, "a consonant that represents a pure sound and a corresponding consonant exists" means a consonant that constitutes a kana to be made into a dull sound by adding a dakuten. "K", "S", "T", and "H" constituting the row kana correspond to this.

Further, the key input means sets the consonant representing a pure sound and the corresponding consonant and the dull point to 1
If a means for assigning the input keys to the input keys is duplicated, it is possible to input a kana having a dull sound by continuously operating the same key. Therefore, it becomes easy to operate the keys associated with the character input.

Further, the kana conversion means for converting the combination of the first consonant, the second consonant and the vowel into a kana character string by the roman character kana conversion is provided, and the key input means is used for the second of the roman characters. Turkey allocates a and the consonants and the sound mark the different input key and also suitable. As the second consonant, a consonant located between the consonant and the vowel (hereinafter,
Half-vowels), and among the three Roman letters when the kana of 1 is spelled out in Roman letters in the Hepburn style, consonants located between the consonants and vowels are considered. According to such a configuration, one of the first consonant, the second consonant and the vowel
Combined by, without having to convert duplicate and Do string and other kana string or with a dullness, it is possible to prevent the flood of candidate strings.

[0072] by the Roman kana conversion, change from a combination of a continuous and vowels of the same consonant to Do string or have a double consonant
Conversion to comprise a double consonant transformation unit, key input means, those hands assigned to different input keys and the consonant and voiced sound mark of the Roman alphabet
It may be. In this case, the combination of 1 of the input keys does not generate the kana character string having a consonant and the kana character string having a dull sound in an overlapping manner, and can prevent the candidate character strings from being scattered.

[0073] The key input means is a consonant representing the Kiyone, if those Ri split <br/> a consonant corresponding voiced consonants are present in different input keys, the first combination of input keys, having dullness It is possible to prevent the occurrence of duplicate candidate character strings without generating a plurality of kana characters. For example, if a consonant representing a pure sound and a corresponding consonant is present as “H”, “K”, “S”, and “T”, the combination of the same vowel produces a line, No more than two kana character strings that belong to a line, a line, or a line.

Further, the character input device of the present invention is based on a plurality of input keys in which Roman characters as Japanese input elements are assigned while allowing duplicate assignment to one key, and the operation of the input keys. The obtained key input is an alphabetic string consisting of the combination of Roman letters assigned to each input key.
And letters sequence converting means for converting, by該英string converting means
A character input device for inputting a character string selected from the candidate character strings, comprising: a conversion unit for converting a candidate character string containing kanji and kana based on the converted alphabetic character string into a candidate character string. to one of a plurality of input keys, the voiced along with the assignment Teru as the Japanese input element, the letters column strange
The conversion method is finally decided based on the result of the conversion performed in the past.
The above-mentioned English corresponding to a candidate character string containing a mixture of kanji and kana
An alphabetic string storage means for storing a character string, and a description in the alphabetic character storage means.
For the above-mentioned remembered alphabetic characters, the priority is set to be higher and
The gist is that it is provided with a priority conversion means for converting the descending power .

According to this structure, the conversion means converts the English character string consisting of the combination of Japanese input elements including the dakuten assigned to each input key into a candidate character string containing kanji and kana by referring to the dictionary. To do. Therefore, it is possible to easily input a kana having a dull sound or a semi-voiced sound.

The character input method of the present invention is a character input method for creating a Japanese character string based on the operation of an input key to which a Roman character is assigned as a Japanese input element and inputting the character string. key obtained based on the operation of the input key assigned duplicate two or more of the Roman alphabet key
-A step of converting the input into a kana character string permitted by the combination of Roman characters assigned to each input key, and a step of converting the converted kana character string into a kanji-kana mixed Japanese character string In the step of converting into the kana character string , when a dakuten is assigned to the input key as the Japanese input element, the combination of the romaji assigned to the input key and the dakuten is a dakuon. From the process of generating kana characters and the results of conversion performed in the past, the maximum
Corresponds to the finally decided candidate character string containing kanji and kana.
Storing the kana character string
For kana character strings, make the priority higher and convert it from the next time
The gist of the present invention is to include the step of performing .

According to the character input method of the present invention, the input key
From the combination of Roman letters and dakuten assigned to
Generate kana characters that are soundsAnd yet, the conversions done in the past
From the results of the above, the final candidate for kanji and kana mixing
The Kana character string corresponding to the character string is stored and stored.
For the kana character string
Do the conversion of. Therefore, when inputting characters,
Does it have a dull sound?Also efficientlyCan be entered
You, The final Japanese character string can also be obtained efficiently
It

The recording medium on which the computer program of the present invention is recorded creates a Japanese character string based on the operation of an input key to which Roman characters are assigned as Japanese input elements,
A recording medium recording a computer program for inputting the character string, based physician to two or more operators of the said input key assigned duplicate Romaji one key
In the step of converting the key input obtained as described above into a kana character string permitted by the combination of Roman characters assigned to each input key and the step of converting into the kana character string , the input key is used as the Japanese input element. When a dakuten is assigned, the combination of the Romaji and the dakuten assigned to the input key is used as a kana character that is a voiced sound .
The process of converting to a character string and the conversion performed in the past
Candidate sentence with kanji and kana finally decided from the result
Storing the kana character string corresponding to the character string;
For the stored Kana character string, set the priority to the next
The gist is that a program for causing a computer to execute the step of performing subsequent conversions and the step of converting the kana character string into a Japanese character string containing kanji and kana is readable by the computer.

According to the recording medium on which the computer program of the present invention is recorded, the computer reads and executes the program recorded on this recording medium,
Generates kana characters that are dakuon from the combination of romaji and dakuten assigned to the input keys , and also performed in the past
The kana and kana mixture finally determined from the conversion results
The Kana character string corresponding to the candidate character string
Then, increase the priority of the memorized kana character strings.
Convert after next time . Therefore, at the time of character input, it is possible to efficiently input kana with a daku sound by a simple method and efficiently obtain the final Japanese character string.
It is Ru can.

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a character input device according to an embodiment of the present invention by means of a claim correspondence diagram,
FIG. 2 is a block diagram showing a specific circuit configuration, and FIG. 3 is a PIM (Perso) to which a character input device is applied.
FIG. 4 shows a module configuration including software. FIG. 4 shows an external appearance of a terminal information manager) terminal 80.

Computer system 7 of this character input device
0 is incorporated in the PIM terminal 80 shown in FIG.
As shown in FIG. 1, the configuration of the function is the key input means 1
0 and a Japanese narrowing-down control means 20. Further, the Japanese narrowing-down control means 20 performs two-stage conversion described later based on the operation of the key input means 10.
The candidate generating means 30 for generating candidates for Japanese character strings, the display means 40 for displaying the generated candidates, the selection control means 50 for selecting a desired candidate from the displayed candidates, and the selection result are described below. The learning control means 60 feeds back the candidate generation. Here, the key
Reduce the number of conversion candidates in Kana-Kanji conversion based on input
The idea is called "narrowing down". That is, Japanese aperture
The inclusion control means 20 is for performing kana-kanji conversion,
"Refine" means to generate a kana character string from key input,
We will show the process in the two-step process of Kana-Kanji conversion.
is doing.

Each of these means is realized by controlling various information inputs and information displays in the computer system 70 shown in FIG. The computer system 70 includes a CPU 71 that controls input / output signals in the PIM terminal 80, a program ROM 74 and a dictionary ROM 75 that store various types of processing executed based on instructions from the CPU 71, a work RAM 76 as an area for storing processing data, and a battery. Backup RAM7
7, a keyboard 73 for inputting an input signal to the CPU 71 via the I / O 73a, and an I / O output signal from the CPU 71
It has a liquid crystal display 78 that outputs through O78a, and these hardwares are connected by a bus 72.

The surface of the display portion of the liquid crystal display 78 is formed using a transparent pressure sensitive film, and a pressure sensitive element 78b capable of detecting the pressing position on the film is arranged on the back side of this transparent pressure sensitive film. In the present embodiment, the arrangement of keys is displayed on the liquid crystal display 78 configured as described above. Therefore, when the key arrangement is displayed on the liquid crystal display 78, the display of the key arrangement on the display, the transparent pressure sensitive film and the pressure sensitive element 78b are integrated to form the keyboard 73.

Each unit shown in FIG. 1 is realized by such a computer system 70. Next, the configuration of each module that realizes the candidate generating means will be described with reference to FIG. The candidate generating means includes a Roman character generation module M1 for generating a Roman character based on a key input, a Roman character kana conversion module M2 for converting a Roman character string into a kana character string, and a kana character string mixed with kanji into a Japanese character string. It has a kana-kanji conversion module M3. In the generation process or conversion process in each of the modules M1, M2 and M3, the key-alphabet table ET, the Roman character-kana conversion dictionary ND and the kana learning dictionary NGD, the kanji dictionary KD and the kanji learning dictionary KGD respectively correspond to dictionary files. Referred to as. The basic one of these dictionary files is stored in the dictionary ROM 75, and the learning dictionary is stored in the battery backup RAM 7.
It is stored in 7.

This character input device is a computer system 7
Since it is realized by 0, it cannot be clearly classified as hardware as shown in FIG.
71 serves as a core, and software and hardware are organically combined to form each means. For example, in the PIM terminal 80, since the input mechanism by the touch display is adopted, the CPU 71 displays the image of the key arrangement on the liquid crystal display 78 and displays the pressed position and the key arrangement displayed by the pressure sensitive element 78b. Since the input keys are detected by associating them with each other, it can be said that these integrally form the key input means 10. In addition, the operation of the input key thus detected is temporarily stored in the work RAM 76, and the dictionary RO is used in accordance with the flow shown in FIG.
Some Japanese character strings are generated by performing Romaji-Kana conversion or Kana-Kanji conversion with reference to the data in M75, and it can be said that these are integrated and constitute the candidate generating means 30. The dictionary ROM 75 and the battery backup RAM 77 will store various dictionary files, and will form a kana arrangement storage means.

Further, since the generated candidates are displayed on the liquid crystal display 78 by the CPU 71, it can be said that these are integrated to constitute the display means 40, and the pressure sensitive element 78b can be used while looking at the displayed candidates. Since the candidate is selected by the touch operation of, the CPU 71 and the like including the pressure-sensitive element 78b and the liquid crystal display 78 are the selection control means 5.
It can be said that it constitutes 0. Further, as will be described later, information on the candidate selected in the selection process (not necessarily the candidate itself) is stored in the battery backup RAM 77 and referred to in the same manner as the dictionary ROM 75 at the next candidate selection. Therefore, it can be said that these together constitute the learning control means 60.

Of course, the hardware circuit configuration example is not limited to that described above, and the character input device is a PI.
It is possible to apply as various devices other than the M terminal 80. For example, FIG. 6 shows an example realized by a normal desktop personal computer 81, and a full keyboard 81
It can be applied to character input using the ten-key pad 81b of a. Further, FIG. 7 shows an example applied to the mobile phone terminal 82. As will be described later, if the Japanese character string can be input with 12 input keys, the mobile phone terminal 82 will be described.
The necessary number of input keys can be secured with the keys 82a (number keys 0 to 9 and * key and # key), and the Japanese character string can be displayed on the display portion 82b. Further, FIG. 8 shows an example in which a Japanese character string can be input on the remote control terminal 83 of the television.
12 input keys can be secured by the channel buttons 83a to 12a, and infrared light output when the channel button 83a is pressed may be detected by the light receiving unit 83b and replaced with keyboard operation. In addition, it is effective to be installed in a device such as a fax machine, a printer, a copier, or the like in which a place for providing an input key is small.

Returning to the PIM terminal 80 of this embodiment, the liquid crystal display 7 is assumed to generate a Japanese character string.
A character string input screen 90 as shown in FIG. 9 is displayed on the screen 8. The screen character string input screen 90 will be described below as the first embodiment. Explaining the main areas, the input key area 91 in the form of a grid of four rows and four columns is provided on the right side of the drawing, and candidates for displaying the narrowed candidate character strings are provided on the left side. There is a display area 92, and above the input key character string display area 93 for displaying the input character keys. In addition, in the non-input state, “switch” is displayed in the lower left cell of the input key area 91 and “space” is displayed in the upper left cell, but during input, as shown in FIG. The display is changed to "conversion" and "decision". All of these contribute to reducing the number of keys because they are selected only in the case of the non-input state and the case of the input state. Further, an up arrow and a down arrow are displayed along the upper side and the lower side of the candidate display area 92, so that the candidates not displayed in the window can be scroll-displayed.

Further, in the information of the input key character string display area 93, a confirmed character string display area 94 for displaying a Japanese sentence consisting of a confirmed Japanese character string is displayed, and the confirmed Japanese character is displayed on the right side thereof. A transfer button 95 for transferring the word character string to the application and a clear button 96 for discarding the same Japanese character string are displayed.

In the character string input screen 90, the input key area 91 is substantially divided into 16 squares, and configured as 16 input keys. When you try to input Japanese in Romaji, you need to input vowels and consonants, punctuation marks, long sounds, and Japanese syllables, so 16 input keys are not enough, and each input key is assigned a single key. Some are assigned and some are assigned multiple keys.
In the example shown in the figure, basically, there is a tendency that a vowel is assigned to a single vowel and a consonant with a high occurrence frequency is assigned to a single vowel. This allocation has a great influence on the input efficiency, and various measures have been added. However,
The arrangement itself has a high degree of freedom, and for example, 12 input keys can be used as in the character string input screen 12 shown in FIG. By using twelve input keys, it becomes possible to input a Japanese character string using a very familiar device such as a telephone or a remote controller of a television as shown in FIGS.

The processing of software for generating a Japanese character string on the premise of such a character string input screen 90 will be described. FIG. 5 is a flow chart showing the Japanese character string generation processing A. In this flow chart, a dictionary file to be referred to is also described together with the software processing. In addition, in the same flowchart, the pressure sensitive element 7
The processing for detecting the input key from the output signal of 8b, the display control on the liquid crystal display 78, and the like are omitted because they are executed by the firmware.

When a touch operation is detected on the pressure sensitive element 78b, the CPU 71 activates the Japanese character string generation process A, and first, in step S100, the type of the input key is determined. When the detected position of the pressure-sensitive element 78b is the input key area 91, it is determined in step S100 that it is a character key, and it is added to the key character string in step S102. Here, the key character string is a physical number assigned to each square of the input key area 91, as shown in FIG. 12, and at this point, the alphabetic characters in the arrangement as shown in the character string input screen 90 are displayed. Does not mean. The generated key character string is temporarily stored in the buffer secured in the work RAM 76 and is sequentially added to the end of the undetermined key character string.

In step S104, the undetermined key character string is converted into an alphabetic character, and the key-alphabet table ET is referred to at this time. As described above, the key character string is a physical number string assigned to each square, and is converted into an alphabetic character string by referring to the key-alphabet table ET shown in FIG. By using the key-alphabet table ET in this way, the key layout can be easily changed by changing the contents of the table. With this, the key character string becomes an alphabetic character string, and in step S106, roman characters are converted into kana. At the time of romaji conversion, conversion is performed by referring to the romaji-kana conversion dictionary ND shown in FIG. 14, but the kana learning dictionary NGD shown in FIG. 15 is also referred to and the conversion result at the time of past romaji-kana conversion is fed back. I am using it. The learning function based on this kana learning dictionary NGD will be described in detail later.

When the Romaji-Kana conversion is performed on the basis of the learning result, in step S108, as shown in FIGS.
Kanji dictionary KD shown in FIG. 20 and Kanji learning dictionary KGD shown in FIG.
Convert to a Japanese character string mixed with kanji by referring to.

The Kanji dictionary KD is provided with an independent word dictionary and an annex word dictionary as a dictionary for storing information of words that can form a Japanese character string. The structure of this independent word dictionary is shown in FIG.
FIG. 18 shows the structure of the annex dictionary, respectively. The independent word dictionary stores information about words that are independent words, and the annex dictionary stores information about words that are auxiliary words.

The word information stored in these dictionaries includes character information and grammatical information, and the character information is stored in the order of the Japanese syllabary corresponding to the index which is a search heading. In this embodiment, the POS information of each word is stored as the grammatical information. This part-of-speech information is shown in FIG. 17 and FIG.
As shown in FIG. 3, the data is stored in the form of a three-digit part-of-speech code determined according to the type of part-of-speech such as noun, verb, adjective, auxiliary verb, and particle.

For example, the Japanese character string "box" is
It consists of two words, "box", which is an independent word that represents a noun, and "wo," which is an annex that represents a case particle. Regarding these two words, the independent word dictionary has a part-of-speech code representing the Chinese character “box” and the common noun “200”, and the auxiliary word dictionary has the kana character “40” and “40”.
The part-of-speech code representing the case particle "0" is stored. Thereby, it becomes possible to detect the part of speech of the dictionary-searched word from the code number.

The Kanji dictionary KD has a connection verification table shown in FIG. This connection verification table is for determining whether or not the words stored in the independent word dictionary or the annexed word dictionary can be connected, and is a matrix table in which each part-of-speech code is assigned to the part-of-speech column on the vertical and horizontal axes. Prepared in format.

In this matrix table, the part-of-speech code in the part-of-speech column on the vertical axis is the part-of-speech code of a particular word in the Japanese character string, and the part-of-speech code in the part-of-speech column on the horizontal axis is in front of the particular word. Indicates the part-of-speech code of a word. Information indicating whether or not there is a grammatical usage of the vertical part of speech following the horizontal part of speech is pre-registered at the intersection of the vertical part of speech column and the horizontal axis of part column. This information is actually registered in the form of a combination of bit information, but for convenience of explanation, a combination of connectable part-of-speech is shown as “◎” and an unconnectable part-of-speech combination is shown as “x”. .

For example, when converting a kana character string such as "hako wo kobubu" into a character string mixed with kanji from the beginning of a sentence, after searching the word "box" from the independent word dictionary, the word "wo" is annexed. When searching from a dictionary, the intersection of the part-of-speech code representing the common noun “200” on the vertical axis and the part-of-speech code “400” on the horizontal axis of the connection verification table is marked with “◎”. Therefore, it is determined that the “box” and “o” can be connected.

As described above, by determining whether or not the preceding and succeeding words can be connected to each other, it is possible to convert each phrase into a Japanese character string. In the present embodiment, the word-of-speech information of a word is stored as grammatical information in the independent word dictionary and the annexed word dictionary, respectively, but the present invention is not limited to such a configuration, and for example, one kanji dictionary KD can be used as grammatical information. It may be configured to store the independent word of the word and the distinction of the auxiliary word. With this configuration,
It is possible to realize conversion in phrase units.

The kanji dictionary KD having such a configuration converts a kana character string into a character string mixed with natural kanji in Japanese. Here, while assigning a plurality of Roman characters to each input key, The alphabetic character string generated based on the operation is narrowed down to a character string that can be converted into kana, and further narrowed down to the character string mixed with kanji registered in the dictionary. By this two-step narrowing down, it will be converted into a character string that is natural in Japanese. The learning function based on the Kanji learning dictionary KGD will be described later.

The description returns to FIG. Regardless of whether or not the character is being input, the character string is converted into a character string containing kanji up to the point possible by operating one character key, and then step S11 is performed.
At 0, a plurality of Kanji candidates at this point are displayed, and the process returns to step S100. In the case of "Haikei", the Romaji-kana conversion is possible at the time of inputting "HA" and at the time of inputting "I", "KE", and "I", and the learning result at those input times is given priority. As a target, a Japanese character string mixed with Kanji corresponding to "Haikei" and a character key to which "H" is assigned are duplicated with "G". The character string is displayed in the candidate display area 92. As described above, the candidate character string is generated and the candidate character string is displayed.

The input person suspends the operation of the character keys at the punctuation unit and operates the candidate move key to scroll until the desired Kanji conversion result is displayed in the candidate display area 92. In this case, as will be described later, the selected candidate character string is displayed in reverse video, and other candidate character strings are displayed as the position of the reverse display moves in the vertical direction.

The candidates are displayed in different colors so that they can be recognized when the conversion source Roman characters are different. For example, a Japanese character string mixed with kanji corresponding to “Haikei” uses yellow, and a Japanese character string mixed with kanji corresponding to “gaikei” uses light blue. This state is shown in FIG. As a result, it is possible to quickly find the desired Japanese character string containing kanji. Of course, whether or not to execute the color change including the case of the black and white display is arbitrary.
A method of changing the shading as shown in (B) may be used.

The candidate movement key means an up arrow or a down arrow, but these are not character keys. For this reason,
After it is determined in step S100 that it is not a character key, in step S112 it is determined whether it is a cancel key, and in step S114 it is determined whether it is a candidate move key. If so, the current candidate is changed in step S116. That is, as described above, the reverse display is moved up and down in accordance with the operation of the candidate move key by the user, and the candidate character string displayed in the candidate display area is reversely displayed. On the other hand, if it is determined in step S114 that the key is not the candidate move key but the enter key, the current candidate is learned in step S118.

On the other hand, although not shown in FIG. 5, a case will be described in which the conversion key is operated after conversion into a character string containing kanji. In the automatic conversion mode, I try to convert kanji as much as possible, but sometimes I want to input hiragana and katakana. Therefore, such input is possible by pressing the conversion key. Specifically, when the conversion key is operated once, it is converted into hiragana, and when the conversion key is operated twice, it is converted into katakana.

The learning process of step S118 will be described. The learning function is roughly divided into two stages. That is, the kana learning function and the kanji learning function.
The latter is the only conventional Japanese Kana-Kanji conversion, but it is useful because it requires a Kana learning function because multiple English characters are assigned to one input key.

First, the kana learning function will be described. When converting from Romaji to Kana, a plurality of English characters are assigned to one input key, so if "Hakai" is entered, the key string shown in FIG. 12 is "11, 5, 10, It will be operated in the order of "9, 7, 10". However, the same operation is necessary even when inputting "gaikei". In this case, it may be difficult to determine which is desired, but if the determination is made within one sentence, the same conversion is often desired. Therefore, in order to display the learned results from the next conversion aside from the first conversion,
The Romaji-kana conversion result adopted when it is determined in step S114 that the decision key has been operated is the kana learning dictionary N.
Store in GD. FIG. 15 shows a state in which, when a candidate character string corresponding to a kana character string “Haikei” is adopted, this kana character string is stored in the kana learning dictionary NGD as a Roman character kana conversion result. The enter key functions in the so-called automatic conversion mode, but at this time, the input key character string display area 93 does not show the kana to be converted. However, internally, because the result of Romaji-kana conversion as described above is known,
This is stored in the same kana learning dictionary NGD.

Therefore, at the time of Romaji-Kana conversion, the Kana learning dictionary NGD is referred to together with the Romaji-Kana conversion dictionary ND, and the Kana character string stored in the Kana learning dictionary NGD is converted with a high priority. As a configuration for increasing the priority, for example, a configuration in which a plurality of kana character strings are the first candidates can be considered.

Next, the kanji learning function will be explained.
In this case, as usual, when it is determined in step S114 that the decision key has been operated, the selected kanji candidate and the kana candidate are associated with each other and the kanji learning dictionary K shown in FIG.
Store in GD. FIG. 20 shows a state in which this kanji character string is stored in the kanji learning dictionary KGD when the kanji candidate “Dear Sir” is adopted. As a result, next time,
When the key character string is operated in the order of "11, 5, 10, 9, 7, 10", "hakaii" is made a priority candidate by the above kana learning, and "hakaii" is a kanji character. By learning, "dearing" is converted as a priority candidate.

In the present embodiment, the learning function is realized by the work of creating such a learning dictionary file and the work of referring with a prefix match at the time of the next conversion, but this is merely an example and other It is possible to add various learning functions. For example, the correspondence relationship between parts of speech may be learned, or suffixes, prefixes, and the like may be learned.

After learning the current candidate in step S118, in step S120, the collocation corresponding dictionary shown in FIG. 22 is referred to by using the conversion result of this time so that the collocation conversion can be performed for the next conversion. In the learning dictionary.
For example, since "Hayai" was selected, "Hayai"
If you search the lexical correspondence dictionary with a prefix, you will find "Haikai, Jika". In this case, the candidate is "Dear Sirs, Dip"
Therefore, in the Kanji learning dictionary KGD, "Jika" and "Tokishita"
And are added. As a result, when the next reference is made, such as "Jikka", "Jikan" is preferentially displayed over other candidates such as "market price". Of course, other than the kanji dictionary KD, another collocation-corresponding learning dictionary may be created. Finally, in step S121, the current candidate is transferred to the confirmed character string display area 94,
The process ends.

By the above processing, the selection operation and the learning processing are performed, and a series of processing including candidate generation, candidate display, selection, and learning is performed as a whole.

The example described above is based on the so-called automatic conversion mode in which conversion is automatically performed according to the key character string that is sequentially input. However, kana-kanji conversion is performed when the conversion key is operated. A mode (so-called successive conversion mode) is also easy. A flowchart in such a case is shown as a Japanese character string generation process B in FIGS. 23 and 24. According to this processing, candidates are generated even in the process of converting from Roman letters to Kana, and the selection result can be effectively reflected in the Kana learning dictionary NGD. Note that the flowcharts shown in FIGS. 23 and 24 generally correspond to the flowchart shown in FIG. 5, and of the steps shown in FIG.
The description is omitted.

When a touch operation is detected on the pressure-sensitive element 78b, the Japanese character string generation processing B is activated to convert the key character string based on the operation of the character key into an alphabetic character (step S20).
4) Convert the converted English characters into a kana character string and display the kana candidate group (steps S206, S208, S21).
0). When it is determined that one candidate is selected from the kana candidate group and the decision key is operated, the one candidate is determined as a kana character string to be converted to kanji, and this kana character string is stored in the kana learning dictionary NGD. Remember (step S21
6), and branches to the kanji conversion processing of FIG. If there is a key input after this, it is determined that the current mode is being converted (step S200), and the process branches to the Kanji selection process shown on the right side of FIG.

The Kanji conversion processing is basically the same as in the case of the flow chart shown in FIG. 5, and the Kana learning string KGD and the Kanji dictionary KD are referred to to convert a kana character string into a kanji character and a kanji candidate group. Is displayed (step S23
0, S232), and branches to the kanji selection process. In the kanji selection process, when it is determined that the input key is the candidate movement key, the desired candidate is highlighted (step S23).
4, S236), when it is determined that the input key is the enter key, the current candidate is decided as a character string mixed with kanji, the character string is stored in the kanji learning dictionary KGD, and kana-kanji conversion learning is performed. . After that, the confirmed kanji-mixed character string is transferred (step S240), and the process ends. In step S238. This makes it easier to reflect the result of conversion from Roman letters to kana at the next conversion.

Although the above description has been focused mainly on the conversion of a single phrase, it goes without saying that the continuous phrase conversion is possible. It seems that the complex clause conversion tends to become more complicated, because multiple kana conversions occur due to duplicate assignment for each clause.
However, when paying attention as a phrase, it is necessary to perform phrase segmentation for the input sentence, and the grammar information in this case can be used for narrowing down the kana conversion itself.

When an example of "Hako to Hakobu" is entered using the input key area 91, the key character string is "HG, A,
K, O, YRW, O, HG, A, K, O, MB. , U "
Will be entered. Considering the combinations of kana that occur, simple clauses include (1) carry boxes, (2) carry boxes, (3) carry boxes *, (4) box *, and box *. However, by checking the connection of independent words and annexes within each clause, the clause marked with "*" will be dropped. In order to conduct a connection test for independent words and annexes, prepare an independent word dictionary and an annex word dictionary, search for an independent word within the range of kana combinations that occur, and determine whether or not there is a combination with an annex word. . As a result, the phrase marked with "*" is dropped as words that cannot be combined.

On the other hand, the connection test between clauses is executed for the remaining clauses dropped by the connection test of independent words and adjunct words. Then, the verb "carry" is not connected to "box yo",
Only "carrying boxes" and "carrying" remain candidates. In other words, by considering the grammatical information such as an independent word or an adjunct word in order to carry out bunsetsu segmentation, further narrowing down as Japanese is performed.

The specific method of the continuous clause conversion is not particularly limited, and various kinds can be adopted. As an example, FIG. 25 shows an example of conversion processing by the minimum cost method.

In this minimum cost method, step S300
After initialization such as erasing the data temporarily stored in and initializing the analysis position to the first digit, step S3
At 05, processing for obtaining an analysis position is performed. What is the analysis position?
This is the position where the input sentence is analyzed next. For example, if a kana character string “Kuru ma wa ko ha kobu” is input, the first analysis position is the position of the first digit “KU”, and the analysis position becomes “ “M” ... In step S310, the CPU 71 performs a process of searching the independent word dictionary and the auxiliary word dictionary stored in the dictionary ROM 75 at this analysis position.
In the previous example, the word "ku" is retrieved from the dictionary.

After the dictionary is searched, step S315
The process of checking the combination with the previous word is performed for the word obtained in step 2. If there is no connection between the words, the word is invalidated, and the dictionary is further searched. For example, the word "ha", which is a particle in the preceding example sentence ("Kurumahakokohakobu"), searched for in the adjunct dictionary for "ha" of "koohahakobu" is the preceding particle " To
Since it is determined that there is no possible combination with and, it is regarded as an invalid word. When the dictionary search and the combination check at one analysis position are completed, the analysis positions are sequentially advanced and the processing is repeated.

For words having a possibility of being combined, the CPU 71 calculates the cost in step S320 and calculates the minimum total cost from the analysis position to the word. The cost is defined as "2" for the independent word and "0" for the annex. Next, in step S325, a cost check is performed to invalidate word combinations having a cost higher than the minimum total cost.

Explaining on the basis of the example shown above, the "car" in the example sentence ("Kurumahakokohakobu") is
It can be divided into various word combinations such as “ku” + “ru” + “ma”, “kuru” + “ma”, and “carma”. A word is applied to each of these combinations to calculate the cost. "Ku" + "Ru" versus "Bitter" (independent word) +
If the word "ryu" (independent word) is applied, "ryu" has a cost of 4. On the other hand, if the word "come" (independent word) is applied to "come", the cost is 2. Since the minimum cost method adopts the combination that minimizes the cost thus obtained, in this case,
"Come" will be adopted. If this analysis is continued, the "car" (independent word) of "car" becomes the minimum cost with cost 2.

Further, considering "until the coming",
Because it is "car" (independent word) + "de" (attached word),
The cost of “de” is the cost 2 corresponding to the total cost of “by car”. Similarly, considering the result of "come" (independent word) + "up" (adjunct), "up" also has a cost of 2.
The clause candidates and the costs thus obtained are stored in the work RAM 76.

In the next step S330, a process of linking the word candidates thus given the cost is performed. That is, by setting a pointer for a word for which the combination is valid, the combination is related. To explain "Kurumachi" in the above example sentence, the minimum total cost is calculated for "Car / By" and "Kuru / To".
The relationship is such that "come" is linked to "up" and "car" is linked to "de". The process of combining check, cost calculation, and linking is repeated until all words have been searched at one analysis position. Further, when the dictionary search at the analysis position is completed, the analysis position is further advanced by one, the establishment of a new word is examined, and similarly, the join check and the cost calculation are repeated.

If the analysis position reaches the position of the last kana character that has already been input and it is determined in step S335 that the analysis has been completed for all words, step S340.
On the premise of the above processing, the processing for searching the path with the minimum cost is performed. This is a process of searching for a combination of valid words that minimizes the sum of costs given to the words. In the example of "Walk until the car comes", "Car (2) / at (2) / box (4) / at (4) /
The total cost of the segmentation of "carry (6)" is 18, which is the minimum cost. The numbers in parentheses mean the cost of each word.

At this time, although the cost is not the minimum, another candidate for segment writing is searched. For example, "Car (2)
In /, (2) / child (4) / is (4) / carries (6) ”. In this way, after the segmentation candidate is created in step S345, the process of creating the candidate inside each phrase is performed in step S350. That is, within one phrase segment,
For example, candidates such as “box” and “box” are prepared for “Hako”.

[0135]

Furthermore, it is also possible to store the semantic connection between clauses as grammatical information. For example, as grammatical information, the information of the clauses having the relation of catching is stored in a dictionary and this information is stored. It is conceivable to refer to when converting Kanji. With such a configuration, it is possible to preferentially display the candidate character strings having high validity even between the separated phrases.

The case of inputting a character string mixed with Chinese characters "carry a load" will be described below as an example. The key character string based on the key operation is “NZJ, I, YWR, O, H
G, A, K, O, MB. , U ”, and even considering the part-of-speech connection test, there are five candidates: (9) Carry similar (10) Carry boiled (11) Carry load (12) Carry time (13) Carry character. Remain. On the other hand, in the dependent dictionary,
It is stored in advance that there are two clauses, "carry" and "carry", that are related to each other. By referring to the dictionary in which such information is stored, the acceptance test is performed on the five candidates, and as a result, "carry cargo" registered in the dictionary is regarded as a candidate character string with a high priority. To be done.

In addition to this example, the hanging acceptance test can be, for example, a test in which a person corresponds to the subject of a certain intransitive verb. Such an acceptance test
This is extremely useful in the case of this case, which is premised on narrowing down in two stages. Of course, these catches are merely examples, and it goes without saying that other tests can be performed.

These bunsetsu candidates and word candidates are used when the user instructs to change the bunsetsu division method or to display the next candidate.

Next, FIG. 26 shows an example in which the way of dividing the phrase segmentation is changed. FIG. 26 shows a case of inputting a character string mixed with Chinese characters such as "Sumomomo Momo_Momo Momo".
A state in which a kana character string such as "sumomomomomomomomomo" based on key operations is displayed after being segmented into phrases is displayed. When the phrase segmentation with the minimum cost is performed using the above-mentioned minimum cost method, the segment is divided into three segments as shown in FIG. In FIG. 26 (A), since the second bunsetsu is divided between the 7th and 8th characters, the second bunsetsu is separated by the 6th character in order to input a desired character string. Need to change between 7th character.

In order to change the punctuation of the bunsetsu, it is necessary to move the bunsetsu bunsetsu and change the bunsetsu length, but in order to implement these works, the bunsetsu bunsetsu of each input area 91 is marked. Functions such as moving and changing phrase length are assigned. This state is shown in FIG. By touching each of these cells, as shown in FIG. 22 (B), the target bunsetsu is moved from the first bunsetsu to the second bunsetsu, and the bunsetsu length is changed to
It is shortened as shown in FIG.

With the above arrangement, continuous clause processing is also possible. Of course, in the conversion within each phrase, the kana-kanji conversion processing as shown in FIGS. 23 and 24 may be applied to select the kanji candidate.

By the way, the Romaji-kana conversion has been performed so far so that it can be easily used for narrowing down. However, as a key operation, the kanji conversion is performed for a key operation having a certain flow. It is not impossible to perform Kanji conversion from a physical keyboard layout. In the example of "Haikei", a dictionary file in which "Dearing" is associated with the key character string "11, 5, 10, 9, 7, 10" may be prepared. The process of generating the key arrangement-Kanji file PRS as such a dictionary file will be described with reference to FIG.

First, the key arrangement-Kanji conversion processing (matching processing) is performed on the Kana-key arrangement file PQF corresponding to the key character string for inputting each kana-Kana arrangement file PQF and the Kana-Kanji file QRF corresponding to the Kanji dictionary KD. (Step S400). The contents of this matching process,
The above-mentioned "Haikai" will be described below as an example.

The kana-key arrangement file PQF contains "1
"Ha" and "ga" for the key string "1,5"
The kana character is registered for the key character string "10", and the kana character for the key character string "9, 7" is registered for the key character string "10". In addition, in the kana-kanji file QRF, kanji character strings such as "Dearing" are provided for kana character strings that say "Haikei", and "outline" for Kana character strings that are "gaikei."
Kanji character strings such as are registered respectively. In the matching process, first, the Kana-Kanji file QRF
Specify the kana character string "Haikei" registered in, and enter this kana character string as "ha", "i", "ke", "i".
These four kana characters are divided, and key character strings corresponding to these kana characters are arranged in order. As a result, "11,
The key character strings are arranged in the order of "5, 10, 9, 7, 10", and this key character string is associated with the kanji character string "Dearing".

Next, with reference to the Kana-key arrangement file PQF, it is determined whether or not kana characters other than "Haikei" are generated from the key character string "11, 5, 10, 9, 7, 10". to decide. For the key character string "11,5", the kana character "ga" is registered in addition to "ha", so the key character string "11,5,10,9,7,10" As a character string to be generated, "gaikei"
Is added. Next, the kana-kanji file QRF for the kana character string "gaikei" is referred to, and the kanji character string "outer shape" is extracted. As a result,
The key character string “11, 5, 10, 9, 7, 10” is associated with the kanji character string “outer shape”.

[11], which are thus associated with each other,
Key string "5,10,9,7,10" and "Dear Sir"
And a kanji character string "outline" are linked and stored in the key arrangement-kanji conversion file PRF. The above is the content of the matching process.

Next, in step S405, the key arrangement-
The key character strings stored in the Kanji conversion file PRF are sorted so as to be in a key arrangement order, and a key arrangement-Kanji conversion file PRS is created. The key arrangement-Kanji conversion file PRF is created through the above steps. FIG. 29 shows a simple flowchart of the creation process.

Therefore, by performing a prefix match search of this key arrangement-Kanji file PRF with the input key character string, it is possible to convert it into a Japanese character string containing Kanji. In this case, the key arrangement-Kanji file PRF is the basis. Since the kana-kanji dictionary KD is configured so that the kana-kanji conversion becomes natural Japanese, as a result, the key character string can be converted into natural Japanese.

The same method can be applied to so-called kana input, and kana may be partially overlapped and assigned to each square of the input key area 91.

If the lower left switching key is operated once with no input, the number input mode as shown in FIG. 30 is entered, and if it is operated twice, the alphabet input mode as shown in FIG. 31 is entered, and if it is operated three times, The symbol input mode is as shown in FIG. After that, when the switching key is further operated, the operation returns to the above-mentioned kana input mode, and the sequence is repeated in the following.

Next, in the PIM 80 of this embodiment,
The manner of inputting a Japanese character string using the automatic conversion mode will be described with reference to FIGS. 33 to 43.

The character string input screen 90 is as shown in FIG. 9 in a non-input state, and since the Japanese character string "Dearing" is input, it is assumed that "HAIKEI" is typed. "H" is assigned to No. 11 of the key character string in which "H" and "G" and the reading point "," are assigned redundantly. When this input key is operated, the candidate display area 92 is displayed as shown in FIG. Is displayed with a reading point "," that can be input independently, and "hg" assigned to the input key is displayed in the input key character string display area 93. Romaji is not converted to kana at this point. If the enter key is pressed here, steps S100 and S11 are performed.
Although the determination key is determined through the determinations in S2 and S114, the learning in step S118 and the collocation process in step S120 are substantially passed, and step S1 is performed.
21, the reading point “,” displayed in the candidate display area 92.
Is transferred to the confirmed character string display area 94. Then, when the transfer button 95 is pressed, it is output to the application.

Next, when the key assigned with "A" is pressed in the state shown in FIG. 10, at least the Roman letters can be converted into "ha" or "ga" (note that for convenience of explanation, the reading point "," The description of the kana character string ", a" generated from the combination with "A" is omitted), and is converted to kana in step S106 and kana-kanji conversion is performed in step S108. As a result,
As shown in FIG. 3, Chinese characters corresponding to “ha” and “ga” are displayed in the candidate display area 92. Here, "leaf" is the highest priority candidate, and is highlighted so that it can be easily identified.

Subsequently, when the key assigned with "I" is typed, it can be converted into "yes" or "gai" as a Roman character. Therefore, as shown in FIG. 34, "yes" and "gai".
The Chinese character corresponding to is displayed in the candidate display area 92. The kanji candidates corresponding to the reading of "yes" are displayed in yellow, the kanji candidates corresponding to the reading of "gai" are displayed in light blue to distinguish them, and the top priority candidate "ash" is highlighted. It is displaying.

When the cancel key is operated, the steps S100 and S112 are followed by the step S12.
In step 2, one character is deleted from the key character string, and step S104
Will return to. Therefore, the state just returns to the state shown in FIG. When the conversion key is operated once, it is converted into hiragana as shown in FIG. 35, and when the conversion key is operated twice, it is converted into katakana as shown in FIG. 36.

The conversion to katakana is not limited to conversion of all input kana to katakana, and only the last one character may be left in hiragana. For example,
This is because the conversion is often done with a particle added after Katakana, such as "Supermarket".

Although not shown in the drawing, by further operating the conversion key after this, full-width display without conversion (“hai”) or half-width display without conversion may be performed. Moreover, since candidates do not occur as much as Kanji conversion, all hiragana, katakana, and non-conversion may be displayed in the candidate display area 92 so that they can be selected, or they can be displayed by operating the conversion key after the display. You may make it change the order which it does. Note that, as shown in FIGS. 35 and 36, when the conversion key is pressed, the display of the conversion state becomes "candidate selection" so that the current situation can be understood.

Next, when the key assigned with "K" is typed, as shown in FIG. 37, the alphabetic character string "hg, a, i, k" is displayed in the input key character string display area 93. . Since this alphabetic character string is an alphabetic character string that cannot be converted into Roman characters, the candidate display area 92 may be replaced with the Chinese character such as "ash" or "outer" that was displayed before the keystroke, and "yes" or "gai". The character string is displayed. Next, when the key assigned with "E" is pressed, as shown in FIG. 38, "hg, a," is displayed in the input key character string display area 93.
An alphabetic string "i, k, e" is displayed. Since the kanji corresponding to the kana character strings such as “Haike” and “Gaike” generated by this alphabetic character string are not registered in the Kanji dictionary KD, “Haike” and “ga” are displayed in the candidate display area 92. Only the kana character string "Ike" is displayed. That is, in a roman character convertible delimiter, the Kana dictionary KD is referred to in step S108 using a convertible kana, but if there is no corresponding Kanji character string at this time, it is displayed in the candidate display area 92. It is only the result of conversion to Rukahana.

However, when there is a combination that can be converted to Kanji in a part of the plurality of combinations that are redundantly assigned at this time, for example, in the Kanji dictionary KD, “Haike” is displayed.
The kanji character string corresponding to the kana character string is not registered, but if the kanji character string "gaiya" corresponding to the kana character string "gaike" is registered, the natural Japanese character string is It is judged to be "outside house" and this is displayed first. FIG. 39 shows such a display. In addition, since it may be desired to input “Haike” after this, the character string that can be converted is always displayed. Specifically, “gaika” with a kanji candidate is given the highest priority and highlighted, followed by “haike”.

Finally, when the key to which "I" is assigned is pressed, the Kanji character strings corresponding to "Haikei" and "Gaikei" in the Romaji-kana conversion result are displayed as shown in FIG. Here, if the decision key is operated to select "dear", this candidate is learned in step S118.
That is, in the kana learning dictionary NGD, “11, 5, 10, 9, 7, 10, haikei” is written for “hakaii” which is the Roman character-kana conversion result corresponding to the above-mentioned key character string. In the Kanji learning dictionary KGD, “Hakaii, Dear Sir” is written as the kana-kanji conversion result. Further, since the operation of this decision key is a confirmation process, the kanji character string “Dearing” is transferred to the confirmed character string display area 94 in step S121. This state is shown in Figure 4.
Shown in 1.

In the automatic conversion mode, since a plurality of Kanji candidates are generated for the kana candidates that can occur, an extremely large number of candidates are displayed. However, one kana candidate should be divided by a fixed number. You may In this case, it is possible to display a kana display for a kana candidate with a kanji candidate still in the kanji candidate and display another kanji candidate for the kana candidate when the kana candidate is selected. is there.

The above is the processing in the automatic conversion mode, and the processing in the successive conversion mode will be described below, focusing mainly on the differences.

In the successive conversion mode, the conversion process is performed according to the flowcharts shown in FIGS. When inputting "Hayai", in the automatic conversion mode,
After converting from alphabetic characters to kana in step S106, the kana characters are converted to kanji in step S108 immediately. However, in the successive conversion mode, unless a control key other than the character key is pressed, the key character string is converted in step S206. Is converted to an alphabetic character, the alphabetic character is converted to a kana character in step S208, and the kana candidate is displayed in step S210. Therefore, when "hg, a" is input, the Chinese character candidates such as "leaf tooth ..." Are displayed in the automatic conversion mode as shown in FIG.
Only the converted "wa" and "ga" are displayed as shown in.

In order to input "Haikei", "hg,
The input keys are operated as "a, i, k, e, i", but finally, as shown in FIG. 43, two kana candidates "Haikei" and "Gaikei" are displayed. Here, if the candidate move key is operated, it is determined in step S202 that the control key is operated, and steps S211 and S2 are performed.
After step 212, the reverse display is changed to “gaikei” in step S214.

On the other hand, if the enter key is operated with the highlighted display set to "YES", steps S202 and S21 are performed.
After the judgments in 1 and S212, kana learning is executed in step S216. That is, as described above, "hg, a,
The operation of “i, k, e, i” is converted into “hakaii” and kana is written in the kana learning dictionary NGD, and thereafter, kanji conversion is executed in step S220. Also,
Even if the conversion key is not operated and the conversion key is operated with "Hakaii" highlighted, "Hayai" is selected and kana learning is performed, and Kanji candidates for "Haayei" are output. Good.

Kanji conversion is executed in step S230,
While referring to the kanji learning dictionary KGD and the kanji dictionary KD, the kanji candidates are displayed in step S232 while appropriately changing the priority display. In the automatic conversion mode, as shown in FIG. 40, the kana has not been decided, so the kanji candidates for “gaikei” will also be displayed, but in the successive conversion mode, the kanji candidates that correspond only to “hakaii” will be displayed. Will be displayed. After that, when the candidate move key is operated, the current candidate for reverse display is changed in step S236 through the determination in step S234, and when the enter key is operated, the kanji learning dictionary KGD is updated in step S238. That is, “dearing” is learned as a kanji conversion for “hakaii”. Since the operation of this decision key is a confirmation process, finally in step S240,
Confirm the kanji character string "Dear Sir" in the character string display area 94
41, and becomes as shown in FIG. 41 as in the case of the automatic conversion mode.

In both the automatic conversion mode and the successive conversion mode, the Japanese character string confirmed when the transfer button 95 is operated is transferred to the application, and the clear button 96 is operated. Discard the string.

As described above, the pressure-sensitive element 78b and the liquid crystal display 78 are used to form the character string input screen 90, and it is possible to input Roman characters while allowing overlapping allocation to each square of the input key area 91. When you try to input a Japanese character string by operating the input keys, the operator assigns an alphabetic character for inputting and generates a possible kana combination to display and select candidates with mixed kanji. However, at this time, by using the past conversion results and dictionaries to preferentially display the ones that have a natural Japanese sequence, the Japanese is narrowed down and all duplicated assignments are performed. It is much easier than searching for a combination.

On the character string input screen 90 as the first embodiment and the character string input screen 12 as the second embodiment described above, input keys for inputting characters (hereinafter referred to as character input keys) and In addition to the candidate movement keys, an input key (hereinafter referred to as a function key) for selecting a character string form such as a change key, a cancel key, a space key, and a decision key is provided, but the function keys are not provided. It is also possible to This configuration is used for the character string input screen 9 described above.
This will be described below in comparison with 0 and the character string input screen 12. FIG. 44 shows a screen (hereinafter referred to as an input start screen) displayed on the liquid crystal display 78 of the PIM terminal 80 when the input process of the Japanese character string is started.

On the input start screen, the character string input screen 22 of the third embodiment is displayed. Character string input screen 22
Includes an input key area 91, a candidate display area 92, and a confirmed character string display area 94, like the character string input screens of the first and second embodiments. Although the input key character string display area 93, which is an area for displaying the input character keys, is not provided in the third embodiment, the area may be provided.

The character string input on the character string input screen 22 is displayed as a document in the text development area 98 above the character string input screen 22 by operating the transfer button 95. Further, below the clear key 96, a help key 29 for displaying a method of inputting a character string on the character string input screen 22 is provided.

In the input key area 91, as in the case of the character string input screen 12 of the second embodiment, twelve keys are displayed in four rows vertically and three rows horizontally. The twelve keys are arranged in a region that can be operated by four fingers excluding the thumb of the five fingers of one hand, and are composed of ten character input keys 27a to j and two movement keys 28a and 28b. .

As with the character input keys 17a to 17j in the second embodiment, the character input keys 27a to 27j are assigned with a plurality of Roman characters and symbols in duplicate. In addition, the numbers from "1" to "9" are also input to each character input key 27
They are assigned to a to j, and the assigned numbers are displayed above each key.

In the third embodiment, like the character input keys 17a, b, c, f of the second embodiment, one character input key 27c is allowed to be assigned three or more Roman characters. By allocating in this way, the number of keys is reduced.

When the character input keys 27a to 27j are operated,
A key character string corresponding to the operated key is input, and this key character string is converted into Roman characters and symbols, numbers, letters, and symbols assigned to each key according to the setting of the input mode. When the input mode is set to the kana input mode, the converted roman characters and codes are converted into a kanji / kana mixed character string by referring to the romaji-kana conversion dictionary ND or the kanji dictionary KD. The converted character string is displayed in the candidate display area 92 as a candidate character string.

In the third embodiment, as in the first and second embodiments, some of the 26 letters of the alphabet are assigned to the respective character input keys 27a to 27j. For this reason, when the alphabet input mode is set, the display of the key tops of the respective character input keys 27a to 27j is changed to the mode in which the 26 letters of the alphabet are assigned redundantly. Even when the symbol input mode is set, the key top display is changed to the symbol display and displayed. This makes it possible to input alphabetic character strings and symbols. This method is similar to that already described with reference to FIGS. 31 and 32.

On the other hand, the types of Roman characters and symbols assigned to the character input keys 27a-j are different from those of the character input keys 17a-j in the second embodiment. This point will be described later.

The move keys 28a and 28b correspond to the candidate move keys of the second embodiment. That is, when the move keys 28a and 28b are operated in the state where the element to be selected is displayed in the candidate display area 92, the position of the reverse display band for reversing and displaying the specific element is moved up and down. The element highlighted by this movement is selected as the element desired by the user. For example, in the input start screen of FIG. 43, the element “waiting for input” is selected. Here, “waiting for input” means the character input keys 27a to 27j.
When you operate, it means that you are ready to enter the key string for the assigned character.

As described above, in the third embodiment, the first,
The switching key 18a, the cancel key 18b, and the space key, which are provided in the second embodiment, are not provided. This is because the input mode is switched and the key character string is deleted by selecting from the display of the candidate display area 92.

In the candidate display area 92, in addition to "candidate character strings after kana-kanji conversion", an element "options for selectable functions in the process of creating character strings (hereinafter referred to as function options)" is displayed. It In this embodiment,
Mode switching function to change the input mode, input waiting function,
A function to delete a key character string that has already been input (hereinafter referred to as the one-input deletion function), and a function to insert a space into a character string or symbol (hereinafter referred to as input data) such as a character string that has already been input with kanji and kana (hereinafter referred to as input data) Hereinafter, the space function), the function to delete characters and symbols from the input data (hereinafter referred to as the single character deletion function) and the function to edit the input data (hereinafter referred to as the
(Edit function) is displayed. By selecting one option from the display of this option, each selected function is executed.

In the initial input screen shown in FIG. 44, since the key character string has not been input yet, in addition to the input waiting function, there are four character deletion functions, a mode switching function, a space function and an editing function. The options for the function are displayed. After the key character string is input, the options for the above four functions are not displayed, but the options for the input waiting function and the one-input delete function are displayed. After that, when the Japanese character string is determined from the input key character string, the key character string returns to the non-input state, so that the above-mentioned four functions are displayed again instead of the one-input delete function.

As described above, in the third embodiment, the switching key 18a and the cancel key 18 are used to execute functions such as mode switching of the input mode and deletion of the input key character string.
It is not necessary to provide b, and a small number of keys can perform many functions associated with document creation. As a result, it is possible to create various types of documents.

Character string input screen 1 of the second and third embodiments
In Nos. 2 and 22, the number of alphabetic characters that exceeds the number of character input keys is assigned as Roman characters for Japanese input. A method of assigning a Japanese input element including such Roman letters to each character input key will be described below.

In the character string input screen 12 as the second embodiment shown in FIG. 11, a plurality of consonants which are combined with the same vowel and give rise to an adjective of a different kana character are input to different character input keys 17a. Assigned to j.
For example, the consonants "D", "G", "H", and "K" are all combined with the same vowel "A", so that the auxiliary verb "da" and the particle "ga", respectively.
Although the frequently used kana characters such as "ha" and "ka" are generated, in the second embodiment, these consonants are assigned to the different character input keys 17e, 17c, 17a, and 17f, respectively. Also, "D", "H", "N",
The consonant "T" is combined with the same vowel "E", so that "de", "he", and
Kana characters, which are the particles "ne" and "te", are generated.
In the second embodiment, these consonants are assigned to different character input keys 17e, 17a, 17b, 17c, respectively.

On the character string input screen 12,
The different vowels "A", "I", "U", "E", and "O" are assigned to different character input keys 17g, 17i, 17 respectively.
It is assigned to h, 17e, and 17j.

A process of generating a Japanese character string containing an auxiliary word of one kana character based on the operation of the character input keys 17a to 17j will be described with reference to FIG. FIG. 45 is an explanatory diagram showing a process of generating a Japanese character string “Shioda”. Each step in this process corresponds to each step performed in the Japanese character string generation process of FIG.

When the character input keys 17d, 17i, 17j, 17e, and 17g are operated in sequence after the Japanese character string generation processing is started, the key character string based on the character input key 17d based on the alphabetic character conversion processing in step S104. Is "S",
Character input to the alphabet "P", the key character string based on the character input key 17i is the character "I", and the key character string based on the character input key 17j is the character "O", ". (Period)" The key character string based on the key 17e is converted into the letters "E" and "D", and the key character string based on the character input key 17g is converted into the letters "A" and ", (comma)", respectively.

Subsequently, in the Roman character conversion process of step S106, the combination of these English characters is converted into kana, and four types of "shioea", "shioda", "pioeaa" and "pioda" are obtained. A kana kana character string is generated. At this time, in the fourth character input key 17e, a consonant other than "D", such as "G", "H", or "K", is combined with a vowel to generate a kana which is a supplementary word of one character. Since no consonant to be assigned is assigned, such as "shioga", "shioha", and "shioka" that include ancillary words that consist of one kana character such as "ga", "ha", and "ka". No kana character string is generated. Further, since the fifth character input key 17g is not assigned with "vowels other than A" such as "E", "shiode" that includes an adjunct consisting of one Kana character "de" A kana character string such as "is not generated.

These four kinds of kana character strings are converted into Japanese character strings mixed with kanji by referring to the kanji dictionary KD and kanji learning dictionary KGD in the kanji conversion process in step S108. Specifically, first, the independent word dictionary is referred to for each kana character string, and words that can be converted into kanji or kana are searched. As a result, of the four types of kana character strings that were generated, for the character strings "shioea" and "shioda", the independent word for the common noun "shio" was used for the character string "shioda". The independent words of the proper noun "Shioda" are searched respectively.

Since the retrieved independent word "salt" can be connected to an annex and constitute one clause, "shio" is excluded from the character strings "shioea" and "shioda". The supplementary word dictionary is referred to for the remaining kana characters "ea" and "da".

As shown in FIG. 18, the word "e" consisting of one kana character or the word "ea" is not an auxiliary word and is not registered in the auxiliary word dictionary. Therefore, it is determined that the kana character "ea" does not form a phrase by combining with the independent word "salt". As a result, only the word that is the case particle "da" is searched from the annex dictionary, and only the character string "shioda" of the above four kana character strings is called "Shioda" and "Shioda". Converted to a mixed Kanji character string.

The character string mixed with Chinese characters thus converted is displayed as a candidate character string based on the processing of step S110. In the case of the above example, as shown in FIG.
Two candidate character strings “Shioda” and “Shioda” are displayed in the candidate display area 92. After that, the desired character string "salt" can be input by moving the position of the reverse display by operating the candidate move key.

As described above, in the second embodiment, the character input key 1 which allows a plurality of Roman characters to be assigned to one key.
7a to j, and converts a plurality of English character strings converted based on the operation of these keys into a kana character string that can be read in Roman letters, and whether or not this kana character string is registered in the annex dictionary The kana character string is converted to a kanji mixed character string by referring to, and the words registered in the annexed word dictionary are extracted as candidate character strings after the kanji conversion. On the other hand, in the character input keys 17a to 17j, the Roman characters "D", "G", "H", and "K" are different character input keys 17 respectively.
Since it is assigned to e, 17c, 17a, and 17f, even if the character input keys 17a to f to which the consonant is assigned and the character input key 17g to which the vowel "A" is assigned are continuously operated, the alphabetic character conversion processing is performed. "DA", "GA",
Multiple alphabetic strings of "HA" and "KA" do not occur.

Therefore, the Roman character conversion process described later does not generate kana characters registered in the annex dictionary such as "ga", "ha", and "ka" in addition to "da".
By the kanji conversion process, "Shioda", "Shioga", "Shionoha",
Candidate character strings such as "salt" do not overlap. Therefore, according to the character string input screen 12 of the second embodiment, when there are a plurality of kana characters that can be an adjective in one character in the same row, it is possible to prevent duplicate occurrence of annex candidates. .

In the second embodiment, "A", "I",
The vowels "U", "E", and "O" are assigned to different character input keys 17g, 17i, 17h, 17e, 17 respectively.
Assign to j. Therefore, even if the character input key 17e to which the consonant is assigned and the character input key 17g to which the vowel "A" is assigned are continuously operated, an alphabetic character string such as "DE" occurs in addition to "DA". There is no such thing.

Therefore, there is no need to generate a kana character registered in the annex word dictionary “de” which is a case particle separately from the kana character “da” which is an auxiliary verb, depending on the Roman character conversion processing to be performed later. As a result of the kanji conversion processing, the candidate character string of mixed kanji "Shio de" does not overlap with "Shio Da." Therefore, even when there are a plurality of kana characters that can be an adjunct in a single character on the same line, it is possible to prevent duplicate adjunct candidates from occurring.

That is, according to the character string input screen 12, 1
It is possible to further narrow down the types of candidate character strings to be generated while assuming that a plurality of Roman characters are assigned to the key and the duplicated candidate character strings are generated. As a result, the number of candidate character strings displayed in the candidate display area 92 is reduced, and the user can easily select the candidate character strings.

In particular, when the character input keys 17a, b, c, f to which three or more Roman characters are assigned are operated, a large number of kana candidates can occur, but consonants that form an annex consisting of one kana character are generated. By assigning to different keys, it is possible to effectively narrow down the candidate character strings that contain mixed kanji. As a result, it is possible to realize easy-to-use character input even if three or more Roman characters are assigned.
There is no need to add additional character input keys to reduce the number of assigned characters.

The structure for prohibiting the assignment of Roman characters that generates two or more single Kana characters that are annexed words has been described above. In addition, in the second embodiment, even when two or more single kana characters, which are annexed dictionaries, are generated as a result of Roman character conversion, it is possible to narrow down to a candidate character string containing one kanji character. . This example will be described with reference to FIG. FIG. 47 is an explanatory diagram showing a process of generating a Japanese character string “Shioga”.

When the character input keys 17d, 17i, 17j, 17c, and 17g are operated in order to input the character string "Shioga", the fourth operated character input key 1
In addition to the consonant "G", other consonants such as "T" and "M" are redundantly assigned to 7c. Therefore, by the Romaji conversion processing in step S106, "shioga", " Kana character strings such as "Shiota" and "Shioma" are generated.

In the Chinese character conversion process of the next step S108, the independent words such as "Shioda" and "Shio" are searched from these kana character strings. Of these, the independent word "salt" can be connected with annexed words to form one clause,
"Ga", which is the remaining kana characters obtained by removing "shio" from the character strings "shioga,""shiota," and "shioma."
Refers to the annex dictionary for "ta" and "ma".

Here, the kana character "ma" is not an adjunct word and is not registered in the adjunct word dictionary. Therefore,
It is determined that the kana character "ma" does not form a phrase by combining with the independent word "salt". Therefore, the character string “Shioma” is excluded from the candidate character strings after the conversion of Kanji.

On the other hand, as shown in FIG. 18, the kana character "ga" is "ta" as a case particle or a connecting particle.
The kana character is an auxiliary verb that means past completion,
Each is registered in the annex dictionary. Therefore, the connection verification table described above for these two kana characters is referred to.

As shown in FIG. 47, in the connection test table, "ga" as a case particle is a noun, "ga" as a connection particle is a verb adnominal form, and "ta" as an auxiliary verb is a verb. It is stipulated that they should be connected together and that they should be connected together. By referring to the connection verification table, it is determined that the connecting particle “ga” and the auxiliary verb “ta” kana character do not form a phrase by combining with the noun “salt” independent word. Therefore, the candidate character strings "Shiota" are not converted, and only candidate character strings "Shioga" and "Shioda" are displayed in the candidate display area 92.

As described above, in the second embodiment, not only the supplementary word dictionary but also the connection verification table is referred to, and the conversion into a kanji-mixed character string is performed in units of clauses. Therefore, even if the Kana character, which is a plurality of one-character adjectives, is generated as a result of the Roman character conversion processing, the grammatical information of each Kana character registered in the adjective dictionary or the words registered in the connection verification table By referring to the grammatical information regarding the adequacy of the connection, it is possible to narrow down to one bunsetsu candidate.

The avoidance of duplicate occurrence of annexed words is as follows.
The character string input screen 22 of the third embodiment already shown in FIG.
Can also be realized by. This character string input screen 22
The arrangement of the character input keys in will be described below.

On the character string input screen 22, as in the character string input screen 12, a plurality of consonants which generate an appendix of a kana character by combination with a vowel are assigned to different character input keys 27a to 27j.

For example, as shown in FIG. 18, a kana one-character ancillary word belonging to Adan has a connecting particle "ba", an auxiliary verb "da", a case particle and a connecting particle "ga", and a dependent particle. There are “ha”, the final particle and the parallel particle “ka”, the final particle “sa”, the final particle and the parallel particle “ya”, and the consonants that make up these kana characters are “B” in Roman letters. , “D”, “G”, “H”, “K”, “S”,
It is "Y". Therefore, in the relationship with the vowel "A" in the Roman alphabet, these consonant Roman alphabets are assigned to different character input keys 27a, 27b, 27c, 2 respectively.
7i, 27g, 27h, 27j.

Similarly, there are case particles and parallel particles "ni" and connective particles "shi" in the adjuncts of the kana character belonging to Idan, and the consonants that make up these kana characters are Roman letters. They are “N” and “S”. Therefore,
In relation to the roman vowel "I", these consonant Roman letters are assigned to different character input keys 27g and 27h.

Kana, which belongs to Udan, is a supplementary word of the kanji character, which has the auxiliary verbs "nu" and "zu", and the consonants that make up these kana characters are "N", which is the Roman character, respectively.
It is "Z". Therefore, in relation to the vowel "U" in the Roman alphabet, these consonant Roman alphabets are assigned to different character input keys 27g and 27c.

The kana one-letter adjectives belonging to Edan include the case particles “de” and “he”, the final particle “ne”, the connecting particle “te”, and the final particle “ze”. The consonants that make up the kana kana character are the "D" in Roman letters,
“H”, “N”, “T”, “Z”. Therefore, in relation to the vowel "E" in the Roman alphabet, these consonant Roman alphabets have different character input keys 2
7b, 27i, 27g, 27f, 27c.

Kana belonging to Odan are the one-letter auxiliary words such as the particle "mo", the case particle and the parallel particle "no" and "to", the case particle "o", and the final particle "zo". The consonants that make up these Kana characters are the "M" in Roman letters,
“N”, “T”, “W”, “Z”. Therefore, in relation to the vowel "O" in the Roman alphabet, these consonant Roman alphabets are assigned different character input keys 2.
7j, 27g, 27f, 27h, 27c.

Further, in the character string input screen 22,
The vowels "A", "I", "U", "E", and "O" are assigned to different character input keys 27d, 27e, 27, respectively.
It is assigned to f, 27a, and 27b.

The processing performed when inputting Japanese using such character input keys 27a to 27j is almost the same as that of the second embodiment already described with reference to FIGS. 45 and 47. That is, even if the one character input key to which a consonant is assigned and the one character input key to which a vowel are assigned are operated in sequence, in the Roman character conversion process of step S106, a plurality of kana characters that are an adjunct of one character Never be generated over.

For example, in order to input the character string "Tokyo ga", the character input keys 27f, 27b, 27f, 27
When g, 27j, 27b, 27f, 27c, and 27d are operated, kana characters such as "ga", "pa", and "za" are displayed based on the eighth character input key 27c and the ninth character input key 27d. Only is generated, "da", "ha",
Adjectives consisting of one kana character such as "ka", "sa", and "ya" are not generated. Therefore, by the kanji conversion process,
"It's Tokyo", "Tokyo is", "Is it Tokyo?", "It's Tokyo",
The candidate character string "Tokyoya" does not overlap with "Tokyo ga". Since "pa" and "za" are not registered in the supplementary word dictionary, it is determined that the independent word "Tokyo" cannot be connected, and as a result, only the candidate character strings "Tokyoga" and "Tokyo" are It is displayed in the candidate display area 92.

The same applies when the kana belonging to a line other than A is generated from the eighth and ninth character input keys. For example, when the 8th and 9th character input keys 27b and 27a are operated, the kanji conversion process causes candidate character strings such as "to Tokyo" and "Tokyo Ne" in addition to "in Tokyo". When the character input key 27g and the character input key 27b are operated, candidate character strings such as "Tokyo", "Tokyo" and "Tokyo" do not occur in addition to "Tokyo".

Therefore, the character string input screen 22 according to the third embodiment.
According to this, the types of candidate character strings generated based on the operation of the character input key can be surely narrowed down, and the user can easily select the candidate character strings.

The arrangements of the character input keys shown as the second and third embodiments are merely examples, and prevent the occurrence of duplicated candidate character strings of annexes consisting of one kana character. Other key layouts may be used as long as

In the first to third embodiments described above,
In order to enable input of kana characters having a dull sound or a semi-voiced sound, Roman letters such as "D" in da line and "P" in pa line are also assigned to the character input keys. A key arrangement when the dakuten and the semi-dakuten instead of the roman letters are assigned to the character input keys will be described with reference to FIG. FIG. 48
Shows a character string input screen 32 as a fourth embodiment.

The character string input screen 32 has substantially the same configuration as the character string input screen 22 of the third embodiment. That is, the input key area 91 has ten character input keys 37a to 37j.
And a part of the 26 letters of the alphabet is assigned to these keys as Roman letters for Japanese input.

On the other hand, in the fourth embodiment, the character input key 37
The dakuten and semi-dakuten are assigned to different keys i and 37h, respectively. As a result, “G” is used to input kana of a line, “Z” is used to input a line, “D” is used to input a line, “B” is used to input a line, and “B” is input to a line. The five alphabetic characters “P” used are excluded from the assignment to the character input keys 37a to j, and a total of 14 alphabetic characters are assigned to the respective character input keys 37a to j as Roman letters for Japanese input.

Roman letters are redundantly assigned to the character input keys 37i and 37h to which dakuten and semi-dakuten are assigned. In the fourth embodiment, "T" is used for inputting the kana of da line and kana of romanized letters that are assigned to overlap with this dakuten.
And, "H" used to input kana of line, line and line
Is called a consonant. That is, the combination of "T" and "" (voiced sign) means the dakan kana, and the combination of "H" and "° (voiced sign)" means the pagana kana. In addition, the character input key 37h to which the semi-voiced point is assigned is input with "K",
The consonants "R" and "W" are assigned redundantly.

On the other hand, "A", "I", "U", "E",
The vowel "O" is a character input key 37d, 37e, 37f, 37a, 37 which is a key different from the dakuten and the semi-dakuten.
assigned to each b.

When the character input keys 37a to 37j are operated, an alphabetic character conversion process (step S104 in FIG. 5) is executed each time a key character string is input, as in the first embodiment.
The Romaji conversion process (step S106 in FIG. 5) and the kanji conversion process (step S108 in FIG. 5) are executed, and candidate character strings mixed with kanji corresponding to the operation stage of the character input keys 37a to 37j are displayed in the candidate display area 92. It is displayed (step S110 in FIG. 5).

In this embodiment, in the English character conversion process, consonant replacement process is performed in which the English character string including the dakuon code or the semi-voice code is replaced with another alphabet string. The contents of the consonant replacement process will be described with reference to FIGS. 49 and 50.
49 shows the consonant replacement processing routine, and FIG. 50 shows the structure of the consonant replacement table referred to in the consonant replacement processing routine. The consonant replacement processing routine is activated as an interruption processing of the alphabetic character conversion processing when the key character string is converted into the voiced character code or the semi-voiced character code in the alphabetic character conversion processing.

As shown in FIG. 49, when this routine is activated, first, the alphabetic character (hereinafter referred to as the immediately preceding alphabetic character) converted immediately before the voiced sound code or the semi-voiced sound code is designated (step S700), and immediately before. It is determined whether the alphabetic character of is a specific alphabetic character (step S710).

Specific alphabetic characters will be described. Japanese kana characters have a clear sound like "ka", "sa", "ta", and "ha" when they have no dakuten, but when they have dakuten or semi-dakuten, they have dakuon and half. There is a kana that produces a dull sound. An alphabetic character that represents a consonant that constitutes such a kana is defined as a specific alphabetic character. In the present embodiment, the specific alphabetic characters are four alphabetic characters “H”, “K”, “S”, and “T”, but other alphabetic characters or combinations of alphabetic characters or symbols may be specified as the specific alphabetic characters. Absent.

The determination as to whether or not it is a specific alphabetic character is made by referring to the index of the consonant substitution table shown in FIG. The consonant substitution table is provided as a sub-table of the key-alphabet table ET, and the index holds information of a specific alphabetic character among the alphabetic characters stored in the key-alphabet table ET.

[0230] If the immediately preceding letter is a specific letter,
With reference to the body of the consonant substitution table, the combination of the immediately preceding letter and the voiced code or the semi-voiced code is changed to another letter,
That is, a process of replacing with a letter representing a consonant constituting a kana having a dull sound or a semi-voiced sound is performed (step S720).
As shown in FIG. 50, "H", "K", "S", "T"
The combination of the English character "" and the dakuon code "" is
The letters "B", "G", "Z", and "D" are replaced with alphabetic characters that represent consonants that form a line, a line, a line, and a line, respectively. After this replacement process, another replaced alphabetic character is added as an alphabetic character string after alphabetic conversion (step S730), and this routine is terminated and returns to the alphabetic character conversion processing.

If the immediately preceding letter is not a specific letter,
It is determined that the immediately preceding letter is not combined with the voiced code or the semi-voiced code, and the conversion to the voiced code or the semi-voiced code is invalid in the combination with the immediately preceding letter (step S750). After this invalidation processing, the processing for determining the last alphabetic character as the alphabetic character string after alphabetic character conversion is performed (step S760), this routine is terminated, and the processing returns to the alphabetic character conversion processing.

Next, the process of actually generating a Japanese character string using the character string input screen 32 of the fourth embodiment will be described with reference to FIG.
1 and FIG. 52. FIG. 51 shows a result of generating a kana character string candidate by performing an alphabetic character conversion process or a roman character conversion process when the character input keys 37a to 37j are operated to input a Japanese character string "is a misery". FIG.

In order to input a Japanese character string mixed with Chinese characters, "It is a miserable loss", first, the character input keys 37a to 37j are input.
By operating, "za", "n", "pa", "i", "da"
It is necessary to enter the Roman characters that make up each kana. The character input keys 37a to 37j that need to be operated to input these five kana are shown on the left side of FIG. 51, each of which is divided into a first operation to a fifth operation. Figure 5
On the right side of 1, the Kana character string candidates generated as a result of the English character conversion processing, the consonant substitution processing, and the Roman character conversion processing based on the respective operations from the first operation to the fifth operation are shown.

First, the first operation for inputting the kana "Za" will be described with reference to FIG. Figure 5
2 (A), the character input keys are keys 37c, 37i, 37.
When the keys are operated in the order of d, the alphabetic character conversion process, the consonant substitution process, and the Roman character conversion process are performed according to the operation of each key.

When the character input key 37c assigned with a roman letter "S" is operated subsequently to the character input key 37i assigned with a dakuten (""), an alphabetic character conversion process is performed based on this operation. As a result, "S
"T", "SH", "MT", "MH", "S", "M"
Six alphabetic characters "" are generated (note that the combination of the character input key 37c with the period is omitted for convenience of description).

Of these, "ST", "MT", "MH"
Since the kana of 1 cannot be generated from these three alphabetic strings, these three alphabetic strings are excluded from the Roman character conversion processing target. That is, in the present embodiment, since the vowels are not assigned to the character input keys 37i to which the dakuten is assigned, only the two character input keys 37c and 37y are set as shown in FIG. 52 (C). At the time of operation, a kana-readable alphabetic character string such as "SI" does not occur. Therefore, even if the character input key 37i to which the dakuten is assigned is operated subsequently to the character input key to which a consonant is assigned, the Roman character conversion process and the Kanji character conversion process are executed, and the kana character string "shi" or " paper,
There is no possibility of generating a kanji character string such as "Mr. child".

Further, since the kana of 1 cannot be identified only from the two alphabetic strings "S" and "M", the Roman character conversion processing for these two alphabetic strings is suspended.
That is, in this embodiment, as shown in FIG. 52 (B), in order to input a kana called "za", the letter input key 37c to which the roman letter "S" is assigned is followed by the roman letter assignment "A". Since the operated character input key 37d is not operated, when only two keys are operated, "S
There will be no readable alphabetic characters such as "A". For this reason, Roman character conversion processing or Chinese character conversion processing may be executed in the middle of inputting a kana having a voiced sound, and a kana character string "sa" or a kanji character string "sand, difference, left" may be generated. Absent.

Of the two English character strings, the English character string "S" is replaced with the English character string "Z" by the consonant replacement process described above, as shown in FIG. 52 (A). On the other hand, "M
The English character string "" is excluded from the Roman character conversion target by the invalidation process shown in step S750 of FIG. As a result, among the six alphabetic strings generated by the alphabetic conversion process, only the alphabetic strings “Z” and “SH” are left as the character strings that can be converted into Roman characters.

After that, when the alphabet string "A" based on the operation of the character input key 37d is combined with the alphabet strings "Z" and "SH" to generate the alphabet strings "ZA" and "SHA". The Romaji conversion process and the Kanji conversion process are executed to convert the Kana-Kana dictionary into the characters "za" and "sha" and the Kana-Kanji dictionary KD to refer to "Za" and "Ka". , "Za", "oblique", and "person" are converted into kanji.

The first operation has been described above. Next, the second to fifth operations will be described with reference to FIG. In the second operation, the character input key 37
g and 37g are the character input keys 37 in the third operation.
i, 37h, and 37d are operated by the character input key 37e in the fourth operation, so that
Kana character strings such as "pa" and "i" are generated.

Here, since the character input key 37h operated by the third operation is not assigned a dakuten which overlaps with the semi-voiced point, the combination of the character input keys 37i and 37h causes the letter "H °" to appear. In addition to rows, "H" and "T"
There is no such a character string. Therefore, a character string such as "ba" or "da" other than "pa" is generated,
Candidate character strings such as "ba, horse, auntie" and "fall, hit" never stand.

In the fifth operation, the character input key 37
When i, 37i, and 37d are operated, kana character strings "da" and "ba" are generated. That is, since "T or H" and "" (voiced code) are overlapped and assigned to the character input key 37i, by operating one character input key 37i continuously, "T" (=
D) "and" H "(= B) are generated. Then, by operating the character input key 37d, "DA" which means "Da", "BA" which means "ba", "TTA" which means "ta", and "ha "HHA", which means "," is generated.

By the above first to fifth operations,
"Zanpaida", "Zanpaiba", "Zanpaita",
Kana character string candidates such as “shampaida” and “shampaiba” are generated. After that, a kanji conversion process is performed on these kana character strings. For kana character strings such as "Zanpaiba" and "Zanpaita", the noun "Zanpai" and the auxiliary verbs "Ba" and "Ta" were determined to be unconnectable by the connection verification table, and For the kana character strings such as "Nampaida" and "Shampaiba",
It is determined that there is no kana character string that can be converted to Kanji in the independent word dictionary. As a result, a mixture of kanji characters such as "disastrous defeat" and kana characters such as "Zampaida", "Zanpaiba", "Zanpaita", "Shampaida", and "Shampaiba" The candidates are displayed in the candidate display area 92.

As described above, according to the character string input screen 32 of the fourth embodiment, the Roman characters for Japanese input are assigned to the ten character input keys 37a to 37j, and the dakuten is set to the character input key 37i. The semi-voiced points are assigned to the character input keys 37h. Therefore, when inputting a kana that has a voiced sound or a semi-voiced sound in Roman characters, "B", "D", "G",
It is not necessary to search the character input keys 37a-j for the keys to which the five types of consonants such as "Z" and "P", which have voiced sounds and semi-voiced sounds, are assigned, and can be easily input. Further, by not assigning these five types of consonants to the character input keys 37a to 37j, the number of characters assigned to the character input keys 37a to 37j is reduced, and the number of keys accompanied by duplicate assignment can be reduced. As a result, character input key 3
It is possible to prevent the generation of many candidate character strings based on the operations of 7a to 7j.

Further, since the Roman characters for Japanese input are redundantly assigned to the character input keys 37i and 37h to which the dakuten and the semi-dakuten are assigned, independent input keys for inputting the dakuon and semi-dakuten symbols are used. No need to provide. As a result, it is possible to realize character input with a small number of input keys. On the other hand, since the dakuten is assigned to a key different from the character input key to which a vowel is assigned, as shown in FIG. 52 (C), in the English character conversion process, "SI" or "SIA"
It is not converted into an English character string such as, but is intended by the Roman character conversion process or the Kanji character conversion process such as "shi", "shia" kana character string, "paper, Mr., child", "koa, jia", etc. A candidate character string that does not exist does not occur. As a result, the user can easily select the candidate character string.

Further, according to the fourth embodiment, the consonant substitution process described above identifies the consonant "Z" and the consonant "Z" from the combination of the consonant "S" and the voiced consonant code "". For the combination of vowel and "A", the Roman character-kana conversion dictionary ND is referred to generate the kana "za". Therefore, in the process of inputting a "kana", a string of English letters "SA" in which consonants and vowels are consecutive is generated, and undesired "sa line" candidates are converted into Roman-Kana conversion dictionary ND Yakana-Kanji conversion. Never look up from a dictionary. Therefore, an efficient dictionary search can be realized.

Further, since the consonant "T" used for inputting the kana of a line and the dakuten are assigned to the character input key 37i in a duplicated manner, the same character input key 37i can be operated continuously to obtain the key character. A column can be converted into the alphabet "T" (= D), which is a consonant that represents a row.
Therefore, it is possible to easily perform the key operation necessary for inputting the voiced sound.

In the fourth embodiment, the consonant having the voiced sound "D" is specified from the alphabetical character string "T" consisting of the consonant "T" and the voiced sound. It is also possible to have a configuration that specifies a consonant having a dull sound from a simple combination with. For example, "T" in the alphabetic string "TA"
It may be possible to specify the consonant "D" from "". By doing so, even if the character input key is accidentally operated in the order of the consonant (T), the vowel (A), and the dakuten ("), the alphabetic character string" TA "is changed to" D ".
It can be replaced with a readable alphabetic character string such as "A", and the user does not need to redo the key operation.

The consonant replacement process shown in FIG. 49 is merely an example, and may be another process for identifying a consonant having a dull sound from a combination of a consonant having a clean sound and a dakuten or a semi-dakuten. .

Next, another key arrangement in which the dakuten and the semi-dakuten are assigned to the character input keys will be described with reference to FIG. FIG. 53 is a screen 4 for character string input as the fifth embodiment.
2 is shown. The character string input screen 42 is provided with ten character input keys 47a to 47j like the character string input screen 32 of the fourth embodiment, and each of these keys has a "G",
Romaji for Japanese input consisting of 14 letters of the alphabet excluding "Z", "D", "B", and "P", and a dakuten and a semi-dakuten are assigned.

On the other hand, in the fifth embodiment, in the above-mentioned consonant replacement process, "H" is replaced with a specific alphabetic character such as "B", "G", "Z", "D" by combining with a dakuten. , "K", "S", and "T" are assigned to different keys of character input keys 47g, 47j, 47c, 47h, respectively.

Also, in the fifth embodiment, as in the fourth embodiment, the dakuten or semi-voiced voice and the roman letters are assigned to the character input keys 47i and 47h in an overlapping manner, while the roman letters assigned to the dakuten and "N" are assigned. Let us say that the consonants are "" and the Roman letters assigned to the half-voiced points are consonants "T" and "W".

Using such a character string input screen 42,
The process of actually generating a Japanese character string will be described with reference to FIG. FIG. 54 shows that, when the character input keys 37a to 37j are operated to input the Japanese character string "dehydration symptom", alphabetic character conversion processing, consonant replacement processing, and Roman character conversion processing are performed, and kana character string candidates are selected. It is explanatory drawing which shows the produced | generated result.

In order to input the Japanese character string of "dehydration symptoms" mixed with Kanji, first, the character input keys 37a-
Operate j to select "da", "tsu", "su", "i",
It is necessary to enter the Roman characters that make up each kana, such as "shi", "yo", "u", "ji", "yo", and "u". The character input keys 47a to j that need to be operated to input these five kana in Roman characters are respectively set to the first
The operation is divided into the seventh operation and shown on the left side of FIG. 54.
On the right side of FIG. 54, candidates for the kana character string generated as a result of the Roman character conversion processing based on the respective operations from the first operation to the seventh operation are shown. Hereinafter, the first operation, the second operation, and the fourth operation that lead to characteristic results will be described.

First, the first operation will be described. When the alphabetic character conversion processing is performed based on the operation of the character input keys 47h, 47i, 47d, the key character string of each key is converted into an alphabetic character string "T" A "or" WA ". To the character input key 47h, only "T" is assigned among the characters representing consonants that are made into dull sounds by adding dakuten points "H", "K", "S", and "T". In addition to "TA", "HA", "KA", and "S"
A readable alphabetic character string such as "A (za)" does not occur.

Of these two alphabetic strings, the consonant replacement processing described above is performed on the alphabetic strings "T" and "W", and as a result, the alphabetical string "TA" is "DA". , And the alphabetic string “WA” is invalidated as a combination that cannot produce a kana. Therefore,
Romaji conversion processing is performed on one English character string "DA" to generate one kana character string "DA".

Next, the second operation will be described. When the alphabetic character conversion processing is performed based on the operation of the character input keys 47c, 47c, 47f, the key character string of each key is "SS
It is converted into an alphabetic character string such as "U" or "MMU". That is, in the present embodiment, since the dakuten is assigned to the character input key 47i together with the consonant "N" that does not generate a consonant by overlapping twice, "SU (Z ) ”Does not produce a readable Kana string. Therefore, Romaji conversion processing is performed on the two alphabetic strings "SSU" and "MMU",
Two types of kana character strings, "susu" and "tsumu" are generated.

Next, the fourth operation will be described. When the alphabetic character conversion processing is performed based on the operation of the character input keys 47c, 47j, 47b, the key character string of each key is "SY".
It is converted into an alphabetic character string such as "O" or "MYO". That is, in the present embodiment, since the voiced sound point is assigned to the character input key 47i together with the consonant "N" which cannot be a half vowel, "S0" is generated based on the operation of the character input key 47j.
There will be no Kana-readable alphabetic character strings such as (zo). Therefore, the Roman character conversion process is performed on the two English character strings "SYO" and "MYO", and two types of kana character strings "SHO" and "MYO" are generated.

Kana character strings such as "Daisuishojo", "Daimaishoujojo", "Daisuimeijojo", etc. are generated based on the above first to seventh operations. Thereafter, the kana character conversion process is performed on these kana character strings, and candidates for the kanji character strings “dehydration symptom” and “dehydrated Akeboshi” are displayed in the candidate display area 92.

As described above, according to the character string input screen 42 of the fifth embodiment, the consonant sound which is made into a dull sound by adding the dakuten points "H", "K", "S", and "T" is added. Different character input keys 47g, 47j, 47c, 4
Assign to 7h. Therefore, the character input keys 47a-47
A combination of 1 of j does not generate a plurality of kana characters having a dull sound.

Further, on the character string input screen 42, the consonant and the dakuten which generate a consonant by overlapping twice are assigned to different character input keys. Therefore, the character input keys 47a-
The combination of 1 of 47j does not generate a kana character string having a consonant and a kana character string having a dull sound in an overlapping manner.

Further, in the character string input screen 42,
The half vowels such as "Y", "W", and "H" and the dakuten are assigned to different character input keys. Therefore, the character input key 4
The combination of 1 of 7a to 47j does not generate a kana character string having a jumble sound and a kana character string having a dull sound in an overlapping manner.

As a result, it is possible to prevent the user from displaying a large number of candidate character strings mixed with Chinese characters, which is unintended by the user, and it becomes easy to select the candidate character strings.

When the kana characters such as "tsu" and "shi" are represented by Hebon-type Roman spelling, three Roman characters such as "TSU" and "SHI" are used. In the present embodiment, the consonants "S" and "H" located between the consonant and the vowel of the three Roman letters are assigned to the character input keys 47c and 47g different from the dakuten. Therefore, by combining 1 of the character input keys 47a to 47j, a kana character string of "tsu" or "shi" and a kana character string of "zu" or "ji" having a dull sound are not generated redundantly. By this, too, the randomization of the candidate character strings can be prevented.

The method of assigning Japanese input elements such as Roman letters and dakuten to each character input key has been described above. Such assignment can be widely applied to a character input device that converts a key signal into a Japanese character string containing kanji and kana. For example, you can convert a key string of a character input key into an alphabetic character string, convert this alphabetic character string into a mixed kanji character string without converting it into a kana character string, and use it in a character input device that inputs a Japanese character string. Can be applied. That is, in such a device, a dictionary that is referred to when converting an English character string into a mixed Kanji character string may be configured to store information on independent words or annexes, or may be used to distinguish between English characters and dakuten. By adopting a configuration in which a combination is converted into a character string containing kanji by referring to a dictionary, it is possible to obtain the same effects as those of the second to fifth embodiments described above.

Further, in the second and third embodiments, the key arrangements in which the consonants forming the one kana character which is an accessory word are assigned to different character input keys are used in the fourth and fifth embodiments. Although a key layout in which a dakuten and a semi-dakuten are assigned to keys in addition to Roman letters is shown, it is of course possible to use a key layout in which the second, third and fourth and fifth embodiments are combined.

The character input device according to one embodiment of the present invention has been described above, but an embodiment as a recording medium recording a program executable by a computer can be considered as another embodiment. That is, if the array data of the character input keys described above and the execution program for displaying the array data based on the execution instruction of the CPU are recorded in a ROM or a CD-ROM as a recording medium, this ROM is stored. The contents recorded in the above are installed in a computer using a display device capable of inputting characters as a device driver, so that the arrangement of the character input keys in the second to fifth embodiments described above can be displayed in a display device connected to the computer. It becomes possible to display.

Further, as an execution program, a process for generating a kana character which is a dakuon from a continuous combination of a roman character assigned to a character input key and a dakuten, is stored in the ROM.
If it is recorded in, it becomes possible to easily input the kana which is a dull sound in romaji by executing this processing at the stage of romaji conversion processing.

Although the present embodiment is premised on inputting Japanese, Japanese is only an example of an indirect input language, and it can be applied to various languages that cannot be directly input by a general keyboard. Needless to say.

[Brief description of drawings]

FIG. 1 is a diagram corresponding to claims of a character input device according to an embodiment of the present invention.

FIG. 2 is a hardware block diagram of a character input device.

FIG. 3 is an external view of a PIM terminal to which a character input device is applied.

FIG. 4 is a diagram showing a module configuration of a character input device.

FIG. 5 is a flowchart showing a Japanese character string generation process A.

FIG. 6 is an external view of a desktop computer to which a character input device is applied.

FIG. 7 is an external view of a mobile phone terminal to which a character input device is applied.

FIG. 8 is an external view of a remote control terminal to which a character input device is applied.

FIG. 9 is a diagram showing a character string input screen as a first embodiment of the character input device.

FIG. 10 is a diagram showing a character string input screen as a first embodiment of the character input device.

FIG. 11 is a diagram showing a character string input screen as a second embodiment of the character input device.

FIG. 12 is a diagram showing a character string input screen as a first embodiment of the character input device.

FIG. 13 is a diagram showing a key-alphabet table.

FIG. 14 is a diagram showing a Roman-Kana conversion dictionary.

FIG. 15 is a diagram showing a kana learning dictionary.

FIG. 16 is a diagram showing a Kanji dictionary.

FIG. 17 is an explanatory diagram showing the structure of an independent word dictionary.

FIG. 18 is an explanatory diagram showing a structure of an annex dictionary.

FIG. 19 is an explanatory diagram showing a connection verification table.

FIG. 20 is a diagram showing a Kanji learning dictionary.

FIG. 21 is an explanatory diagram showing a state in which two types of candidate character strings having different readings are displayed in the candidate display area 92.

FIG. 22 is a diagram showing a collocation dictionary.

FIG. 23 is a flowchart showing Japanese character string generation processing B.

FIG. 24 is a flowchart showing Japanese character string generation processing B.

FIG. 25 is a flowchart of a continuous phrase conversion process.

FIG. 26 is a diagram showing a process of adjusting a phrase length in continuous phrase conversion.

FIG. 27 is a diagram showing a phrase length adjusting operation key in continuous phrase conversion.

FIG. 28 is a block diagram showing a process of generating a key arrangement-kanji file.

FIG. 29 is a diagram showing a flowchart for generating a key arrangement-kanji character file.

FIG. 30 is a diagram showing a character string input screen in the numeral input mode of the character input device.

FIG. 31 is a diagram showing a character string input screen in the alphabet input mode of the character input device.

FIG. 32 is a diagram showing a character string input screen in the symbol input mode of the character input device.

FIG. 33 is a diagram showing an input state of the first character on the character string input screen in the conversion process of the character input device.

FIG. 34 is a diagram showing an input state of the next character on the character string input screen in the conversion process of the character input device.

FIG. 35 is a diagram showing a state where hiragana conversion is performed on the character string input screen in the conversion process of the character input device.

FIG. 36 is a diagram showing a state in which katakana conversion is performed on the character string input screen in the conversion process of the character input device.

FIG. 37 is a diagram showing a state in which the last input cannot be converted to Roman characters on the character string input screen in the conversion process of the character input device.

FIG. 38 is a diagram showing a state in which there is no Kanji candidate even though a roman character conversion delimiter is reached on the character string input screen in the conversion process of the character input device.

FIG. 39 is a diagram showing a state where there is a kanji candidate for one kana combination on the character string input screen in the conversion process of the character input device.

FIG. 40 is a diagram showing a final state on the character string input screen in the conversion process of the character input device.

FIG. 41 is a diagram showing a character string input screen in a conversion completed state of the character input device.

FIG. 42 is a diagram showing a display state in the successive conversion mode on the character string input screen in the conversion process of the character input device.

FIG. 43 is a diagram showing a display state immediately before Kanji conversion in the successive conversion mode on the character string input screen in the conversion process of the character input device.

FIG. 44 is an explanatory diagram showing a character string input screen as a third embodiment of the character input device.

[FIG. 45] Using the character string input screen of the second embodiment,
It is explanatory drawing which shows the process in which the Japanese character string "It is salt" is produced | generated.

FIG. 46 is an explanatory diagram showing a state where a candidate character string corresponding to the operation of the character input key is displayed on the character string input screen of the second example.

[FIG. 47] Using the character string input screen of the second embodiment,
It is explanatory drawing which shows the process in which the Japanese character string "Shioga" is produced | generated.

FIG. 48 is an explanatory diagram showing a character string input screen as a fourth example of the character input device.

FIG. 49 is a flowchart showing a consonant replacement processing routine.

FIG. 50 is an explanatory diagram showing a consonant substitution table.

FIG. 51 is an explanatory diagram showing a result of Roman character conversion processing in the case of inputting a Japanese character string “It is a misery”.

FIG. 52 is an explanatory diagram showing a difference in conversion processing between the case of using a combination of character input keys of the fourth embodiment and the case of using another combination of character input keys.

FIG. 53 is an explanatory diagram showing a character string input screen as a fifth embodiment of the character input device.

FIG. 54 is an explanatory diagram showing a result of Roman character conversion processing when a Japanese character string “dehydration symptom” is input.

[Explanation of symbols]

10 ... Key input means 12 ... Character string input screen 17a-j ... Character input keys 18a ... Switching key 18b ... Cancel key 20 ... Control means 22 ... Character string input screen 24 ... Input mode display area 27a-j ... Character input keys 28a ... Upward movement key 28b ... Downward movement key 29 ... Help key 30 ... Candidate generating means 32 ... Screen for inputting character strings 37a-j ... Character input keys 38a ... Upward movement key 38b ... Downward movement key 40 ... Display means 42 ... Screen for inputting character strings 47a-j ... Character input keys 48a ... Upward movement key 48b ... Downward movement key 50 ... Selection control means 60 ... Learning control means 70 ... Computer system 71 ... CPU 72 ... Bus 73 ... Keyboard 73a ... I / O 74 ... Program ROM 75 ... Dictionary ROM 76 ... Work RAM 77 ... Battery backup RAM 78 ... Liquid crystal display 78a ... I / O 78b ... Pressure sensitive element 80 ... PIM terminal 81 ... Desktop PC 81a ... Full keyboard 81b ... Numeric keypad 82 ... Mobile phone terminal 82a ... key 82b ... Display section 83 ... Remote control terminal 83a ... Channel button 83b ... Light receiving unit 90 ... Screen for inputting character strings 91 ... Input key area 92 ... Candidate display area 93 ... Input key character string display area 94 ... Confirmed character string display area 95 ... Transfer button 96 ... Clear button 98 ... Text development area

─────────────────────────────────────────────────── ─── Continuation of the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G06F 17/21-17/24

Claims (29)

(57) [Claims] 1. A key input unit having a plurality of input keys, wherein a Japanese input element based on a combination of a vowel and a consonant is allowed to be assigned to each of the input keys, and a key input unit is assigned to the input keys. When inputting Japanese by referring to the dictionary assuming the Japanese input element, the character string that can occur in the Japanese input element assigned to the same input key is converted into the Japanese Kana character string. Japanese conversion control means for performing conversion and second-stage conversion for converting the Japanese kana character string into a Japanese kanji mixed character string, wherein the Japanese conversion control means has the results of carried out transformation, and Do list storing means or storing the kana character string in the conversion of the first stage corresponding to the finally determined kanji mingled string, stored in the kana sequence storage means Wherein either the Do string, the character input device and a first converting means for converting the next time and high priority in the conversion of the first stage. 2. The character input device according to claim 1, wherein the Japanese input element based on a combination of a vowel and a consonant assigned to the key input unit is a Roman character, and the Japanese conversion control unit. Is a Roman character kana conversion in the conversion of the first step and converted into a combination of kana. 3. The character input device according to claim 1 or 2, wherein the first conversion unit is a kana character string that can be generated by the Japanese input element as the conversion with the higher priority. A character input device comprising means for preferentially displaying the result of the second-stage conversion using the kana character string stored in the kana sequence storage means as a candidate. 4. The character input device according to claim 3, wherein the kana sequence storage means is capable of retrieving a kana-kanji conversion file storing a kana sequence of Japanese and a kana sequence corresponding to the reading of the kana-kanji conversion file. A character input device characterized by storing a sequence. 5. The character input device according to any one of claims 1 to 4, wherein the Japanese conversion control means uses the converted Chinese character mixture character string as a Japanese candidate at the conversion in the second step. Candidate generating means for generating, display means for displaying the converted Japanese candidates, and selection control for selecting one candidate from the Japanese candidates displayed by the displaying means based on a user's operation. A character input device comprising: means and learning control means for preferentially subjecting the selection result by the selection control means to the conversion in the next second conversion. 6. The character input device according to any one of claims 1 to 4, wherein the Japanese language conversion control unit includes a candidate generation unit that generates a converted Kanji / Kana mixed character string as a Japanese candidate. Display means for selectively displaying the converted Japanese candidates, the display means displaying the Japanese candidates in a different manner for each combination of kana that can occur from the Japanese input element. Character input device. 7. The character input device according to claim 1, wherein the Japanese conversion control unit includes a grammar information storage unit that stores grammatical information of words in the dictionary, and the first unit. Phrase segmentation and writing means for segmenting a kana character string obtained by the step conversion into phrase units based on the stored grammatical information, and a connection between the segmented phrases based on the stored grammatical information. A character input device, which is provided with a connection verification means for performing a verification, and is a means for converting into a character string containing the Japanese kanji by performing the verification by the connection verification means. 8. The key input means comprises a plurality of input keys.
-Use some of the keys provided on the device.
It is a means configured by using one of claims 1 to 7.
The character input device described in the slip. 9. The character input according to any one of claims 1 to 7.
In the input device, the key input means is a display means capable of displaying the Japanese input element in the form of characters.
And the display means provided integrally with the display means.
By contacting a specific part of the
And a character displayed on the part specified by the part specifying means
The Japanese input element corresponding to the character
It is characterized by comprising input element input means for inputting
Character input device. 10. The character input device according to claim 9.
The display means has at least an area in which the characters are displayed.
And a transparent pressure-sensitive film for covering the transparent pressure- sensitive film.
A character input device that is a means of identifying a part. 11. The character according to claim 1 to claim 7.
On the input device, enter the array of input keys to which the Japanese input elements are assigned.
A key layout display means for displaying is provided, and the key input means has a specific input on the display by the display means.
When a force key is designated, it is assigned to the designated key
Character input device that is a means for inputting the specified Japanese input elements
Place 12. The key input means is used for the specific input.
A hand that uses a pointing device to give key instructions
The character input device according to claim 11, which is a stage. 13. A day based on a combination of vowels and consonants
A key that arranges multiple input keys to which the original language input elements are assigned.
Keyboard, and the input keys arranged on the keyboard.
-Enter a mixed Kanji character string based on the operation
A character input device , wherein the Japanese input element to the input key
Sentences that allow overlapping assignments and result from the duplicated Japanese input elements
Convert a character string into a Japanese kana character string by referring to a dictionary
The kana character string conversion means and the kana character string converted by the kana character string conversion means
Convert to a character string mixed with Japanese Kanji by referring to a dictionary
A kana-kanji conversion means, and the kana- character-string conversion means refers to a dictionary to display candidate kana-character strings.
The final decision of the Han from the results of the conversion performed in the past
Memorize the kana character string corresponding to the mixed character string,
The stored kana character string is given priority in the next conversion.
A character input device comprising: the means as the candidate. 14. An operation of an input key of a key input means is detected.
A method for inputting characters in Japanese by using multiple input keys and combining vowels and consonants
Japanese input elements based on are assigned to each input key
From the keyboard that allowed duplicate assignment to
Input the real language input element and assume the Japanese input element assigned to the input key.
And enter the Japanese by referring to the dictionary
The character string that can occur in the assigned Japanese input element
The first stage conversion to convert to the Japanese kana character string
To convert the kana string of words in Chinese characters mingled with a string of Japanese
It is assumed that the second stage conversion is performed , and the final decision is made from the results of the conversions performed in the past.
In the conversion of the first step corresponding to mixed character strings
The kana character string is stored, and the stored kana character string is converted in the first step.
Character input for higher conversion priority and subsequent conversion
Method. 15. The operation of the input key of the key input means is controlled.
A character input control program that allows the computer to detect and input Japanese.
A medium in which a program is recorded, having a plurality of input keys and combining vowels and consonants
Japanese input elements based on are assigned to each input key
From the keyboard that allowed duplicate assignment to
Assuming the function to input real language input elements and the Japanese input elements assigned to the input keys.
And enter the Japanese by referring to the dictionary
The character string that can occur in the assigned Japanese input element
First-stage conversion to convert to the kana character string of the main language and the conversion
Converts a kana-kana character string into a character string mixed with Japanese kanji
The function to perform the second-stage conversion and the result of the conversion performed in the past.
In the conversion of the first step corresponding to mixed character strings
The function of storing the kana character string and the conversion of the stored kana character string in the first step
Function and perform the conversion from the next time onwards.
A medium that stores a program to be realized by a computer. 16. A part of 26 letters of the alphabet
Of the input key that is assigned as the Roman alphabet for Japanese input
A dictionary that stores words and grammatical information of the words based on the operation
Create a Japanese character string while referring to it and enter the character string
A method of inputting characters , wherein two or more Roman characters are assigned to one key,
And the consonants that make up the adjuncts "ga", "da", and "ha"
Operation of the input keys that are assigned to different keys
Based on , the combination of Roman letters assigned to each input key
A step of converting into a permissible kana character string, and the converted kana character string with reference to the dictionary.
Character with a step of converting into a candidate character string mixed with kana
input method. 17. A part of 26 letters of the alphabet
Of the input key that is assigned as the Roman alphabet for Japanese input
A dictionary that stores words and grammatical information of the words based on the operation
Create a Japanese character string while referring to it and enter the character string
Is a recording medium recording a computer program for
Then, assign two or more of the above Roman letters to the key of 1,
And the consonants that make up the adjuncts "ga", "da", and "ha"
Operation of the input keys that are assigned to different keys
Based on , the combination of Roman letters assigned to each input key
A step of converting into a permissible kana character string, and the converted kana character string with reference to the dictionary.
A program for causing a computer to execute the process of converting into a kana-mixed candidate character string is compiled.
A recording medium that is readable by a computer. 18. The key input means is for inputting the plurality of inputs.
Divide the dakuten mark as one of the Japanese input elements into one of the keys.
The character input device according to claim 2, which is applied. 19. The input key, the Roman character
Claims that are assigned by excluding consonants that represent dakuon at least
18. The character input device as described in 18. 20. The key input means is arranged such that
Assigned to at least one of the Roman alphabets
The character input device according to claim 18. 21. The key input means is for inputting the Roman characters.
Assign the vowels and the dakuten to the different input keys.
The character input device according to claim 20. 22. The character input device according to claim 18.
, A consonant that represents a pure sound and for which there is a corresponding voiced sound
Represent the above-mentioned voiced sound from a continuous combination of sound and voiced point.
A consonant identification means for identifying a consonant is provided, and the Japanese conversion control means is characterized by the consonant identification means.
Romaji-kana conversion for a specified combination of consonants and vowels
Then, a character input device that is a means for converting into a kana character string of 1
Place 23. The key input means is for inputting the Roman characters.
It is a consonant that represents the above-mentioned Kiyone, and the corresponding dakuon is
Duplicate existing consonants and the dakuten into the input key of 1
23. The character input device according to claim 22, wherein the character input device is assigned as follows. 24. The first Romaji-kana conversion results in the first
Kana sentences from combinations of consonants, second consonants, and vowels
The key input means comprises a conversion means for converting into a character string, and the key input means is the second of the Roman characters.
A contract in which the consonant and the dakuten are assigned to the different input keys.
The character input device according to claim 22. 25. The same child is obtained by the Romaji-kana conversion.
A kana sentence that has a consonant from a combination of continuous sounds and vowels
The key input means includes a consonant conversion means for converting into a character string, and the key input means
3. The input key different from the dakuten is assigned to the input key.
The character input device described in 2. 26. The key input means is a child representing Kiyone.
A consonant that is a sound and has a corresponding voiced sound
The character input device according to claim 18, wherein the character input device is assigned to a force key. 27. Romaji as a Japanese input element,
Assigned while allowing duplicate assignment to key 1
Enter multiple input keys and the key input obtained based on the operation of the input keys.
Consists of a combination of Roman letters assigned to force keys
An alphabetic character string conversion means for converting into an alphabetic character string and a dictionary based on the alphabetic character string converted by the alphabetic character string conversion means.
A character string that is converted to a candidate character string that contains kanji and
And a character input for inputting a character string selected from the candidate character strings.
Force input device , wherein the Japanese input element is connected to one of the plurality of input keys.
And assigns a dakuten to the alphabetic character string conversion means , and the alphabetic character string conversion means determines the Chinese character finally determined from the result of the conversion performed in the past.
Memorize the above-mentioned English character string corresponding to the candidate character string mixed with Kana and Kana
Priority is given to the alphabetic string storage means and the alphabetic string stored in the alphabetic character storage means.
Equipped with priority conversion means to increase the degree of conversion and perform subsequent conversions
Character input device. 28. A Roman character is divided as a Japanese input element.
Create a Japanese character string based on the operation of the entered input key
Then, in the character input method for inputting the character string , two or more Roman characters are assigned to one key in duplicate.
The key input obtained based on the operation of the input key is
Allows combinations of Roman letters assigned to input keys
The process of converting into a kana-kana character string and the converted kana-character string into Japanese characters mixed with kanji and kana
And a step of converting the sequence, the step of converting into the kana character string is voiced divide those as the Japanese input element in the input keys
If it is, the Roman character assigned to the input key
The kana character that is a voiced sound is produced from the combination of
The final decision was made based on the process to be performed and the result of the conversion performed in the past.
The Kana character string corresponding to the candidate character string mixed with Kana
The process of storing and the stored kana character string are given high priority.
A character input method including a step of performing subsequent conversions . 29. A Roman character is divided as a Japanese input element.
Create a Japanese character string based on the operation of the entered input key
Computer program for inputting the character string
Is a recording medium in which 1 or more of the above-mentioned Roman characters are assigned to one key in duplicate.
The key input obtained based on the operation of the input key is
Allows combinations of Roman letters assigned to input keys
In the step of converting into a kana character string and the step of converting into a kana character string,
If you assign a dakuten as the Japanese input element,
Of the Romaji and the dakuten assigned to the input key
As the kana character that is a dakuon from the combination, the kana character
The process of converting to a column and the result of the conversion performed in the past determine the final
The Kana character string corresponding to the candidate character string mixed with Kana
The process of storing and the stored kana character string are given high priority.
And the subsequent conversion process, and converting the Kana character string into a Japanese character string containing Kanji and Kana.
A step of, a program for causing a computer to execute the competent
A recording medium that is readable by a computer.