JP2003132053A - Apparatus for character string conversion, method therefor, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve speedup in conversion processing from a Kana (Japanese) character string to a character string of a mixture of Chinese character and Kana, and to facilitate selection of a conversion candidate, while maintaining the conversion accuracy of single-clause conversion, in the characters inputs using a key in which a plurality of similar but different characters are assigned. SOLUTION: When the conversion of the input kana character string to that of the mixture of Chinese character, and Kana is instructed by depressing a conversion/successor candidate key 71b after inputting Kana characters, a CPU34 executes single-clause conversion display processing. In the single-clause conversion display processing, the input Kana character string is converted to the mixture of Chinese characters and Kana, by assuming that a clause is constituted of the input Kana character string. When this is converted, a Kana character string (conversion object Kana character string), that becomes an object to be converted into Chinese character of Katakana (Japanese) among the input Kana character strings is fixed, and a content word having the same reading with the conversion object Kana character string is extracted from a content word dictionary. Next, the propriety of connection of the extracted content word with a remaining character string other than the conversion object Kana character string is decided, and only the content word clearing the decision is determined as the conversion candidate that is browsable and selectable.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for converting a character string. The present invention relates to a technique of inputting and converting at least a part of the input kana character string by referring to a dictionary.

[0002]

2. Description of the Related Art In recent years, even for products whose main purpose is not to create data such as documents, there are many cases in which a means for inputting a character string is provided in order to input setting information regarding the operation of the product. Among such products, it is not desirable to externally connect a full keyboard which is a general character string input means because the product is portable, and the product itself is small. It is difficult to secure a space for arranging the input keys similar to the full keyboard. Therefore, as a character string input means in such a product, there has been proposed one in which a plurality of different characters having the same character type are assigned to one input key and a smaller number of input keys are provided in the product itself than in a full keyboard.

For example, in a mobile phone, a numeric keypad for dialing is used as a key for data input.
That is, by assigning a plurality of different characters of the same character type to each key, the numeric keypad is configured as an input key capable of inputting 26 letters of the alphabet or the Japanese syllabary, and by operating the input key, numbers, letters, hiragana, katakana, etc. It realized the input of kana characters and the conversion of kana characters into kanji.
Such a method of inputting and converting a character string has been adopted in other portable information devices such as a small information terminal and an electronic notebook.

In such an input key to which a plurality of different characters of the same character type are assigned, in order to input one desired kana character, the user continuously taps the same key or selects a kana character candidate. It was necessary to perform complicated work such as selecting a desired kana character from the inside, which is not possible with a full keyboard. Taking the case of inputting the kana character "to" by kana, for example, with the full keyboard, you only have to press the key to which one hiragana character "to" is assigned once. I had to type the key to which the five hiragana characters, "Tetsu, to," were assigned five times in a row. In addition, when inputting the kana character “to” in Roman letters, “T”,
You only have to press each key, which is assigned one roman letter "O", once, but for small information terminals, etc., "T,
After pressing each key assigned two Roman letters “M” and “O, Y” once, all kana characters that can be considered from this combination of Roman letters, that is, “to” in the case of “TO” , "MO" for "MO", "TY"
In the case of, you must select the kana character of "to" from the four kana characters of "Undefined character of Cha" and "MY" indeterminate character of "My". It was

As described above, since the operation of the device provided with the input keys to which a plurality of different characters of the same character type are assigned is extremely limited as compared with the full keyboard, most of the users use the consecutive phrase conversion method (multiple Input a long kana character string that spans all bunsetsu and convert the input kana character string into a kana-kana mixed character string all at once. Was used to convert the kana character string to kanji etc. In other words, if a long Kana character is entered all at once with the input keys to which different characters of the same character type are assigned, in the unlikely event that a Kana character disappears during input or the Kana character is confirmed as it is due to an erroneous operation, etc. In addition, you have to redo the complicated operations that are incomparable to the full keyboard from the beginning. In order to avoid the situation where such a complicated operation is required again and to avoid wasting hard-to-enter kana characters, a device equipped with input keys to which different characters of the same character type are assigned is used. A single phrase conversion method has been widely used in which kana characters are converted or fixed to a desired kanji character or the like in advance when they are input one phrase at a time, and then the subsequent kana characters are input one phrase at a time.

[0006]

However, in the conventional single-clause conversion method, when the input kana character string is converted, even conversion candidates that have almost no probability of being adopted are extracted, so that the conversion candidate Not only does the generation process take a long time, but also the user's selection of the conversion candidate is wasted.

For example, when a kana character string "masuno ni" is input and converted to obtain a character string "masuno ni", conventionally, "masuno ni" and "masu no ni" are used. A character string having no meaning in Japanese or a character string having a different independent word region such as "increase" was extracted as conversion candidates. As a result, the number of extracted conversion candidates increases, which is a cause of delay in the conversion candidate generation processing. In addition, the user scrolls a conversion candidate group including a large number of conversion candidates other than the desired candidate to select "Masuno ni" which is a desired candidate, and selects a candidate "Masino ni" from this conversion candidate group. It had to be discovered and required complicated work.

Further, when a kana character string "Ikoto" is input and converted in order to obtain a character string "Ikoto", conventionally, the Japanese "Ikoto" is rarely used. Even the character string was extracted as a conversion candidate, and there was the same difficulty as above. Furthermore, "say"
In the case of an input device that adopts a conversion format that converts "and" into different words, the unintended conversion candidate "" may be displayed after the candidate "to say" is confirmed. In such a case, the kanji character string “thing” extracted or displayed as a conversion candidate must be returned to the original kana character string “koto” again, and both the input device and the user must do so. Wasted processing and work.

Therefore, the present invention solves the above problems,
When inputting characters using keys to which multiple different characters of the same type are assigned, while maintaining the conversion accuracy of single clause conversion, speeding up the conversion process from kana character strings to kanji and kana mixed character strings and selecting conversion candidates. The following configuration is adopted for the purpose of realizing the simplification of.

[0010]

Means for Solving the Problem and Its Action / Effect The character string conversion device of the present invention has an input section equipped with two or more one keys to which a plurality of different characters of the same kind are assigned. A character string conversion device for converting at least a part of a kana character string input by key operation by referring to a dictionary, the conversion target being a kana character string capable of converting the input kana character string into an independent word. A specifying unit that specifies the Kana character string, and a delimiter position in the input Kana character string between the conversion target Kana character string specified by the specifying unit and the remaining Kana character string other than the conversion target Kana character string As a grammatically appropriate combination of an independent word corresponding to the conversion target kana character string and an adjunct word forming the remaining kana character string; Was done An independent word according to the suit look, the longest test independent words that the conversion target kana character string is the longest,
When a candidate is selected from among the conversion candidates displayed by the display unit and the display unit that displays the conversion candidate in preference to other independent words, the character string of the selected candidate and the determination unit are used. A gist includes a determining means for determining a residual kana character string related to a combination determined to be grammatically appropriate as a converted character string without displaying a conversion candidate for the residual kana character string. To do.

"One key to which a plurality of different characters of the same type are assigned" means that a plurality of different characters of the same type are assigned to one key. As the "characters of the same type", for example, hiragana characters, Roman characters, etc. can be considered. As "different characters of the same type", for example, "A" and "I" can be considered in the case of Hiragana characters, and "A" and "B" can be considered in the case of Roman characters. It is only necessary that multiple "different characters of the same type" are assigned,
The number of assigned characters is 2, 3, 4, 5, 6
Any number may be used. For example, when different hiragana characters such as "A", "I", "U", "E", and "O" are assigned to one key, five characters are assigned to one key. It will be.

The adjunct word means a word other than an independent word.
As the auxiliary words, for example, in addition to particles and auxiliary verbs, inflectional endings, auxiliary nouns, auxiliary nouns, formal nouns, etc. can be considered.

In the character string conversion device of the present invention, the character string specifying means can convert the kana character string input by operating one key to which a plurality of different characters of the same kind are assigned to an independent word into a kana character string. Is determined as the conversion target kana character string, and the determination means corresponds to the conversion target kana character string as the delimiter position in the input kana character string between the conversion target kana character string and the remaining kana character string. The independent word relating to the combination determined to be grammatically appropriate by the determining means by determining the grammatical suitability of the combination of the independent word and the auxiliary word forming the remaining kana character string, ,
The longest test independent word having the longest conversion target kana character string is displayed as a conversion candidate in preference to other independent words. When one candidate is selected from the displayed conversion candidates, the confirming unit determines the character string of the selected candidate and the remaining kana character string relating to the combination determined to be grammatically appropriate by the determining unit, The converted character string is confirmed without displaying the conversion candidates for the remaining kana character string.

According to such a character string conversion device of the present invention, in determining the input kana character string as the post-conversion character string, the longest test independent word is given priority in the conversion target kana character string portion. It is displayed as a conversion candidate, and the conversion candidate is not displayed for the remaining kana character string portion. Therefore, it is possible to efficiently narrow down the conversion candidates displayed for the input kana character string while maintaining the conversion accuracy at a high level. As a result, it is possible to realize a higher conversion speed in the character string conversion device.
In addition, the user can easily select a conversion candidate.

In the above invention, the longest test independent word means that the combination with the adjunct word forming the remaining kana character string is determined to be grammatically appropriate (hereinafter referred to as requirement 1) and the conversion target. It means an independent word that satisfies the two requirements that the kana character string is longest (hereinafter referred to as requirement 2). In the above invention, such a longest test independent word is given priority over other independent words to be a conversion candidate. Display as. In this case, what kind of independent word other than the longest test independent word is displayed can be appropriately determined.

For example, it is possible to adopt a structure in which candidates satisfying any one of the above two requirements (requirement 1 and requirement 2) are displayed as other independent words. With this configuration, the number of displayed conversion candidates can be reduced as compared with the conventional case.

Further, only the longest test independent word is displayed as a conversion candidate, and the independent word satisfying the above two requirements (that is,
Independent words other than the longest test independent word) may not be displayed as conversion candidates. This makes it possible to narrow down the displayed conversion candidates to a smaller number. Therefore, it is possible to further accelerate the conversion speed in the character string conversion device and facilitate the user's selection work of conversion candidates.

Change means for changing the delimiter position in the kana character string input after the conversion candidates are displayed by the display means, and the kana character string located in front of the delimiter position changed by the changing means for conversion. A re-identifying unit that identifies the Kana character string, an independent word corresponding to the conversion-target Kana character string identified by the re-identifying unit, and an adjunct that forms the remaining Kana-character string other than the conversion-target Kana character string. Redetermination means for determining the grammatical suitability of a combination, and a test independent word, which is an independent word related to the combination determined to be grammatically suitable by the redetermination means, as a conversion candidate with priority over other independent words. Re-display means for displaying, and when the one candidate is selected from the conversion candidates displayed by the re-display means, the confirming means, and the character string of the selected candidate and the re-determination The kana character string of the remaining of the combination is determined grammatically more suitable stages, without displaying the conversion candidate for the residuum of the kana character string, it is also desirable that the means for determining a converted character string.

In this way, even when the delimiter position is changed and the converted character string is decided after the conversion candidates are displayed, the converted independent kana character string has priority over the converted independent word. It is displayed as a candidate, and the conversion candidate is not displayed for the remaining kana character string portion. Therefore, even when the break position is changed,
The conversion candidates displayed for the entered Kana character string are
It is possible to efficiently narrow down the conversion accuracy while maintaining a high level, and it is possible to ensure a high conversion speed and easy selection of conversion candidates.

In this case, it is possible to appropriately determine what independent word is displayed as an independent word other than the test independent word. For example, the conversion target kana character string with another displayed independent word may be the same as the test independent word, and only the independent word corresponding to one conversion target kana character string may be displayed as a conversion candidate. This way
The displayed conversion candidates can be further narrowed down.

The changing means may be a means for sequentially changing the delimiter position by a predetermined operation, and the test independent words corresponding to different conversion target kana character strings can be displayed one after another along with the operation of changing the delimiter position. It is preferable in that

When the converted character string containing one candidate is confirmed by the selection from the conversion candidates displayed by the redisplay means, the character string containing the selected one candidate of the converted character string is selected. Equipped with registration means to register as a learning independent word,
The display means may be a means for displaying the learning independent word registered by the registration means as a conversion candidate in preference to the test independent word. In this way, when the same kana character string as the kana character string corresponding to the converted character string is input again and a conversion instruction is given, the learned independent word is displayed as the highest priority conversion candidate. Therefore, the user can determine the desired character string without instructing to change the delimiter position or selecting the next candidate after displaying the conversion candidates, which can greatly reduce the operation labor of the user. it can.

The independent word corresponding to the conversion target kana character string is
The longest test independence excluding the kana-independent word is provided with a determining means for determining whether or not the kana-independent word has a kana character portion written in kana characters at the end of the independence word. A word or a test independent word excluding the independent word with the kana may be displayed as a conversion candidate. This makes it possible to display only independent words that collectively convert all the conversion target kana character strings as conversion candidates. Therefore, when selecting a conversion candidate, the user can correctly recognize which kana character string is converted.

After the judgment by the judging means, a registering means for registering an independent word with a kana name in the holding list and a specifying means for specifying a kana character string excluding the kana character part of the inputted kana character string. And when the same kana character string as the kana character string specified by the specifying means is specified as the conversion target kana character string by the specifying means or the re-specifying means, the kana with independent words registered in the holding list is ,
It is also preferable to provide a priority display means for displaying the longest test independent word or other independent words including the test independent word as a conversion candidate in priority. By doing this, the kana-independent word whose display as a conversion candidate is suspended is displayed as a conversion candidate having a high priority when the kana character string excluding the kana character portion becomes the conversion target kana character string. Therefore, it is possible to reduce the effort required by the user to select the candidate. For example, it is particularly effective when there are a number of longest test independent words or test independent words corresponding to the conversion target kana character string.

It is also preferable that the input section is provided with a ten-key pad having two or more keys, because the keys can be efficiently arranged even in a narrow space.

It is also possible to grasp the invention as an electronic device provided with the above character string conversion device. As such an electronic device, a mobile phone having a character string converter incorporated therein, a small information terminal having a character string converter incorporated therein, an electronic notebook having a character string converter incorporated therein, and the like can be considered. Further, the above character string conversion device can be incorporated in various electric products having a function of inputting operation information and setting information.

The input unit is not limited to one provided in advance in the product itself, but may be one externally connected to the product such as a character string converter or an electronic device for use.

According to the character string conversion method of the present invention, a kana character string is input by a key operation of an input section equipped with two or more one key to which a plurality of different characters of the same kind are assigned, and the inputted kana character string is input. Is a character string conversion method for converting at least a part of the above by referring to a dictionary, wherein a kana character string that can be converted from the input kana character string into an independent word is specified as a conversion target kana character string, and the specified kana character string is specified. Between the conversion target kana character string and the remaining kana character string other than the conversion target kana character string as the delimiter position in the input kana character string,
The grammatical suitability of the combination of the independent word corresponding to the conversion target kana character string and the auxiliary word forming the remaining kana character string is determined, and the independence of the combination determined to be grammatically suitable by the determination is determined. The longest test independent word, which is a word having the longest conversion target kana character string, is displayed as a conversion candidate in preference to other independent words, and one candidate is selected from the displayed conversion candidates. At this time, after the conversion of the selected candidate character string and the residual kana character string relating to the combination determined to be grammatically appropriate by the determination means without displaying the conversion candidates for the residual kana character string. The gist is to confirm as a character string.

According to the character string conversion method of the present invention, when the input kana character string is decided as the converted character string, the longest test independent word is given priority as the conversion candidate for the conversion target kana character string portion. The conversion candidate is not displayed for the remaining kana character string portion. Therefore, it becomes possible to efficiently narrow down the conversion candidates displayed for the input kana character string while maintaining the conversion accuracy at a high level, which speeds up the conversion speed and allows the user to select conversion candidates. It is possible to realize both of the above.

In the recording medium of the present invention, a kana character string is input by a key operation of an input section equipped with two or more one key to which a plurality of different characters of the same kind are assigned, and at least the inputted kana character string is inputted. A recording medium recording a program for converting a part by referring to a dictionary, wherein a kana character string that can be converted from the input kana character string to an independent word is specified as a conversion target kana character string, , The specified conversion target kana character string and the remaining kana character strings other than the conversion target kana character string are set as the delimiter position in the input kana character string, and the self-supporting kana character string corresponding to the conversion target kana character string Determining the grammatical propriety of a combination of a word and an adjunct word constituting the remaining kana character string, and an independent word relating to the combination determined to be grammatically appropriate by the decision, The step of displaying the longest test independent word having the longest kana character string as a conversion candidate in preference to other independent words, and when one candidate is selected from the displayed conversion candidates, the selected candidate is selected. A step of determining a candidate character string and a residual kana character string related to the combination determined to be grammatically appropriate by the determination means as a converted character string without displaying a conversion candidate for the residual kana character string. The gist is that a program for causing a computer to execute is recorded in a computer-readable manner.

The program recorded on the recording medium is read and executed by a computer. The computer executes this program and realizes the above character string conversion method. Therefore, by using the recording medium of the present invention in a device having a computer capable of reading a program, the device inputs a kana character string by operating a key to which a plurality of different characters of the same kind are assigned. However, it is possible to efficiently narrow down the conversion candidates displayed for the input kana character string while maintaining a high level of conversion accuracy, speeding up the conversion speed and selecting conversion candidates by the user. It becomes possible to realize both of the simplification of the work.

It is also possible to realize the recording medium of the present invention by recording the respective steps of the above recording medium separately on two or more recording media and recording these two or more recording media in combination.

As the recording medium, various recording media such as flexible disk, CD-ROM, DVD-ROM, semiconductor memory (ROM, PROM, EEPROM, flash memory, etc.) can be used. Of course, it is also possible to store these programs in a server placed on a network such as the Internet and download them to the client computer for use.

[0034]

BEST MODE FOR CARRYING OUT THE INVENTION In order to further clarify the configuration and operation of the present invention described above, the embodiments of the present invention will be described based on Examples. FIG. 1 is an explanatory diagram showing an external appearance of a mobile phone 70 according to a first embodiment of the present invention, and FIG. 2 is an explanatory diagram showing a hardware configuration of the mobile phone 70.

As shown in FIG. 1, the mobile phone 70 includes an input section 72 in which a large number of push buttons are arranged and a display 74. The input section 72 is provided with twelve push buttons 76a to l, six push buttons 75a to f and one cross key 71. this house,
Push buttons 76a to l arranged in four rows vertically and three rows horizontally
Consists of a numeric keypad, which is assigned the numbers "0" to "9" necessary for dialing, and auxiliary keys, which are assigned the symbols "* (asterisk)" and "# (sharp)". .

Further, each push button 75a arranged between the push buttons 76a to 76c and the display 74.
Each of ~ f has the following roles. The push button 75a is a call button for starting the input of a call dial signal, and the push button 75c is the mobile phone 7.
A power button for turning the power of 0 on / off. The push button 75b is a clear button for clearing input data, and the push button 75d is a phonebook button for calling a phonebook. The push button 75e is a mail button 75e for selecting the mail function.
f is a web button 75f for selecting a web function.

Around the push button 75d, a cross key 71 is arranged which can be pushed down in the up, down, left and right directions. The cross key 71 is used for the up key 71a,
It is composed of a downward direction key 71b, a leftward direction key 71c and a rightward direction key 71d. Each of these keys 71a-d
Along with the operation of, various setting menus such as the listening volume are called up.

The mobile telephone 70 of the present embodiment is prepared to be capable of executing functions such as mail creation and transmission / reception, and information transmission / reception by accessing the Web, in addition to dial call. These functions are selected by operating the mail button 75e or the web button 75f.

In the mail or web function, it may be necessary to create data after the function is selected. Therefore, in the mobile phone 70 of the present embodiment, the push buttons 76a to 76a to l are provided in order to give the push buttons 76a to l the role of the data input keys 76a to l for inputting character / symbol data. The kana characters of the Japanese syllabary are assigned to j along with the numbers from "0" to "9".

Specifically, the push button 76a includes
Ten characters, "a", "i", "u", "e", and "o", are assigned to the kana characters in the line and the kana sound of each kana character. Which of these ten kana characters is to be input is determined by the number of times the push button 76a is pressed. For example, if the push button 76a is pressed once, "A" is input, and if it is continuously pressed twice, "I" is input.

Like the push button 76a, the push buttons 76b to i are assigned with kana characters from the kana line to the kana line and squeaks and audible sounds that each kana character can take. Kana characters such as ",""" and "n" are assigned. Further, "" (voiced point) and "° (semi-voiced point)" are assigned to the push button 76k, and various symbols are assigned to the push button 76l. In this way, each data input key 7
Two or more types of characters, codes and symbols are assigned to 6a to l.

The push buttons 76a to 76l function as data input keys 76a to 76l for inputting kana characters and symbols when creating data in the mail function or the web function. By operating the data input keys 76a to 76l, data of kana characters, symbols, and numbers assigned to each button can be input. In addition, the input undetermined kana character string (hereinafter referred to as input kana character string) is converted to a character string in which kanji and katakana are mixed, and kanji and katakana and kana are mixed (hereinafter, kanji and kana mixed character string). It is also possible (hereinafter referred to as Kana-Kanji conversion). The kana character string input in this manner and the converted kanji / kana mixed character string are displayed on the display 74.

In this embodiment, the input kana character string is converted into a kanji / kana mixed character string, but it is of course possible to convert the input roman character into a kanji / kana mixed character string. In this case, push buttons 76a-l
26 letters of the alphabet may be assigned to. For example, by assigning approximately three letters of alphabet to each of the twelve push buttons 76a to 76l, it is possible to realize kana input using roman letters.

Upward key 71a and downward key 71
The b, left direction key 71c, right direction key 71d and push button 75d play the following roles when creating data in the mail function and the web function. The down key 71b functions as a conversion / next candidate key 71b for instructing execution of kana-kanji conversion for an input kana character string and display of a conversion next candidate. Further, in the state where the conversion next candidate is displayed, the up key 71a and the down key 71b function as a next candidate selection key for selecting one desired conversion next candidate. Push button 75d
Functions as a confirmation key for confirming the input kana character string or the character string after kana-kanji conversion. The push button 75d that realizes such a function will be referred to as an enter key 75d below.
Say. The left arrow key 71c is used for the display 74.
The left move key 71c for moving the position of the cursor CU in the character string displayed in
The right direction key 71d functions as a right movement key 71d that moves the position of the cursor CU (see FIG. 8) or the range of the reverse display band HT (see FIG. 8) to the right.

The block diagram of FIG. 2 shows the hardware configuration required to create or display such data. As shown in FIG. 2, the mobile phone 70 includes a bus 35 around a CPU 34 for executing various arithmetic processes.
The following parts are connected to each other by. ROM
Reference numeral 36 is a memory in which programs and reference data necessary for the CPU 21 to execute various arithmetic processes are stored in advance. The ROM 36 stores a program relating to a kana-kanji conversion process executed by a mail or web function described later.

The flash memory 37 is a ROM that can be erased / written in block units. The flash memory 37 stores a Japanese independent word dictionary 38, an adjunct word dictionary 39, a connection verification table 40, a phrase end determination table 41, etc., which will be described later. This independent word dictionary 38, adjunct word dictionary 3
9. The connection verification table 40 and the phrase end determination table 41 store various data referred to when executing the kana-kanji conversion processing.

The ROM 36 and the flash memory 37
The program or reference data stored in the storage device is stored in an external storage device or a recording medium such as a CD-ROM, and various programs and reference data are loaded from the storage device or recording medium to be executed by the mobile phone 70. It is also possible.

The RAM 44 is a memory for temporarily storing various data necessary for the CPU 21 to execute various arithmetic processes. The input interface 33 controls the input of the key signal from the input unit 72. A display controller (DC) 46 controls signal output to a display 74 as a display device.

A reception circuit 50 and a transmission circuit 51 are also connected to the bus 35. The transmission circuit 51 sends the information such as the input mail address and the created mail to the server SV of the relay station via the antenna 79.
The server SV transmits the information of the received mail to the device (the other mobile phone 95, the personal computer 90, etc.) having the received mail address directly or via the network NW. The receiving circuit 50 receives information such as a telephone number and a mail from the relay station via the antenna 79. Information such as mail can be transmitted and received by the receiving circuit 50 and the transmitting circuit 51.

The structures of the independent word dictionary 38, the auxiliary word dictionary 39, the connection verification table 40, and the phrase end determination table 41 stored in the flash memory 37 will be described with reference to FIGS. 3 to 6. The independent word dictionary 38 is a dictionary in which independent words are recorded. As shown in FIG. 3, the independent word dictionary 38 includes an index part JD in which readings of independent words are recorded and a dictionary body JM. The dictionary body JM contains notation corresponding to each reading of independent words and part-of-speech information when this notation is taken,
The relationship between reading and notation has a hierarchical structure in which the reading is the apex and is divided into multiple branches. For example, in FIG. 3, for one reading "Ma", "increase", "benefit", "between",
A plurality of notations “Ma” and pre-part-of-speech information and post-part-of-speech information of each notation are registered.

The pre-part-of-speech information is the part-of-speech information used when determining whether or not a word can be connected to the word immediately preceding the word, and the post-part-of-speech information is the part of the word immediately after the word. It is part-of-speech information used when determining whether or not a word can be connected. A specific example of this will be described later.

In the dictionary body JM, as a notation of an independent word,
Kanji and Katakana notations are registered. In addition,
The katakana notation is not registered for all readings, but only the readings normally represented by katakana are registered in the dictionary body JM. For example,
In the dictionary body JM, the kanji notation "tobacco" and the katakana notation "tobacco" are registered for the reading "tobacco", but "masuno ni", "masuno", "ma"
"Masunoni", "masuno", "ma" for the reading
The katakana notation is not registered because it is not commonly used in Japanese.

Here, in the description of the independent word registered in the dictionary body JM, the range composed of kanji and katakana is called the kanji part, and the length of the kana character string corresponding to the kanji part is the kana character corresponding to the kanji part. It is called the column length Q. In the dictionary body JM, the value of the kana character string length Q corresponding to the kanji part is registered for each independent word. For example, in the example shown in FIG. 3, since "Masuno" is composed entirely of kanji, the kanji part is "masuno" and the corresponding kana is "masuno". The value is "3". Since "sabo" is composed entirely of katakana, the kanji part is "sabo", the corresponding kana is "sabo", and the value of the kana character string length Q is "2". Since the "translation" is composed entirely of kanji, the kanji part is "translation" and the corresponding kana is "translation".
The value of the kana character string length Q corresponding to the kanji part is “2”. on the other hand,
In the “separation”, “minute” is kanji and “ke” is hiragana, so the kanji part is “minute”, the corresponding kana is “wa”, and the value of the kana character string length Q is “kanji”. 1 ”. The kanji part means the maximum range of kanji and katakana that starts from the beginning. For example, "win"
In this case, “Kachiki” which is the whole from “Katsu” to “Ki” becomes the Kanji part, and the value of the kana character string length Q corresponding to the Kanji part becomes “3”.

The structure of the auxiliary word dictionary 39 is shown in FIG. The adjunct word dictionary 39 is a dictionary in which adjunct words are recorded.
In addition to particles and auxiliary verbs, inflectional endings, formal nouns, auxiliary verbs, etc. are registered as auxiliary words in the auxiliary word dictionary 39. As shown in FIG. 4, the adjunct word dictionary 39 includes reading and notation of an adjunct word and a dictionary body FM in which the part of speech of this notation is recorded. That is, only the same Hiragana notation as reading is recorded in the dictionary body FM, and the Kanji notation corresponding to reading is not recorded. Therefore, the reading and notation are
There is a one-to-one correspondence between reading and notation. For example, in FIG. 4, for one reading "ni",
One hiragana notation "ni", and pre-part-of-speech information and post-part-of-speech information (case particles) of this notation are registered. Regarding the formal noun "koto", only one hiragana notation "koto" is registered for one reading, and the kanji notation "koto" is not registered.

Here, the pre-part-of-speech information and the post-part-of-speech information will be described with a concrete example. In the auxiliary word dictionary 39 illustrated in FIG. 4, the word "su" in which the front part-of-speech information is a five-step inflection ending and the rear part-of-speech information is a five-step union form is registered. When deciding whether or not to connect to the word immediately preceding the word "su", the part-of-speech information of "su" is judged to be a Saline five-step inflection ending, and the word immediately following the word "su" is determined. When determining whether or not to connect, the part-of-speech information of "su" is determined as a five-stage union form.

For example, "erasing" including the above "su"
In the case of the word "su", the word "su" means that the pre-part-of-speech information and the post-part-of-speech information are both five-stage verbs, and that the front-part-of-speech information and the post-part-of-speech information are both formal nouns. It is connected with the word ". In addition,
The word “erasing” is stored in the independent word dictionary 38,
The words “su” and “koto” are stored in the auxiliary word dictionary 39. When determining the connection relationship between the preceding word "erased" (hereinafter referred to as the preceding word) and the subsequent word "su" (hereinafter referred to as the subsequent word), the part-of-speech information of the subsequent word "su" is used. Is determined as the "sayuki five-step inflection ending". Further, when determining the connection relationship between the preceding word "su" and the succeeding word "koto", the part-of-speech information of the preceding word "su" is determined as "five-dan form".

The criteria for determining the connection relation between such words are defined in the connection verification table 40 described above. An example of the connection verification table 40 is shown in FIG. In the case of the word "erasing" described above, the post-part-of-speech information of the preceding word "erasure" is the Sago five-step verb, and the pre-part-of-speech information of the subsequent word "su" is the Sago five-step Because of the endings, it is determined that the word “erasing” and the word “su” can be connected (see FIG. 5). Also, because the post-part-of-speech information of the preceding word "su" is a five-dial form, and the pre-part-of-speech information of the subsequent word "su" is a formal noun, the words "su" and "koto" Are determined to be connectable (see FIG. 5).

Next, the phrase end determination table 41 will be described. The phrase end determination table 41 is a table in which a criterion for determining whether a certain word can be the end of a phrase is set. An example of the phrase end determination table 41 is shown in FIG. As shown in FIG.
The names of the posterior part-of-speech information of all types are written in the column of "post-part-of-speech information of the last word" in the phrase-end determination table 41, and It is written whether the last word having each post-part-of-speech information can be the end of a phrase. For example, in the case of the phrase "erasing", the last word "koto" is
Since the post-part-of-speech information is a formal noun, it is determined that it can be the end of a phrase. On the other hand, in the case of the phrase "erasing", the last word "erasing" is determined to be not the end of the phrase because the post-part-of-speech information is the Sadan five-stage verb. Also,
In the case of the phrase "erasing", the last word "su" has two parts of the post-part-of-speech information, the five-stage ending form and the five-stage union form, but either one of them (the five-stage ending form). (Shape) can be the end of a phrase, so it is determined that it can be the end of a phrase.

Next, the contents of the characteristic kana-kanji conversion processing in the mobile phone 70 will be described with reference to FIGS. 7 to 18. FIG. 7 is a flowchart showing the contents of the kana-kanji conversion process performed during the execution of the mail function and the web function as a kana-kanji conversion process routine. Figure 7
The kana-kanji conversion routine shown in is executed by the CPU 34 when the mail function or the web function is executed by pressing the mail button 75e or the web button 75f and the mail or the data is specified by pressing the enter key 75d. The content of this processing is stored in the ROM 36. In the following description, the content of the process when the mail function is executed will be described as a representative example, but the process that is substantially the same as when the mail function is executed is also executed when the web function is executed.

When this routine is started, first, after the process of setting the cursor CU to the reference position X0 is performed (step S100), it is determined whether or not a kana character has been input by the data input keys 76a to 76l. A determination process is performed (step S105). If it is determined that the kana character has not been input, the standby state is entered until the kana character is input.

When it is determined in the process of step S105 that a kana character has been input, a process of moving the position of the cursor CU in the kana character input direction by the number of input kana characters is performed (step In step S110, it is determined whether or not the conversion / next candidate key 71b has been pressed based on the presence / absence of a key signal input from the conversion / next candidate key 71b (step S120).

If it is determined in step S120 that the conversion / next candidate key 71b has not been pressed,
A process of determining whether or not the confirm key 75d is pressed is performed based on the presence / absence of a key signal input from the confirm key 75d (step S150). When it is determined that the confirmation key 75d is pressed, the input kana character string is confirmed as the kana character as it is, the position of the cursor CU is moved one by one in the input direction of the kana character, and the position of the cursor CU after the movement is changed. Processing for resetting to the reference position X0 is performed (step S15).
5). Enter key 75d in the process of step S150
If it is determined that is not pressed, step S12
Returning to the processing of 0, it is in a standby state until the conversion / next candidate key 71b or the confirm key 75d is pressed.

When it is determined in step S120 that the conversion / next candidate key 71b has been pressed, a single phrase conversion display process described later is performed (step S125).
In this single phrase conversion display process, it is considered that one phrase is composed of all input kana character strings, the input kana character string is converted to a kanji kana mixed character string, and the converted kanji kana mixed character string is converted. A process of displaying the first candidate is performed. For example, if the conversion / next candidate key 71b is pressed for the input kana character string "masuno ni", "masuno ni"
Is converted into a kanji-kana mixed character string assuming that is a single clause.

After the converted kanji and kana character string is displayed by the single phrase conversion display process, a process for determining whether or not the left move key 71c is pressed is performed (step S1).
30), if it is determined that the left move key 71c has been pressed, the reverse display band H that covers the converted kanji / kana mixed character string.
After the process of shifting the range of T to the direction opposite to the input direction of kana characters by the number of keystrokes of the left movement key 71c, a modified single phrase conversion display process described later is performed (step S135). In this modified single-phrase conversion display processing, the reverse display band HT is moved in response to the keystroke of the left movement key 71c, and it is instructed to change the word delimiter position for one phrase consisting of the entire input kana character string. It is determined that the input kana character string is converted to a kanji-kana mixed character string based on the changed delimiter position, and the converted kanji-kana mixed character string is displayed.

If it is determined in step S130 that the left move key 71c has not been pressed, conversion /
If it is determined that the conversion / next candidate key 71b is pressed, a process of displaying the next candidate of the kanji / kana mixed character string is performed. A certain next candidate display process is performed (step S145). Details of the next candidate display process will be described later with reference to FIGS. 14 to 15.

In this way, after the single phrase conversion display process, the modified single phrase conversion display process, and the next candidate display process are performed, the first candidate of the kanji / kana mixed character string is displayed or the next candidate is selected. A process of determining whether or not the confirm key 75d has been pressed is performed (step S150). When it is determined that the confirmation key 75d is pressed, the input character string is confirmed as the first or next candidate kanji / kana mixed character string, and the position of the cursor CU is moved by one in the kana character input direction. , The process of resetting the position of the moved cursor CU to the reference position X0 is performed (step S155).

After the processing of step S155 is completed, a processing for judging whether or not an instruction to end the kana-kanji conversion processing is given is performed (step S160), and until the instruction is given,
Returning to step S105, the above processing is repeated. This routine is ended when an instruction to end the kana-kanji conversion processing is given.

Next, the contents of the above-mentioned single phrase conversion display process, modified single phrase conversion display process, and next candidate display process will be described.

Before explaining the contents of each processing, the terms used in the following description will be defined in advance. FIG. 8 is an explanatory diagram showing a display example on the display 74 during execution of the kana-kanji conversion processing. The upper part of this FIG. 8 shows the display mode of the display 74 in the state where no input kana character string has been input, and the middle part of FIG.
The display mode of the display 74 when a kana character string "masuno ni" is input is shown. In the lower part of FIG. 8, the conversion / next candidate key 71b is pressed from the state of the middle part of FIG.
The display mode of the display 74 when the input kana character string "masuno ni" is converted into the kanji-kana mixed character string "masuno ni" is shown.

In FIG. 8, the triangle mark with diagonal hatching is the cursor CU, and the input kana character string "Masuni"
The underline immediately below is an unconfirmed display line MM that indicates that the underlined portion is an unconfirmed character string. The quadrangle surrounding the kanji-kana mixed character string "Masuno ni" is an inverted display band HT indicating that the portion surrounded by the square is selected.

As shown in FIG. 8, in the following description, the position of the cursor CU (hereinafter referred to as a reference position) when inputting an undetermined kana character is referred to as "X0", and the position after the reference position X0 is represented. Move the cursor CU position to "X1, X2, X
3, ..., Xi ”. Further, the length of the input kana character string from the reference position X0 (hereinafter abbreviated as kana character string length) is represented by a symbol "N". For example, the kana character string length N of the input kana character string shown in the middle part of FIG.
The value of is "4".

Of the input kana character strings, the kana character string to be converted into a character (kanji, katakana, etc.) different from the input kana character is called the conversion target kana character string, starting from the reference position X0. The length of the kana character string to be converted (hereinafter abbreviated as kana character string length) is represented by the symbol "K". When an independent word having the same reading as the conversion target kana character string is extracted, the input kana character string other than the kana character string that is the reading of the independent word is called the "residual kana character string" in this embodiment. A detailed description of the points will be given later.

For example, in the examples shown in the middle and lower rows of FIG. 8, the input kana character string "masuno ni" is converted to the kanji / kana mixed character string "masuno ni". In this case, the conversion target kana character string becomes "masuno" which is the kana character string of the part converted to the kanji "masuno", and the value of the conversion target kana character string length K is from "ma" to "no". The length is up to "3". In addition, the remaining kana character string becomes “ni”, and between the “no” and “ni” in the input kana character string “masuno ni”, the separation position S between the independent word and the adjunct word
P.

Hereinafter, the simple phrase conversion display process (step S125 in FIG. 7), the modified simple phrase conversion display process (step S135 in FIG. 7), and the next candidate display process (step S14 in FIG. 7).
The contents of each process of 5) will be described based on the flowcharts of FIGS. 9 to 11. For the sake of clarity, the display 74 during execution of each process is displayed.
The description will be made with reference to the drawings showing the display state of FIG.

FIG. 9 is a flowchart showing the contents of the single-clause conversion display processing as a single-clause conversion display processing routine, and FIG. 12 is an explanatory diagram showing a concrete example of the single-clause conversion display processing. Single phrase conversion display process The single phrase conversion display process routine shown in FIG.
This is a process executed by the CPU 34 when the kana-kanji conversion for the input kana character string is instructed by pressing 1b, and the contents of this process are stored in the ROM 36.

When this routine is started, first, the value of the kana character string length N, which is the length of the input kana character string from the reference position X0, is specified (step S200), and the value of this kana character string length N is specified. Is set as the value of the conversion target kana character string length K (step S205). For example, in FIG.
As shown in (A), when the input kana character string is a string "masuno ni", the value of the conversion target kana character string length K is "4" which is the value of the kana character string length N. Is set.

Subsequently, the index portion JD of the independent word dictionary 38 is referred to, and a process for determining whether or not the same reading as the kana character string to be converted is recorded (step S2).
10, S215). When it is determined that the same reading as the conversion target kana character string is recorded, it is considered that the dictionary body JM of the independent word dictionary 38 stores the independent word in which the conversion target kana character string is written in kanji or katakana. , The information of this independent word is extracted from the independent word dictionary 38.

Subsequently, with respect to the extracted independent word, a process of determining whether or not the entire reading is in the Kanji part is performed (step S220). When the entire reading is in the Kanji part, the phrase end determination table 41 is referred to for the post-part-of-speech information of this independent word, and a process of determining whether this independent word can be the end of a phrase is performed (step S225). . When the entire reading is in the Kanji part and can be the end of a phrase, the extracted independent word is registered in the candidate list TR, and the process proceeds to the next process (steps S230 and S240). The candidate list TR is an area for storing conversion candidates for independent words, and is provided in a predetermined area of the RAM 44.

On the other hand, when the extracted independent word does not read the entire kanji part, the information of the extracted independent word is registered in the holding list HR, and the process proceeds to the next process ( Steps S235, S240). In the holding list HR, in addition to reading and notation of the independent word, pre-part-of-speech information and post-part-of-speech information, the number of kana character strings that become the reading of the independent word and the value of the kana character string length Q corresponding to the kanji part are registered. The hold list HR is a list used in the processes of steps S250 to S275, which will be described later, and is provided in a predetermined area of the RAM 44.

If the extracted independent word is a kanji part of the reading but cannot be the end of a phrase, it is considered unfavorable to extract this independent word as a conversion candidate. The process moves to the next process (step S240) without being registered in the candidate list TR or the hold list HR.

Such steps S220 to S
A specific example of the processing up to 235 is shown in FIG.
FIG. 12B shows, as an example, a case where the conversion target kana character string is the character string “separation”. In the dictionary body JM of the independent word dictionary 38, two types of notation "translation" and "dividing" are recorded as the notation corresponding to the reading "translation" (see FIG. 3). Among these, the notation "translation" means that the entire reading of "translation" is the kanji part (step S220: YES) and can be the end of a phrase as a noun (see FIG. 6) (step S225: Yes),
It is registered in the candidate list TR. On the other hand, the notation "division" means that only "wa" in the reading "translation" is used as the kanji part (step S220: NO).
It is registered in the hold list HR. In this case, the hold list H
In R, information such as "translation (reading of independent words)", "dividing (notation)", "ka line lower verb (pre-part of speech information)", "ka line lower verb (post-part of speech information)", Information such as "2" (the number of kana character strings to be read) and "1 (value of kana character string length Q)" is registered.

The above steps S220 to S235 are performed for all the independent words in the independent word dictionary 38 which have the same pronunciation as the character string to be converted (step S240). As a result, when the independent word is registered in the candidate list TR, the independent word which is the first candidate is extracted from the candidate list TR, and the process of displaying the character string of the extracted independent word on the display 74 is performed. (Step S245: YES, Step S295), this routine ends. In the present embodiment, the independent word which is the first candidate is not always fixed by learning, but its detailed configuration will be described later.

On the other hand, in the processing of step S215, the independent word having the same reading as the character string to be converted is the independent word dictionary 38.
If it is determined that the independent word corresponding to the reading of the entire input kana character string is not included in the independent word dictionary 3
8), or step S24
When no independent word is registered in the candidate list TR in the process of 0, the conversion candidate for the conversion target kana character string cannot be displayed. Therefore, the condition of “conversion target kana character string length K ← N” is relaxed and a process of subtracting 1 from the value of the conversion target kana character string length K is performed (step S250). This is because the length of the kana character string to be converted is shortened to make it easier to find a conversion candidate. In the present embodiment, as a process for finding this conversion candidate, a pending list reference process (step S260) and an independent word dictionary reference process (step S270) are performed in this order. Details of the holding list reference processing and the independent word dictionary reference processing will be described later.

Steps S250 to S270
A specific example of the above process is shown in FIG. FIG. 12 (C)
Shows an example in which the kana character string to be converted is the character string "masuno ni". Independent word dictionary 38 is "masunoni"
Since the notation corresponding to all the readings is not registered (step S215: NO), “4” in FIG.
The value of the conversion target kana character string length K set as is set to "3" (step S250). After that, "masuno"
The conversion candidate corresponding to the reading
, And then the independent word dictionary 38 for extraction.

After completion of the holding list reference processing and the independent word dictionary reference processing, the independent words that are conversion candidates are candidate lists TR.
A process for determining whether or not it has been registered is performed (step S275). When the independent word is registered in the candidate list TR, the independent word that is the first candidate is extracted from the candidate list TR, and the extracted character string of the independent word is combined with the remaining kana character strings other than this independent word. The process of displaying on the display 74 is performed (step S275: YES, step S2).
95) and terminates this routine.

On the other hand, if no independent word is registered in the candidate list TR in the process of step S275, the conversion candidates for the conversion target kana character string cannot be displayed. Therefore, in order to further relax the condition of the conversion target kana character string length K, the process returns to step S250 to perform a process of further subtracting 1 from the value of the conversion target kana character string length K. Such subtraction processing is repeatedly performed until an independent word that is a conversion candidate is registered in the candidate list TR. However, when the value of the conversion target kana character string length K becomes 0, further subtraction processing cannot be performed,
A process of extracting an indefinite word corresponding to the input kana character string is performed, and a process of registering the extracted indefinite word in the candidate list TR is performed (step S255: YES, step S28).
0, S290). In the present embodiment, the whole input kana character string written in hiragana and the whole input kana character string written in katakana are extracted as indefinite words. After that, the indeterminate word which is the first candidate is extracted from the candidate list TR, and the processing for displaying the character string of the extracted indefinite word on the display 74 is performed (step S295), and this routine is ended.

Pending list reference processing (step S2 in FIG. 9)
The contents of 60) are shown in FIG.
It shows in 0. As described above, the value of the conversion target kana character string length K is subtracted before the start of this routine (step S250 in FIG. 9). When this routine is started, first, the hold list HR is referred to (step S300), and the value of the kana character string length Q corresponding to the kanji part is converted to the kana character string length K.
A process of determining whether or not an independent word that matches the value of is registered is performed (step S310).

In the processing of step S310, when it is determined that an independent word in which the value of the kana character string length Q corresponding to the kanji part matches the value of the conversion target kana character string length K is registered,
A process of extracting information on the registered independent word from the holding list HR is performed (step S320). When there are a plurality of such independent words, it is desirable to preferentially extract the independent word having the largest number of kana character strings to be read. This is because conversion candidates can be extracted for a long range of the input kana character string (the purpose of longest match).

Then, it is determined whether or not there is an input kana character string (hereinafter referred to as a residual kana character string) other than the kana character string which is the reading of the extracted independent word (step S330). If there is no remaining kana character string, it is regarded as a conversion candidate for the entire input kana character string and the candidate list T
A process of registering in R is performed (step S350).

On the other hand, if it is determined in the process of step S330 that there is a residual kana character string, the input word and the independent word in the input kana character string are inserted between the end of the conversion target kana character string and the beginning of the residual kana character string. A process of determining whether a reading of a residual character string is included in the auxiliary word dictionary 39 is performed with the word delimiter position SP (step S33).
5). When the reading of the residual kana character string is recorded, is it possible to connect the kana character string that is the reading of the extracted independent word and the residual kana character string grammatically, and A process of determining whether or not two conditions, that is, the end of a phrase, are satisfied is performed (step S34).
0). Only when these two conditions are satisfied, the candidate list T
A process of registering in R is performed (step S350). In the processing of step S350, the independent word is registered in the candidate list TR with a higher priority than the registration in the processing of step S450 of the independent word dictionary reference processing routine described later.

The determinations in steps S330 to S340 are made as follows. First, after identifying the reading of the residual character string, the auxiliary word dictionary 39
, The possible combinations of readings are determined, and the pre-part-of-speech information and the post-part-of-speech information for each reading (reading matches the notation) are extracted. By applying the relationship between the pre-part-of-speech information for each reading extracted from the auxiliary word dictionary 39 and the post-part-of-speech information of the independent word extracted in step S320 to the connection verification table 40 (FIG. 5), grammatical connection is established. It is possible to judge whether it is possible or not. Also,
Whether the residual character string can be the end of the phrase by applying the post-part-of-speech information of the last reading of each reading extracted from the attached word dictionary 39 to the phrase end determination table 41 (FIG. 6). Can be judged.

When it is determined in the processing of step S335 that the reading of the remaining character string is not recorded in the auxiliary word dictionary 39, when it is determined that the two conditions are not satisfied in the processing of step S340, or step S350.
When the registration process for the candidate list TR is completed, another independent word (an independent word in which the value of the kana character string length Q corresponding to the kanji part matches the value of the conversion target kana character string length K) in the holding list HR. ) Is present (step S370), and if there is such another independent word, the process returns to step S320 and the above processing is repeated. When it is determined that there is no other independent word in the process of step S370, or when it is determined that the required independent word is not registered in the holding list HR in the process of step S310, this routine is ended, and the independent word dictionary. Reference processing (step S in FIG. 9)
270).

Next, FIG. 11 shows the contents of the independent word dictionary reference processing (step S270 in FIG. 9) as an independent word dictionary reference processing routine. At the time of starting this routine, the value of the conversion target kana character string length K is maintained at the same value as that at the time of the pending list reference processing. When this routine is activated, first, the index portion JD of the independent word dictionary 38 is referred to, and a process of determining whether or not the same reading as the kana character string to be converted is recorded (steps S400 and S41).
0). When it is determined that the same reading as the conversion target kana character string is recorded, it is considered that the dictionary body JM of the independent word dictionary 38 stores the independent word in which the conversion target kana character string is written in kanji or katakana. , A process of extracting the information of the independent word from the independent word dictionary 38 (step S42).
0).

Subsequently, with respect to the extracted independent word, a process of determining whether the entire reading is in the Kanji part is performed (step S425). When the entire reading is in the Kanji part, the remaining kana character string is specified (step S430), and the space between the end of the conversion target kana character string and the beginning of the remaining kana character string is input. Separation position SP from the word
After that, a process for determining whether or not the reading of the remaining character string is included in the auxiliary word dictionary 39 is performed (step S
435).

When the reading of the residual kana character string is recorded in the auxiliary word dictionary 39, the kana character string which is the reading of the extracted independent word and the residual kana character string can be connected grammatically. And whether the remaining character strings may end up at the end of a phrase satisfy the two conditions (step S440), and only when the two conditions are satisfied, the process of registering in the candidate list TR is performed. Perform (step S450). The process relating to connection verification and phrase end determination in step S440 is performed in the same manner as the process in step S340 of the above-mentioned pending list reference routine. In the process of step S450, the independent words are registered in the candidate list TR in a lower priority than the independent words registered in the process of step S350 of the above-mentioned holding list reference routine.

On the other hand, when it is determined in the process of step S425 that the entire reading is not the kanji part, a process of registering the extracted independent word information in the holding list HR is performed (step S460). As described above, the holding list HR includes the reading and notation of the independent word, the pre-part-of-speech information, the post-part-of-speech information, the number of kana character strings that become the reading of the independent word, and the kana character string length Q corresponding to the kanji part. The value is registered.

When it is judged in the processing of step S435 that the reading of the remaining character string is not recorded in the auxiliary word dictionary 39, when it is judged that the two conditions are not satisfied in the processing of step S440, or step S45.
0, when the registration process to the candidate list TR and the hold list HR in S460 is completed, whether or not there is another independent word (independent word having the same reading as the conversion target kana character string) in the independent word dictionary 38. (Step S470), and if there is such another independent word, the process returns to step S420 to repeat the above process.

When it is determined in the processing of step S470 that there is no other independent word, or when it is determined in the processing of step S410 that the requested independent word is not recorded in the independent word dictionary 38, this routine is terminated. , And proceeds to the next process (step S275 in FIG. 9). The holding list reference process and the independent word dictionary reference process are repeatedly performed until the independent word is registered in the candidate list TR (step S275 in FIG. 9). After the independent word is registered in the candidate list TR, the first candidate among the registered independent words is extracted from the candidate list TR and displayed on the display 74.

FIG. 13 shows a specific example of the above-mentioned holding list reference processing and independent word dictionary reference processing. FIG. 13 shows a process in which an independent word is registered in the holding list HR or the candidate list TR when the single phrase conversion display processing is performed on the character string of the input kana character string “separate”. . First, the value of the conversion target kana character string length K is set to "3" which is the value of the kana character string length N, and the conversion target kana character string is set to the character string "separate" (step S2 in FIG. 9).
00, S205). In this case, since the free-standing word "yaru" is not included in the independent-word dictionary 38, FIG.
As shown in FIG. 3 (A), the holding list HR and the candidate list T
Nothing is registered in R.

Subsequently, the value of the conversion target kana character string length K is set to "2", and the conversion target kana character string is set to the character string "translation" (step S215: NO → step S250, step of FIG. 9). S255: NO). Hold list H
Since nothing is registered in R, after that, the process goes out of the holding list reference processing (step S310 in FIG. 10: NO →
NEXT) and the independent word dictionary reference process (FIG. 11) are performed.

In the independent word dictionary 38, there are two kinds of independent words, "translation" and "dividing", which are read as "independence" (see FIG. 3). this house,
As for the notation "translation", the whole reading of "translation" is taken as the kanji part, so the remaining kana character string is specified as "ru" (step S425: YES → step S43).
0), the delimiter position SP between the independent word and the adjunct word in the input kana character string "separating" is determined to be a position between "ke" and "ru". On the other hand, the preceding word "translation" and the subsequent word "ru", which is composed of the remaining character strings, are It is a connection with a subsequent word, and it is determined by the connection verification table 40 that connection is not possible (see FIG. 5). Therefore, as shown in FIG. 13B, the independent word “translation” is not registered in the candidate list TR (step S445: NO → step S470).

On the other hand, the notation "division" means that only "wa" in the reading "translation" is used as the kanji part, so that it is registered in the holding list HR as shown in FIG. 13 (B). (Step S425: NO → S460). In this case, in the holding list HR, "Translation (reading of independent words)", "division (notation)", "ka line lower verb (pre-part of speech information)", "ka line lower verb (post-part of speech information)" Information, "2 (the number of kana character strings to be read)", and "1 (value of kana character string length Q corresponding to kanji part)" are registered.

Since the independent word is not yet registered in the candidate list TR, the value of the conversion target kana character string length K is set to "1", and the conversion target kana character string is set to the character string "wa" ( Step S275 of FIG. 9: NO → step S25
0, step S255: NO).

After that, the pending list reference process is performed. Information on the independent word “dividing” is registered in the hold list HR. Since the value of the kana character string length Q corresponding to the kanji part of the independent word "division" matches the value of the kana character string length K to be converted (both are 1), this independent word "division" is extracted from the pending list HR. The remaining character string is identified as “ru”. After that, the delimiter position SP between the independent word and the adjunct word in the input kana character string "separating" is "ke" and "ru".
A position between and is determined and it is determined whether or not the reading of the remaining character string is included in the auxiliary word dictionary 39 (step S335). The connection relationship with the remaining character string and the end of the phrase are determined (step S31 in FIG. 10).
0: YES → step S320 to step S340).
The antecedent word "separation" and the subsequent word "ru", which is composed of the remaining character strings, are the antecedent word with the post-part-of-speech information of "ka or lower one-stage verb" and the preceding part-of-speech information of "ka "
Is a connection with a subsequent word of, and is determined to be connectable by the connection verification table 40 (FIG. 5). Further, in the phrase end determination table 41 (FIG. 6), since the latter part of the lower part of the post-part-of-speech information of the word “ru” is included, it is determined that the word “ru” can be the end of a phrase. . As a result, FIG.
As shown in (C), the independent word "division" is registered in the candidate list TR as the first candidate (step S340: YES in FIG. 10 → step S350).

After the processing of referring to the holding list is completed,
The independent word dictionary reference process (FIG. 11) is performed on the conversion target kana character string “wa”. In the independent word dictionary 38,
As an independent word with the reading "wa", "ring",
Independent words written in many notations such as "wa", "ring", "feather", "wa" are recorded. In these independent words, the entire reading of "wa" is the kanji part. Therefore, the remaining kana character string is specified as "ke", and the input kana character string "separate"
The separation position SP between the independent word and the attached word in is determined to be a position between "wa" and "ke". After this, the connection relationship with the remaining character string and the end of the phrase are determined, but the reading of "Ke", which is the remaining character string, is not included in the auxiliary word dictionary 39.
Is not recorded in (step S425 of FIG. 11: YE
S → step S435: NO). As a result, FIG.
As shown in (C), "ring", "sum", "ring",
Independent words such as "feather" and "wa" are not registered in the candidate list TR.

In this way, only the independent word "division" is registered in the candidate list TR, and the single-phrase conversion display processing ends. Therefore, when the kana character string "separating" is input and the conversion / next candidate key 71b is pressed, the character string "separating" in the input kana character string is displayed on the display 74 as shown in FIG. 13 (D). The character string “divided” registered in the candidate list TR is converted and displayed, and the character “ru” in the input kana character string is displayed as it is in an undetermined state. When the confirm key 75d is pressed from this state (see FIG. 7).
Step S150: YES), conversion candidates (eg, “ru”, “ryu”) for the undetermined character string “ru”
A character string mixed with kanji and kana such as “divide” is determined as it is, without being displayed, and is displayed on the display 74 in the state shown in FIG. 13 (E).

When the input kana character string is the character string "masuno ni", the single phrase conversion display processing is performed on this character string and the independent word is added to the holding list HR or the candidate list TR. Another example of the registration process is shown in Fig. 1.
Shown in FIG. As shown in FIGS. 14 (A) and 14 (B),
If the input kana character string is "Masuni", the delimiter position SP between the independent word and the adjunct word in the input kana character string "Masuni" is set to a position between "no" and "ni". Determined,
Four independent words “Masuno”, “Masuno”, “Masuno”, and “Masuno” are registered in the candidate list TR in this priority order, and the single phrase conversion display process ends.

Therefore, when the kana character string "Masuni" is input and the conversion / next candidate key 71b is pressed, the screen shown in FIG.
As shown in (C), the character string "masuno" in the input kana character string is converted and displayed on the display 74 as the character string "masuno" registered as the first candidate in the candidate list TR. , The character "ni" in the input kana character string is displayed as it is in an undetermined state. When the enter key 75d is pressed from this state (step S150 in FIG. 7: Y
ES), the kana-kana mixed character string “Masuno ni” is fixed as it is without outputting conversion candidates (for example, “ni”, “load”, etc.) for the undetermined character string “ni”, and FIG. It is displayed on the display 74 in the state shown in (D).

The display 7 shown in FIG.
When the conversion / next candidate key 71b is pressed from the state of 4, the sub window 74s as shown in FIG. 14 (E) is displayed. In the sub window 74s, all the words "masuno", "masuno", "masuno", and "masuno" registered in the candidate list TR are displayed together with a scroll bar.
The process of performing such display is the next candidate display process (step S145 in FIG. 7). The content of this next candidate display processing is shown in FIG. 15 as a next candidate display processing routine.

When this routine is started, first, the candidate list TR is referred to, and the independent word registered in the candidate list TR is read out (steps S500 and S51).
0). Next, the independent words registered in the candidate list TR are displayed in a list in the sub window 74s together with the numbers indicating the priority order as candidates (step S52).
0), this routine ends.

As described above, in the candidate list TR, only the independent word having the longest conversion target kana character string length K (independent word of K = 3 in the example of FIG. 14) is registered, and the words before and after it are registered. Only independent words that clear grammatical conditions such as connection with and ending of bunsetsu are registered. In the example of FIG. 14, the value of the kana character string length N to be converted is “2”, “box”, “masu”,
Independent words "mass" and "ma", "magic", "increase", and "benefit" in which the value of the conversion target kana character string length N is "1" are not registered in the candidate list TR. Therefore, when the conversion / next candidate key 71b is pressed from the state of the display 74 shown in FIG. 14 (C), "box", "cell", "mass", "between",
"Ma", "increase", "benefit" conversion candidates, and combinations of these independent words and residual character strings "Masu no ni", "Masu ni ni", "Masu ni" etc. The conversion candidates that have no meaning as Japanese are not displayed in the sub window 74s. In this way, subwindow 7
By accurately narrowing down the conversion candidates displayed in 4s, the number of conversion candidates displayed is reduced, and as a result,
The effort for selecting the next candidate can be reduced.

It should be noted that when the sub window 74s as shown in FIG. 14 (E) is displayed, the up arrow key 71 is pressed.
When a and the down key 71b are operated, the position of the reverse display band HT covering the conversion candidates is moved in the vertical direction. When the confirmation key 75d is pressed after the inverted display band HT is set to the desired conversion candidate (for example, “2, Masuno”) by such an operation (step S150: YES in FIG. 7), the unconfirmed “ni” is displayed. Conversion candidates for character strings (for example,
A character string including kanji and kana such as "Masuno ni" is confirmed and displayed on the display 74 as it is without outputting "boiled", "load", etc.).

From the state in which the conversion candidate character string "Masuno ni" shown in FIG. 14C is displayed, the delimiter position SP between the independent word and the adjunct word in the input kana character string is changed and conversion is performed again. May want to. For example, in the case where the character string “Masuno ni” is displayed even though the user was supposed to convert the kana-kana mixed character string “masu ni ni”, FIG. By changing the range of the reverse display band HT as shown by the two-dot chain line, the delimiter position SP between the independent word and the adjunct word in the input kana character string "Masuni" is between "su" and "no". Change to the position, change the value of the conversion target kana character string length N from the current "3" to "2" (the conversion target kana character string in this case is "masu"), and only the "masu" part It is necessary to convert the string to "increase" again. The process of changing the delimiter position SP between the independent word and the adjunct word in the input kana character string after the conversion into the kana-kana mixed character string and performing the conversion again based on the changed delimiter position SP is a corrected single phrase conversion display. This is the processing (step S135 in FIG. 7).

FIG. 16 shows the contents of this modified single-phrase conversion display processing as a modified single-clause conversion display processing routine. This routine is started when the key signal of the left movement key 71c is detected after the candidate character string after the conversion is displayed by the single clause conversion display processing. When this routine is activated, first, a process of specifying the number of consecutive keystrokes of the left movement key 71c (hereinafter referred to as the number of consecutive keystrokes) is performed (step S).
600).

Next, the range of the current reverse display band HT is shifted by the number of consecutive keystrokes in the direction opposite to the input direction of kana characters (step S610), and the currently set conversion target kana character string length K is set. A process of subtracting the value of the number of continuous keystrokes from the value of is performed (step S620). 14
When the left move key 71c is pressed only once as indicated by the chain double-dashed line in (F), the delimiter position SP between the independent word and the adjunct word in the input kana character string is "su" and "no". It is changed to the position between. As a result, the input kana character string “masu” located in front of the delimiter position becomes the conversion target kana character string, and the value of the conversion target kana character string length K is set to “2”.

Next, it is determined whether or not the value of the conversion target kana character string length K is 0 (step S63).
0). When the value of the conversion target kana character string length K is 0, a process of displaying the entire input kana character string as it is in kana characters (step S635) is performed, and the right move key 71
A process of waiting for the keystroke d is performed (step S64).
0). After that, when the right move key 71d is pressed, the range of the reverse display band HT and the delimiter position SP and the value of the conversion target kana character string length K are corrected based on the number of consecutive key presses of the right move key 71d. It returns to the process of S630.

When the value of the conversion target kana character string length K is not 0 in the process of step S630, the above-mentioned pending list reference process (FIG. 10), the independent word is based on the value of the conversion target kana character string length K. The dictionary reference processing (FIG. 11) is executed in this order (steps S650 and S660).

After the holding list reference processing and the independent word dictionary reference processing are completed, the independent words that are conversion candidates are candidate lists TR.
A process for determining whether or not the registration has been performed is performed (step S670). When the independent word is registered in the candidate list TR, the independent word that is the first candidate is extracted from the candidate list TR, and the extracted character string of the independent word is combined with the remaining kana character strings other than this independent word. The process of displaying on the display 74 is performed (step S680), and this routine is finished.

On the other hand, if no independent word is registered in the candidate list TR in the process of step S660, conversion candidates for kana and katakana for the conversion target kana character string cannot be displayed. Processing for displaying the target kana character string as it is in kana characters (step S
690), and this routine is completed.

FIG. 17 shows an example of reconversion by the modified single-clause conversion display process. FIG. 17A is the same as FIG.
It is explanatory drawing which shows the registration state to the candidate list TR and the display state on the display 74 when the left move key 71c is tapped once from the state shown in (C). 14
When the left move key 71c is pressed once in the state shown in (C), the delimiter position SP between the independent word and the adjunct word in the input kana character string is changed to a position between "su" and "no". ,
The value of the conversion target kana character string length K is changed to “2”. As a result, the kana character string to be converted is separated at the delimiter position S.
"Masu" which is the input kana character string located in front of P
Becomes

Then, the process for referring to the holding list is performed, and the independent word dictionary reference process is performed. This is because no independent word is registered in the holding list HR in the state shown in FIG. 14C (see FIG. 14B). Referring to the independent word dictionary 38, in the independent word dictionary 38, as the independent word of the same reading as “masu” which is the conversion target kana character string, “box”,
Three independent words whose readings are "kasu" and "mass" are all kanji parts are registered (see Fig. 3). Therefore, the remaining character strings "Noni" for these three independent words
Whether or not connection with and the end of phrase are determined (steps S400 to S440 in FIG. 11).

The post-part-of-speech information of the preceding words "Sho", "Masu", and "mass" are all nouns (see FIG. 3). On the other hand, the remaining part of the character string “Noni” has the prepositional part of speech information as the connecting particle “Noni” or the prepositional part of speech information as the semi-positional particle “NO” and the prepositional part information as the case particle “ It is possible to think of any of the combinations with "". When the following word is considered to be "Noni", it cannot be connected to the preceding words "Masu", "Masu", and "Mas", but when the following word is considered to be "No", Connections with the words "box", "cell", and "mass" are possible (see connection verification table 40 in FIG. 5). Therefore, it is determined that “box”, “cell”, “mass” and “no” can be connected. The preceding word "no" and the following word "ni"
As for the relation with, it is a connection between the front part-of-speech information: the semi-positional particle and the post-part-of-speech information: the case particle, and both are determined to be connectable (see the connection verification table 40 in FIG. 5). The part of speech information after the last word "ni" is a case particle, so "ni"
Is determined to be the end of a phrase (see the phrase end determination table 41 in FIG. 6).

As a result, as shown in FIG.
The independent words “box”, “masu”, and “mass” are registered in the candidate list TR in this priority order (step S67 in FIG. 14).
0). After that, the character string “Masu”, which is the independent word of the first candidate, is displayed in the reverse display band HT, and the remaining character strings “no” and “ni” are displayed on the unconfirmed display line MM. (Step S680 in FIG. 14).

FIG. 17B shows the registration state in the candidate list TR and the display 74 when the left move key 71c is tapped once from the display state shown in FIG. 17A.
FIG. 6 is an explanatory diagram showing a display state of “”. When the left move key 71c is pressed once in the state shown in FIG. 17 (A), the delimiter position SP between the independent word and the adjunct word in the input kana character string becomes a position between "ma" and "su". The value of the conversion target kana character string length K is changed to "1". As a result, the conversion target kana character string becomes the input kana character string “ma” located in front of the delimiter position SP.

In the state shown in FIG. 17A, the holding list H
Since no independent word is registered in R, the process for referring to the independent word dictionary is performed without executing the pending list reference process. In the independent word dictionary 38, "increase", "benefit", "ma", and
Four independent words whose entire reading is "Kama" are Kanji parts are registered (see Fig. 3). Therefore, it is judged whether or not the four independent words can be connected to the remaining character string "sunoni" and the ending of the phrase (steps S400 to S440 in FIG. 11). As a result, "increase",
When the Sayuki 5dan verb "Masu" is used as the antecedent, it can be connected to the subsequent word "Su (Sayuki 5dan inflection ending)", but the nouns "Ma" and "Ma" precede. When it is defined as a word, it is determined that the connection with the subsequent word "su (sa row five-stage inflection ending)" is impossible (see the connection verification table 40 in FIG. 5). As a result, as shown in FIG. 17B, the independent words “increase” and “benefit” among the above four independent words are registered in the candidate list TR in this priority order (step S670 in FIG. 14). . After this, as shown in FIG. 17B, the first
The character string "masu", which is the candidate's independent word, is displayed in reverse video.
"Su", which is the remaining character string displayed in T,
"NO" and "NI" are displayed on the unconfirmed display line MM (step S680 in FIG. 14).

FIG. 17C shows the registration state in the candidate list TR and the display 74 when the left movement key 71c is tapped once from the display state shown in FIG. 17B.
FIG. 6 is an explanatory diagram showing a display state of “”. When the left move key 71c is pressed once in the state shown in FIG. 17B, the value of the conversion target kana character string length K is changed to "0". Since this value "0" means that there is no kana character string to be converted, the pending list reference processing and the independent word dictionary reference processing are not performed. As a result, as shown in FIG. 17C, nothing is registered in the candidate list TR, and the input kana character string "masuno ni" is displayed as the kana character on the unconfirmed display line MM.

When the confirm key 75d is pressed from the display state shown in FIG. 17B (step S150 in FIG. 7:
YES), Kanji and Kana mixed characters such as "Increase" without outputting conversion candidates (eg, "Nest", "Field", "Package", etc.) for the undetermined character string "Suni". The row is confirmed. The display state on the display 74 after the confirmation is shown in FIG. In the present embodiment, the character string confirmed after changing the delimiter position SP by the keystroke of the left move key 71c is learned in this way, and when the same kana character string is input next time, the first priority is given to the other. Processing for displaying as candidates (hereinafter referred to as learning processing) is performed.

The content of such learning processing is shown in FIG. 18 as a learning processing routine. This routine executes C when the key signal of the enter key 75d is input after the corrected single phrase conversion display processing is performed (step S150 in FIG. 7: YES).
The processing is executed by the PU 34, and the content of this processing is stored in the ROM 36.

When this routine is started, first, the input kana character string before being determined is specified (step S700), and the specified input kana character string is first extracted in the single phrase conversion display processing (FIG. 9). Processing for requesting reading of an independent word is performed (step S710). This "reading of the independent word extracted for the first time" corresponds to steps S205 to S2 in FIG.
15 and the processing results of steps S250 to S270. For example, if the input kana character string is "Masuni", the independent words "Masuno", "Masuno", "Masuno", and "Masuno" are extracted for the first time.
The required reading is "masuno".

Next, a process of specifying the confirmed kanji-kana mixed character string (step S720) and cutting out the same reading part as the reading of the independent word extracted from the specified kanji-kana mixed character string for the first time as a registered character string. Is performed (step S730). If the delimiter of the adjunct word in "increase" does not match here, the part to be cut out is extended to match it. For example, if the kanji-kana mixed character string after confirmation is "Masuno ni", the part of "Masuno" that is the same reading as "Masuno" etc. is cut out as a registered character string. "Noni" is a single break, so the cut-out part of "Add" is "Add"
Expanded to the whole.

Next, referring to the independent word dictionary 38, the cut-out registered character string is registered with priority over other notations of the same reading (step S740). For example,
When the registered character string “to increase” is cut out, the registered character string “to increase” is indicated by the first part of the notation corresponding to the reading “Masuni” as shown in brackets in FIG. Registered as one candidate.

Next, the processing for setting the condition for the cut out registered character string is performed, the processing for registering the set condition together with the registered character string in the independent word dictionary 38 (step S750), and this routine is ended. . This condition setting is a condition that is intentionally set in order to be registered in the candidate list when the conversion target kana character string length K is the longest test character string length. For example, in the case of the registered character string "Add to ni", as shown in parentheses in FIG. 3, <kana character string length Q = 3>, <post-part-of-speech = connection particle “noni”> The condition is set or registered.

According to such learning processing, when the same kana character string is input again after the delimiter position has been changed and the kana-kana mixed character string has been confirmed, the same character string that was previously confirmed is first Can be displayed as a conversion candidate. For example, suppose that the confirmation key 75d is pressed from the display state shown in FIG. 17 (B) and the kanji / kana mixed character string “to increase” is confirmed and is displayed on the display 74 in the state shown in FIG. 17 (D). . After this, the kana character string "Masuni" is entered again and the conversion / next candidate key 71 is entered.
When b is pressed (step S120 in FIG. 7), the word "increase" is extracted from the independent word dictionary 38 in the pending list reference process (step S260 in FIG. 9) in the simple phrase conversion display process, and the independent word is extracted. Dictionary reference processing (Fig. 9
In step S270), the words “Masuno”, “Masuno”, “Masuno”, and “Masuno” are extracted from the independent word dictionary 38. Therefore, the word “to increase” is registered in the candidate list TR as the first candidate. As a result, FIG.
After the conversion / next candidate key 71b is pressed in step S120, the same character string as the one fixed by the previous division change can be displayed as the first candidate on the display 74 as the first candidate. .

According to the mobile phone 70 of the present embodiment described above, in inputting a short kana character string equivalent to one phrase and converting the input kana character string to determine the desired kanji / kana mixed character string, Only independent words having the longest conversion target kana character string are registered in the candidate list TR as conversion candidates for the input kana character string. Therefore, a number of conversion target kana character strings (for example, conversion target kana character string length K is 1,
It is not necessary to extract conversion candidates for kana character strings in each of the cases 2, 3, ..., And the conversion processing can be speeded up by reducing the number of extractions of conversion candidates. Also,
Only independent words having the longest conversion target kana character string are displayed on the display 74 as conversion candidates for the input kana character string. Therefore, a large number of conversion candidates having different lengths of kana character strings to be converted will not be displayed on the display 74. Therefore, it is not necessary to search for a desired candidate from a large number of conversion candidates having different lengths of kana character strings to be converted, and the user can easily select a conversion candidate.

Moreover, the conversion candidates registered in the candidate list TR and displayed on the display 74 are limited to those which have passed the connection test with the remaining character strings and the sentence end test.
Efficiently narrows down only the conversion candidates that match the usage as Japanese. Therefore, the conversion processing can be speeded up and the conversion candidate selection work can be facilitated while maintaining the conversion accuracy at a high level. In addition, since the conversion candidates are not extracted and displayed for the remaining character strings, it is possible to further accelerate the conversion processing speed and the conversion candidate selection work.

That is, since the adjunct word dictionary 39 does not contain the kanji or katakana notation for reading, when the independence word corresponding to the kana character string to be converted is extracted, the remaining words as the independence word and the adjunct word are not extracted. The combination with the character string can be grammatically analyzed with reference to the adjunct dictionary 39 to quickly determine the remaining character string. For example, when the input kana character string is "teach", first, the independent word "teach" is extracted, and then the word dictionary 39 for "koto" is extracted.
Refer to. After that, based on the information of the inflectional ending “ru” and the formal noun “koto” obtained from the attached word dictionary 39, the connection test for “teaching” and “ru” or the connection test for “ru” and “koto” If the end of sentence is determined for "koto", the conversion target kana character string part "teaching" and the remaining character string part "koto" are
Without extracting or displaying conversion candidates such as “things”, it is possible to immediately settle on a kanji-kana mixed character string “teaching”.

Further, the mobile telephone 70 of the present embodiment, when the conversion candidate character string is displayed, the left move key 71.
When the delimiter position SP between the independent word and the adjunct word in the input kana character string is changed by the operation of c, the kana character string located in front of the delimiter position is re-specified as the conversion target kana character string, and The grammatical propriety of the combination of the independent word corresponding to the specified conversion target kana character string and the adjunct word forming the residual kana character string is determined, and only the independent word related to the combination determined to be grammatically appropriate is determined. Display as a conversion candidate. When one candidate is selected from the displayed conversion candidates, the character string of the selected candidate and the remaining kana character string are displayed without conversion candidate for the remaining kana character string. Confirm as a mixed character string.

Therefore, even when the conversion position SP is changed and the conversion is performed again after the conversion candidates are displayed and the character string is to be decided, the conversion candidate for the input kana character string is converted to the remaining kana character string. Only independent words that have passed the connection test and sentence ending test are registered in the candidate list TR and displayed on the display 74, and conversion candidates are not extracted and displayed for the remaining character strings. Therefore, even when the delimiter position SP is changed, it is possible to efficiently narrow down the conversion candidates for the input kana character string while maintaining the conversion accuracy at a high level, thereby increasing the conversion speed and selecting the conversion candidates. It is possible to ensure the ease of.

Further, in this embodiment, the delimiter position SP is set.
When the mixed Kanji / Kana character string is confirmed after changing
A learning process of learning the determined character string is performed. Therefore, when the same kana character string is input again and a conversion instruction is given, the learned character string is displayed as the first candidate conversion candidate having a higher priority than other independent words. Therefore, the user can confirm the desired character string without instructing to change the delimiter position SP or selecting the next candidate after displaying the conversion candidates, and the operation labor of the user can be greatly reduced. You can

When the independent phonetic word having the same reading as the kana character string to be converted is extracted from the independent word dictionary 38, the mobile phone 70 of the present embodiment determines whether or not the entire reading of the extracted independent word is the kanji part. Judgment is made and the conversion candidates are displayed by excluding the independent words whose readings are not entirely kanji. As a result, only independent words that collectively convert all the kana character strings to be converted can be displayed as conversion candidates. Therefore, the user can correctly recognize which kana character string of the input kana character string is converted when selecting a conversion candidate.

In this embodiment, an independent word whose reading is not entirely in the kanji part is registered in the holding list HR. After that, when the character string to be converted is changed, first, the holding list H
By referring to R, an independent word having a kana-character-corresponding kana character string length Q having the same value as the conversion target kana character string length K is extracted and registered in the candidate list TR with the highest priority. That is, the independent word whose display as a conversion candidate is suspended is displayed as the conversion candidate having the highest priority when the kana character string forming the kanji part of the independent word becomes the conversion target kana character string. Therefore, it is possible to reduce the effort required by the user to select the candidate. Such a configuration is particularly effective when there are many independent words whose reading is the same as that of the kana character string to be converted and whose connection with the remaining kana character string is determined to be grammatically appropriate. .

Although the embodiments of the present invention have been described based on the embodiments, the present invention is not limited to the embodiments and can be implemented in various modes without departing from the gist of the present invention. Of course, you can do that.

For example, in the single phrase conversion display processing in the above-described embodiment, the kana character string to be converted is the longest, and the independent word that clears the connection test and sentence end test with the remaining kana character string (hereinafter, longest test independent word The number of conversion candidates is strictly narrowed down by using only) as a conversion candidate, but it is also possible to relax the conditions for narrowing down the number of conversion candidates and extract or display independent words other than the longest test independent word as conversion candidates. Is. As such other independent words, an independent word whose conversion target kana character string is different from the longest test independent word, or a conversion target kana character string which is the same as the longest test independent word, and does not clear the connection test or end-of-sentence test. It is possible to consider an independent word, an independent word that is written in hiragana or katakana, and is a character string that is the same as the conversion target kana character string as the longest test independent word. In such a case, it is preferable that the longest test independent word is registered in the candidate list TR with priority over the other independent words, because the conversion candidates can be easily selected.

Further, in the modified single clause conversion display processing in the above-described embodiment, conversion is performed by using only independent words (hereinafter referred to as verification independent words) that have cleared the connection verification with the remaining character strings and the end-of-sentence verification as conversion candidates. Although the number of candidates is strictly narrowed down, it is also possible to relax the condition for narrowing down the number of conversion candidates and extract or display other independent words other than the test independent words as conversion candidates. Such other independent words are Kana character strings that have the same conversion target kana character strings and different Kanji part lengths, or Kana character strings that are the same conversion character strings as test independent words and are written in hiragana or katakana. You can think of independent words. In such a case, it is preferable that the test independent word is registered in the candidate list TR with priority over the other independent words, because the conversion candidate can be easily selected.

Although the portable telephone 70 of the above-described embodiment is configured so that kana characters can be input by assigning kana characters to the 10 input keys, other characters, for example,
It does not matter if the alphabetic characters are assigned and the Roman characters can be input.

Further, in the above embodiment, the push button or the cross key is used as the character input key or the like, but the push button or the cross key is displayed on the display,
The displayed keys may be touched with a pen or the like to input characters.

In the above-mentioned embodiment, the portable telephone 70 has been described as an example, but the kana-kanji conversion method as in the above-mentioned embodiment is applied to the input section provided with two or more one key to which a plurality of different characters of the same kind are assigned. It may be applied to other devices having. Such other device may be any device that needs to input the created data, for example, business computer, desktop personal computer, notebook computer, word processor, electronic notebook, bank ATM, boarding ticket or betting ticket. Ticket issuing devices such as security devices, home appliances such as televisions and refrigerators, various vending machines, karaoke devices, registers, order input devices, fax machines, copy devices, game devices, personal information input devices, and authenticity of contracts, etc. It is possible to consider an apparatus or the like for creating and transmitting / receiving a document required by the.

In the embodiment described above, the program and reference data relating to character input and kana-kanji conversion are stored in the ROM.
36 and the flash memory 37. In this recording medium, such programs and reference data are stored in another recording medium in a computer-readable form, and the contents stored in this recording medium are stored. , CP
When loaded into a device including a computer including U, ROM, RAM, etc., it becomes possible to execute all the processes described in the present embodiment.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing an external appearance of a mobile phone 70 that is a first embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a hardware configuration of the mobile phone 70.

FIG. 3 is an explanatory diagram showing a structure of an independent word dictionary 38.

FIG. 4 is an explanatory diagram showing a structure of an auxiliary word dictionary 39.

5 is an explanatory diagram showing an example of a connection verification table 40. FIG.

FIG. 6 is an explanatory diagram showing an example of a phrase end determination table 41.

FIG. 7 is a flowchart showing a kana-kanji conversion processing routine.

FIG. 8 is an explanatory diagram showing a display example on a display 74 during execution of a kana-kanji conversion process.

FIG. 9 is a flowchart showing a single phrase conversion display processing routine.

FIG. 10 is a flowchart showing a pending list reference processing routine.

FIG. 11 is a flowchart showing an independent word dictionary reference processing routine.

FIG. 12 is an explanatory diagram showing a specific example of single phrase conversion display processing.

FIG. 13 is an explanatory diagram showing a specific example of a holding list reference process and an independent word dictionary reference process.

FIG. 14 is an explanatory diagram showing another specific example of the single phrase conversion display processing.

FIG. 15 is a flowchart showing a next candidate display processing routine.

FIG. 16 is a flowchart showing a modified single-phrase conversion display processing routine.

FIG. 17 is an explanatory diagram showing an example of re-conversion by a modified single clause conversion display process.

FIG. 18 is a flowchart showing a learning processing routine.

[Explanation of symbols]

33 ... Input interface 34 ... CPU 35 ... bus 36 ... ROM 37 ... Flash memory 38 ... Freestanding dictionary 39 ... Adjunct dictionary 40 ... Connection test chart 41 ... Phrase end judgment table 44 ... RAM 46 ... Display controller 50 ... Receiving circuit 51 ... Transmission circuit 70 ... Mobile phone 71 ... Cross key 71a ... Up key 71b ... Down arrow key, conversion / next candidate key 71c ... Left direction key, left movement key 71d ... right direction key, right movement key 72 ... Input section 74 ... Display 74s ... Sub window 75d ... Confirm key 75e ... mail button 75f ... Web button 76a-l ... Data input keys 79 ... Antenna 90 ... Personal computer 95 ... Mobile phone CU ... cursor FM ... dictionary HR ... holding list HT ... Inverted display band JD ... Index part JM ... dictionary K ... Kana character string length to be converted MM ... Indeterminate display line N ... Kana character string length NW ... Network Q: Kana character string length for kanji SP: Separation position SV ... server TR ... Candidate list X0 ... reference position

Claims (13)

[Claims] 1. An input unit having two or more keys each having a plurality of different characters of the same type assigned thereto, and referring to a dictionary for at least a part of a kana character string input by a key operation of the input unit. A character string conversion device for converting the input kana character string into a kana character string that can be converted into an independent word as a conversion target kana character string; and a conversion specified by the specifying means. An independent word corresponding to the conversion target kana character string and the remaining kana character are used as a delimiter position in the input kana character string between the target kana character string and the remaining kana character string other than the conversion target kana character string. Determination means for determining grammatical suitability of a combination with an accessory word forming a character string, and an independent word relating to the combination determined to be grammatically suitable by the determination means, wherein the conversion target kana character string is Longest And a display means for displaying the longest test independent word that becomes as a conversion candidate in preference to other independent words, and when one candidate is selected from the conversion candidates displayed by the display means, the selected candidate And a residual kana character string relating to the combination determined to be grammatically appropriate by the determination means as a converted character string without displaying conversion candidates for the residual kana character string. A character string conversion device having and. 2. The character string conversion device according to claim 1, wherein the display means is means for displaying only the longest test independent word as a conversion candidate. 3. The character string conversion device according to claim 1, wherein after the conversion candidate is displayed by the display unit, a changing unit that changes a delimiter position in the input kana character string, Re-identifying means for identifying a kana character string located before the delimiter position changed by the changing means as a conversion target kana character string, and an independent word corresponding to the conversion target kana character string specified by the re-identifying means And a rejudgment means for judging the grammatical adequacy of the combination of the remaining kana character strings other than the conversion target kana character string with adjuncts constituting the remaining kana character string, and And a redisplay means for displaying the test independent word, which is such an independent word, as a conversion candidate in preference to other independent words, wherein the confirming means is one of the conversion candidates displayed by the redisplay means. When complement is selected, the remaining kana character strings relating to the selected candidate character string and the combination determined to be grammatically appropriate by the re-determining means are displayed, and conversion candidates for the remaining kana character string are displayed. A character string conversion device that is a means for determining a converted character string without performing the conversion. 4. The character string conversion device according to claim 3, wherein the redisplay means is means for displaying only an independent word corresponding to one conversion target kana character string as the conversion candidate. 5. The character string conversion device according to claim 3, wherein the changing unit is a unit that sequentially changes the delimiter position by a predetermined operation. 6. The character string conversion device according to claim 3, wherein the converted character string including one candidate is determined by selection from the conversion candidates displayed by the redisplay unit. In this case, there is provided registration means for registering a character string including one candidate selected from the converted character strings as a learning independent word, and the display means is adapted to register the learning independent word registered by the registration means. A character string conversion device which is a means for displaying as a conversion candidate in preference to a test independent word. 7. The character string conversion device according to claim 1, wherein an independent word corresponding to the conversion target kana character string is represented by a kana character at the end of the independent word. A determining means for determining whether or not the kana with a kana having a character portion, the display means or the redisplay means, the longest test independent word excluding the kana independent word or the excluding kana independent word A character string conversion device that is a means for displaying a test independent word as a conversion candidate. 8. The character string conversion apparatus according to claim 7, wherein after the judgment by the judgment means, a registration means for registering the kana-independent word in a holding list, and a kana character string inputted A specifying means for specifying a kana character string excluding the kana character portion, and a kana character string which is the same as the kana character string specified by the specifying means is used as the conversion target kana character string by the specifying means or the re-specifying means. When specified, the priority display means for displaying the independent word with a kana registered in the holding list as a conversion candidate in preference to the longest test independent word or other independent words including the test independent word. Character string converter. 9. The character string conversion device according to claim 1, wherein the input unit includes a ten-key pad including the two or more keys. 10. An electronic device comprising the character string conversion device according to claim 1. 11. The electronic device is any one of a mobile phone in which the character string conversion device is incorporated, a small information terminal in which the character string conversion device is incorporated, and an electronic notebook in which the character string conversion device is incorporated. The electronic device according to claim 10, wherein 12. A kana character string is input by a key operation of an input unit provided with two or more one keys to which a plurality of different characters of the same kind are assigned, and at least a part of the input kana character string is stored in a dictionary. A character string conversion method for converting by reference, wherein a kana character string that can be converted into an independent word from the input kana character string is specified as a conversion target kana character string, and the specified conversion target kana character string and An independent word corresponding to the conversion target kana character string and the residual kana character string are configured with the remaining kana character string other than the conversion target kana character string as a delimiter position in the input kana character string. The grammatical propriety of the combination with the adjunct word is determined, and the longest test independent word that is the independent word related to the combination determined to be grammatically appropriate by the determination and in which the conversion target kana character string is the longest, Other independent words When a candidate is selected from the displayed conversion candidates and is displayed as a conversion candidate with priority over
The converted candidate character string and the residual kana character string relating to the combination determined to be grammatically appropriate by the determination means are displayed as a converted character string without displaying conversion candidates for the residual kana character string. String conversion method that is determined as. 13. A kana character string is input by a key operation of an input section provided with two or more one keys to which a plurality of different characters of the same kind are assigned, and at least a part of the input kana character string is stored in a dictionary. A recording medium recording a program for conversion by reference, which comprises: specifying a kana character string that can be converted from the input kana character string into an independent word as a conversion target kana character string; and The conversion target kana character string and the remaining kana character strings other than the conversion target kana character string are set as delimiter positions in the input kana character string, and the independent word corresponding to the conversion target kana character string and the residual A step of determining grammatical suitability of a combination with an accessory word forming a kana character string, and an independent word related to the combination determined to be grammatically suitable by the judgment, wherein the conversion target kana character string is the highest A step of displaying the longest longest test independent word, which is long, as a conversion candidate in preference to other independent words; and when one candidate is selected from the displayed conversion candidates,
The converted candidate character string and the residual kana character string relating to the combination determined to be grammatically appropriate by the determination means are displayed as a converted character string without displaying conversion candidates for the residual kana character string. A recording medium having a program readable by a computer, the program causing the computer to execute the step of determining as.