JPH0969104A - Device and method for kana-kanji conversion - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simplify conversion processing step in a KANA (Japanese syllabary)/KANJI (Chinese character) converter and to reduce the size of dictionary. SOLUTION: In the case of executing KANA/KANJI conversion processing by referring to a dictionary, word retrieval is executed by using a previously determined representative notation for words having various derived notations shown in the table. Information related to the representative notation is prepared as grammatical information and grammatical check whether the representative notation is formed as a clause or not is executed. When a character string candidate is displayed as a result of clause separated writing processing, the representative notation is rewritten to a display notation to be one of derived notations and displayed. Consequently, processing such as the grammatical check of a derived notation other than the representative notation in the clause sepatated writing processing can be removed and it is unnecessary to prepare the grammatical information of derived notations other than the representative notation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a kana-kanji conversion device and kana-kanji conversion method, and more specifically, a kana-kanji conversion device and kana-kanji conversion device for generating a kana-kanji mixed character string candidate by referring to a dictionary from an input character string. Regarding the conversion method.

[0002]

2. Description of the Related Art Hitherto, various kana-kanji conversion devices have been proposed for converting an inputted kana-character string into a desired kana-kanji mixed character string. In particular, recently, when performing Kana-Kanji conversion, if the user does not specify the delimiter of words or phrases in accordance with the Kana input, and if the Kana-Kana character string is entered, the desired Kana-Kanji mixed characters can be entered. Various efforts have been made to obtain rows.

As a process for correctly writing a kana solid character string into bunsetsus, two bunsetsus are analyzed as a basic unit, and two bunsetsus that can obtain the longest bunsetsu are set as first candidates. There is a longest phrase matching method, a minimum cost method in which a word that can be a candidate of a word constituting a phrase and a combination of words are given a cost, and the phrase having the smallest total score is the first candidate. In addition, there are many homonyms in Japanese, but in order to select the correct word from these homonyms, it has a dependency information dictionary that stores information on the dependency of certain words, There has been proposed a method of searching for a phrase including a word corresponding to dependency information. For example, the relationship between "book" and "book" of "book author" is registered as a dependency relationship, and priority is given to combinations such as "represent book" and "represent book". , "Authoring a book" is a candidate to avoid inappropriate conversion.

Further, as a device for making it easier to obtain a desired kana-kanji mixed character string, learning or use of a homonym word in which a word selected by the user immediately before the homonym word is given the highest priority as the next candidate. Memorize the phrase break that the person specified,
There is known a learning process such as phrase segmentation learning that preferentially uses this.

[0005]

However, there are many words in Japanese, for which a plurality of notations are allowed, as represented by Japanese characters and syllabary. For example, beginning with the difference in the notation depending on the way the feed is sent, such as "Handling", "Handling", "Handling", "Violin" and "Violin", "Buffer" and "Buffer" and "Buffer",
Examples are "10,000 yen" and "Ichiman". The fact that plural notations are allowed in this way is referred to as "notation fluctuation" or derivative notation. When there is such fluctuation of the notation, to select the desired notation, select the desired notation from a large number of conversion candidates when converting the character string, or It is necessary to set in advance the method of sending the feed.

Further, such fluctuations in the notation impose a heavy burden on the phrase segmentation writing process. For example,
In the case of the character string "Toriokai", "Handling"
Since "Handling" and "Handling" are allowed notations, when performing phrase segmentation processing, it is necessary to check the grammatical validity etc. for all of these, and for each of these allowed notations You must prepare the information in the dictionary.

The kana-kanji conversion device and the kana-kanji conversion method of the present invention solve these problems, reduce the size of the dictionary, reduce the steps of the segmentation and segmentation processing, and correctly segment the segmented words. It was made for the purpose and adopted the following composition.

[0008]

[Means for Solving the Problem and Its Action / Effect] The first kana-kanji conversion device of the present invention inputs a kana character string, refers to a dictionary, and writes the input kana character string into phrase segments. A kana-kanji conversion device that generates a kana-kanji mixed character string candidate, wherein a word dictionary that stores word readings and one representative notation as constituent elements, and a word that allows a plurality of notations A derived notation dictionary that stores the notation associated with the representative notation, and a phrase segmentation and writing means that performs segmentation and segmentation processing using the representative notation with respect to the input kana character string by referring to the word dictionary. A word notation selecting means for deciding a predetermined notation as a display notation when the other notations allowed for the words constituting the phrase obtained by the phrase segmentation processing exist, and the decision Word represented in the display representation that is, the gist that a display means for displaying the first candidate words constituting the kana and kanji character string candidate.

A kana-kanji conversion method corresponding to this kana-kanji conversion device inputs a kana character string, refers to a dictionary, and scribes the kana character string thus input, and generates kana-kanji mixed character string candidates. Kana-Kanji conversion method
When performing phrase segmentation processing on the input kana character string, refer to a word dictionary that stores word reading and one representative notation as constituent elements, and perform segment segmentation processing using the representative notation. When the words constituting the phrase obtained by the phrase segmentation processing are words for which a plurality of notations are allowed, a predetermined notation is determined as the display notation, and the word represented by the display notation is The gist is to display it as the first candidate of the words that form the character string candidate mixed with kana and kanji.

According to the first kana-kanji conversion device and kana-kanji conversion method of the present invention configured as described above, word reading and one representative notation are stored in the word dictionary as constituent elements, For a word for which a plurality of notations are allowed, this derivative notation is stored in the derivative notation dictionary in association with the representative notation. When a kana-kanji mixed character string candidate is generated from the input kana character string, the phrase segmentation processing is performed using the representative notation with reference to this word dictionary.
When the words constituting the bunsetsu obtained as a result have derivative notations, a predetermined notation is determined as the notation for display, and as the first candidate of the words constituting the generated kana-kanji mixed character string candidate, this Display the word represented in display notation. Therefore, in the dictionary used to determine whether or not a grammatical phrase is valid when performing phrase segmentation processing, it is sufficient to store only the grammatical information about the word represented by the representative notation, and only the word represented by the representative notation is stored. The phrase segmentation processing may be performed by using this.

Here, the word notation selecting means may be configured to include means for determining the display notation by referring to the derived notation dictionary. At this time, if the word obtained as a result of performing the phrase segmentation processing using the representative notation has a derivative notation, the predetermined notation is determined as the display notation by referring to the derivative notation dictionary. Therefore, the information about the display notation may be stored in the derivative notation dictionary, and only the information about the representative notation may be stored in the word dictionary.

Further, the word notation selecting means refers to means for storing a corresponding notation in the word dictionary from the derivative notation dictionary based on the use of past derivative notations, and information stored in the word dictionary. Then, it may be configured to include means for determining the display notation. With such a configuration, the corresponding notation is stored in the word dictionary from the derivative notation dictionary based on the past use of the notation. Therefore, when the word obtained as a result of the phrase segmentation processing using the representative notation has the derivative notation, the predetermined notation is determined as the display notation by referring to the information stored in this word dictionary. At this time, since the information about the display notation is stored in the word dictionary, it is possible to obtain the information about the display notation at the same time as the representative notation when referring to the word dictionary in performing the phrase segmentation processing. Therefore, when displaying a kana-kanji mixed character string candidate, it is possible to display the word represented by the display notation as the first candidate without referring to the dictionary again.

Further, the word dictionary and the derived notation dictionary may be integrated. At this time, when the phrase segmentation is performed using the representative notation with reference to the integrated dictionary, the notation for display is also read at the same time. Therefore, after the phrase segmentation processing, there is no need to refer to the dictionary again, and the kana-kanji character string candidates are displayed with the already-read word represented by the display notation as the first candidate.
Also, by integrating the word dictionary and the derivative notation dictionary,
The dictionary file is easy to handle.

Further, the display means for displaying the kana / kanji character string candidates is capable of distinguishing the homonyms of the word displayed as the first candidate from the notations allowed for at least the first candidate word. It is also possible to adopt a configuration including a split display means for displaying. With such a configuration,
When selecting a conversion candidate, it is not necessary to search for a desired notation from among various homophones and their notations mixed in an order that has no grammatical or semantic connection, and the desired notation can be easily selected. Become.

Further, the division display means expands the first candidate word and the derivative notation of each homonym in two different directions, and two-dimensionally displays the other word notation with respect to the first candidate word notation. It may be configured to. With such a configuration, it becomes easier to select a desired notation.

Here, when the conversion candidates below the second candidate are displayed by the division display means, when the word candidate displayed as the first candidate is not the desired word notation, a conversion instruction is given by a predetermined conversion key or the like. When input, the remaining candidates may be displayed two-dimensionally as described above, or the notation candidates of the word included in the phrase being converted may be two-dimensionally expanded as described above. good.

Further, in the divided display means, not all the word notation candidates that can be converted are not two-dimensionally expressed at once, but the word currently displayed as the first candidate is the word of the first candidate. It is also possible to develop and display the homologous synonyms expressed in the same style as the first candidate word notation and the derivative notation. At this time, if the desired word notation is not found in the displayed candidates, the derivative notation of the desired word notation among the displayed homonyms is once again selected as the first candidate, and a new one is newly selected. A desired word notation is selected from the displayed word notation candidates.

On the other hand, the divided display means displays the homonyms with the word to be displayed as the first candidate, and then receives a predetermined instruction to display the word and / or the first candidate.
Alternatively, it may be provided with means for displaying an allowable notation for at least one of the homonyms. In this case, when selecting the desired notation, first select the desired word from the homonyms and then select the desired notation from the multiple allowed notations for the selected word. Will be selected. Therefore, it is possible to easily reach the desired notation step by step.

It is also preferable that the derived notation dictionary also stores word notations using different characters in association with the same representative notation, and has the following configuration. That is,
When the first candidate word is displayed by the display means, if there is a variant character in the word representation of the first candidate, the variant character presence display means displays that there is a variant character, and at that time, the variant character If desired, the variant character selecting means converts into the notation using the desired variant character according to the notation style of the first candidate. In this case,
It is not necessary to search for a notation using a desired variant from a large number of candidate notations, and a word notation using a desired variant can be easily obtained according to the notation style that the user prefers.

Further, it is also preferable that the word notation selecting means is configured to learn the selection situation selected by the user from the derivative notations and to determine subsequent display notations based on the learning result. is there. This is because the word notation that the user has preferred and selected in the past becomes the first candidate, and the desired kana-kanji mixed sentence can be easily obtained.

Here, based on the result of the learning, the order of a plurality of allowed notations stored in the derivative notation dictionary is rearranged and stored, and the plurality of allowed notations is rearranged in this rearranged order. It is also possible to adopt a configuration provided with a derived notation dictionary learning means for preparing so that can be displayed as the conversion candidate rank. In this case, not only the word notation of the first candidate but also the word notation of the second candidate and below reflects the selection status selected by the user in the past, and it becomes easy to select the desired word notation.

Further, the display notation determined based on the result of the learning is stored in association with the representative notation,
It may be configured to include a display notation learning unit that prepares to be displayed as a first candidate of a word that constitutes a kana-kanji mixed character string candidate. With such a configuration, as long as the user continues to use the same word notation, it is not necessary to perform the operation for determining the display notation each time the first candidate word notation is displayed.

When one word notation is selected by the user from the plurality of allowed notations, the representative notation corresponding to the selected word notation is included in the homonyms in the word dictionary. The word dictionary learning means for changing the priority order to be used as the candidates of the words forming the mixed kana / kanji character string candidates may be provided.
In the phrase segmentation process, if the pronunciation and the part-of-speech are the same, which word is more suitable as the candidate word of the words that form the character string candidates mixed with Kana and Kanji unless other grammatical information is prepared. Can't judge any more. Therefore, by reflecting the word selection status of the user in the priority order of homonyms in the word dictionary, it becomes easy to obtain a desired kana-kanji mixed character string.

Further, the word notation selecting means is configured to prohibit learning of the selection status of the word notation when the user selects a notation using different characters from the plurality of allowed notations. Is also suitable. The use of variant characters is a special case, and it is not a good idea for the user to learn the special selection situation and make it easier to obtain a notation using variant characters that is a special notation as the first candidate. Because it will be done.

Further, in the process of generating the kana-kanji mixed character string candidates, learning means for learning that two or more words have been converted with a predetermined dependency relationship,
In at least one of the above words, when a predetermined derived expression with low frequency of use is selected and confirmed in a kana-kanji mixed character string, there is also a configuration including a learning prohibition means for prohibiting learning of the dependency relationship. preferable. Learning the dependency relationship is an effective means for displaying a desired word as a candidate word from a plurality of homonyms.
However, with regard to special notation, the situation in which it is used is likely to be special. Therefore, it is useful to further provide a means for not learning the dependency relation for the special notation that is not used later.

As described above, in the first kana-kanji conversion device and kana-kanji conversion method of the present invention, only the representative notation of each word is read and the segmentation and segmentation processing is performed. Therefore, the dictionary required for segmentation and segmentation processing is The dictionary size can be reduced by providing only the information about the representative notation. In addition, since the phrase segmentation writing process is performed only for the representative notation, there is an excellent effect that the operation in the segment segmentation segmentation is greatly reduced and the processing speed becomes higher.

On the other hand, the second kana-kanji conversion device of the present invention inputs a kana character string, refers to a dictionary, and writes out the input kana character string for each phrase to generate a kana-kanji mixed character string candidate. A kana-kanji conversion device that stores time and date information indicating a predetermined time such as time and date, and a date and time term conversion dictionary that stores a specific headword representing the time and date information in association with each other. A current date / time information generating means for generating date / time information, a date / time term searching means for checking the presence / absence of a specific entry word representing the date / time information when performing the phrase segmentation processing according to predetermined grammatical information, and the date / time term searching means When a specific entry word representing the date and time information is found, the date and time term conversion dictionary and the current date and time information generating means are referenced to correspond to the specific entry word representing the date and time information. Date / time term conversion means for converting into date / time information, date / time display format storage means for storing a plurality of date / time display formats, and a predetermined format determined from the plurality of date / time display formats when displaying the conversion result Date and time display means to display by
When the used format is undesired, a date / time display format selecting means for displaying the conversion result in a desired format selected from the rest of the plurality of date / time display formats, and the stored plurality of formats. It is provided with a date / time display format changing means for changing the date / time display format to be used when displaying one of the date / time information on the basis of the result of the learning that one has been selected by the user. Use as a summary.

A kana-kanji conversion method corresponding to this kana-kanji conversion device inputs a kana character string, refers to a dictionary, and scribes the input kana character string into phrases, and generates kana-kanji mixed character string candidates. Kana-Kanji conversion method
When the phrase segmentation processing is performed on the input kana character string, a specific entry word representing date and time information indicating a predetermined time point such as time and date is searched for, and a specific entry word representing the date and time information is searched. When is found, the date / time term conversion dictionary that stores the date / time information and the specific entry word representing the date / time information in association with each other, and the date / time information at the present time, are referenced to identify the date / time information. Is converted to the corresponding date and time information, the conversion result is displayed in a predetermined format determined from a plurality of stored date and time display formats, and when the displayed format is undesired, The conversion result is displayed in a desired format selected from the rest of the plurality of date and time display formats, and one of the plurality of stored formats is displayed by the user. Learns that it has been selected, based on the result of the learning, is summarized in that to change the date and time display format to be used when displaying the date and time information.

According to the second kana-kanji conversion device and kana-kanji conversion method of the present invention configured as described above, date and time information indicating a predetermined time point such as time and date, and this date and time information are represented. It is stored in the date / time term conversion dictionary in association with a specific headword. When the phrase segmentation processing is performed in accordance with predetermined grammatical information, the date / time term search means checks whether or not there is a specific headword representing this date / time information.
When a specific entry word representing the date and time information is found, the date and time term conversion means refers to the date and time term conversion dictionary and the current date and time information generation means for generating the date and time information at the present time to identify the specific date and time information. Convert entry words to corresponding date and time information. The conversion result is displayed using a predetermined format determined from the stored plurality of date and time display formats. When the format used at this time is undesired, the conversion result is displayed in the desired format selected from the rest of the plurality of stored date and time display formats. The selection status of this format is learned, and the date and time display format used when displaying date and time information is changed based on the result of this learning.

Therefore, by inputting a specific entry word representing date / time information, it is possible to obtain date / time information corresponding to the specific entry word representing this date / time information as a conversion result in a display according to a predetermined format. Further, since a plurality of date and time display formats are registered, the selection status of the date and time display format is learned, and the date and time display format used when displaying the first candidate is changed based on the learning result. When converting a specific entry word representing date and time information into date and time information, it becomes easy to display the conversion result in a date and time display format that the user prefers to select.

Whether the result obtained by converting a specific headword representing date and time information into corresponding date and time information is displayed, or whether the result converted as a normal word is displayed instead of the specific headword representing date and time information. It may be configured to include a date / time conversion switching means for selecting. In this case,
When it is not necessary to perform date / time conversion, unnecessary date / time-converted conversion candidates are not displayed.

As described above, according to the second kana-kanji conversion device and kana-kanji conversion method of the present invention, the date and time is converted by inputting a specific entry word representing the date and time information, and the date and time information corresponding thereto is predetermined. Since the displayed date and time display format is used and the used date and time display format is learned, a document including date and time information described using a desired format can be created very easily.

As a kana-kanji conversion device capable of converting a specific entry word representing the date and time into date and time information, in addition to the configuration of the first kana-kanji conversion device of the present invention, time, date, year, etc. A date / time term conversion dictionary that stores date / time information indicating a predetermined time point and a specific entry word representing the date / time information in association with each other, a current date / time information generating unit that generates the date / time information at the current time point, and predetermined grammatical information When performing the phrase segmentation processing according to, the date and time term searching means for checking the presence or absence of a particular headword representing the date and time information, and when the particular headword representing the date and time information is found by the date and time term searching means, A date and time term conversion unit that refers to the date and time term conversion dictionary and the current date and time information generation unit and converts a specific entry word representing the date and time information into the corresponding date and time information, and the date and time When displaying a conversion candidate of a specific headword representing information, as one of the homonyms of the specific headword representing the date / time information, or one of the allowed notations of the specific headword representing the date / time information. As one of the configurations, a configuration including a date / time candidate mixed display means for displaying the conversion result into the date / time information in a predetermined format is also preferable.

According to the above configuration, the result of converting the specific entry word representing the date and time information to the corresponding date and time information is as one of the homonyms of this specific entry word.
Alternatively, since it is displayed in a predetermined format as one of the allowed notations of this particular headword, the desired notation can be selected extremely easily. In addition, when converting a specific entry word representing date and time information, it is not necessary to input an instruction as to whether to convert to corresponding date and time information or to display the result of conversion as a normal word.

Further, the third kana-kanji conversion device of the present invention includes, for a word in which a plurality of notations allowed in Japanese exist, one of the allowed notations and one representative notation. Derived notation, which is an allowable notation, is arranged, and the representative notation is used in the processing of segmentation and segmentation in Kana-Kanji conversion, and one of the representative representation and the derivative notation is used in the segmentation after segmentation and segmentation. The point is to use one.

The third kana-kanji conversion device of the present invention as described above has the same effects as the first kana-kanji conversion device of the present invention.

[0037]

BEST MODE FOR CARRYING OUT THE INVENTION In order to further clarify the structure and operation of the present invention described above, the embodiments of the present invention will be described below based on Examples. FIG. 1 is a block diagram showing a hardware configuration in which a kana-kanji conversion control logic actually operates in a kana-kanji conversion device according to a preferred embodiment of the present invention. As shown in FIG. 1, this device is a CPU 21 that executes various arithmetic processes for controlling operations related to Kana-Kanji conversion according to a preset program.
And the following parts connected to each other by a bus 31. The ROM 22 stores in advance programs and data relating to kana-kanji conversion necessary for the CPU 21 to execute various arithmetic processes, and the RAM 23 also stores C
The PU 21 is a memory in which various data related to Kana-Kanji conversion necessary for executing various arithmetic processes are temporarily read and written. The keyboard interface 25 controls key inputs from the keyboard 24, the CRTC 27 controls signal output to the CRT 26 that can be displayed in color, and the printer interface 29 controls data output to the printer 28. The hard disk 32 has a RAM 2
The Kana-Kanji conversion processing program provided in the form of a device driver and various programs to be loaded and executed in the memory 3 or various conversion dictionaries are stored.
DC) 30 controls. The timer 33 generates date and time information indicating a predetermined time, such as a current time, a date, and the like.

With the hardware thus configured, input of a character string, conversion of kana to kanji, editing, display, printing, etc. are performed. That is, the character string input from the keyboard 24 is subjected to predetermined processing by the CPU 21 while referring to various data stored in the hard disk 32,
It is stored in a predetermined area of the RAM 23 and is displayed on the screen of the CRT 26 via the CRTC 27.

Next, the details of the kana-kanji conversion processing executed by the hardware thus configured will be described. First, the configuration and operation of each part involved in the kana-kanji conversion process will be outlined with reference to FIG. It is a thing. The CPU 21 performs all processing. Regarding kana-kanji conversion, when the keyboard 24 is operated, a predetermined interrupt process is activated, and a device driver for converting the input key image into a corresponding kana-character string and further converting this into a kana-kanji mixed character string. to start. Of course, if the computer is capable of parallel processing, then kana-kanji conversion may be performed by one application (input method), and the conversion result may be passed to the required application. In this case, the input method collectively accepts the input from the keyboard 24.

The key image from the keyboard 24 is accepted by the character input unit 40, and converted into a corresponding character string here. When inputting romaji, it refers to a predetermined conversion table and converts into a kana character string. Every time one kana character is obtained, the character input unit 40 sends the kana character to the conversion control unit 42. This conversion control unit 42
It plays a central role in Kana-Kanji conversion, controls various Kana-Kanji conversions described later, and sends the result to the converted character string output unit 44. Converted character string output unit 44
Actually sends a signal to the CRTC 27,
Display the converted string.

The conversion control unit 42 delivers the received kana characters to the character string input unit 50. The character string input unit 50 stores the kana character string in the character storage unit 52. Based on this character string, the independent word candidate creation unit 54 and the adjunct word candidate creation unit 64
And create word data candidates. The independent word candidate creation unit 54 uses the independent word dictionary 58 stored in advance in the hard disk 32, and under the management of the independent word analysis position management unit 56, performs processing for extracting independent word candidates from the obtained kana character string. To do. On the other hand, the adjunct word candidate creating unit 64 also performs the process of extracting an adjunct word candidate from the obtained kana character string under the control of the adjunct word analysis position managing unit 66 using the adjunct word dictionary 68. The process of extracting independent word candidates and adjunct word candidates while moving the analysis position will be described later.

Here, the independent word dictionary 58 is learned by learning.
Change the priority of homonyms and affixes. This learning process is performed by the independent word learning unit 72 and the auxiliary word learning unit 7
4, an affix learning unit 76 and a character conversion learning unit 78. The independent word learning unit 72 learns the word selected last in the independent word group in which homonyms exist as the highest priority candidate. The auxiliary word learning unit 74 learns in which word form an auxiliary word such as "please" is converted to "please" or "please". Furthermore,
The affix learning unit 76 is for learning conversion formats such as prefixes and suffixes (for example, "Go" and "Go"). The character conversion learning unit 78 learns the input character string as it is as hiragana or katakana and outputs the confirmed hiragana or katakana as a candidate in the subsequent conversion process. .

The independent word candidate creating unit 54 and the auxiliary word candidate creating unit 64 obtain the created word candidates, and the word data creating unit 80 creates data for each word candidate. That is, as a result of the connection verification unit 82 referring to the connection verification table 84, the connection between the obtained independent words and adjunct words, independent words and independent words, and further the connection between the clauses consisting of “independent words + adjunct words”,
And the cost calculation unit 86 obtains the result of the cost calculation performed by the cost calculation unit 86 and outputs the result as data for each word.
This word data is once stored in the word data storage unit 100 and is used for the processing of segmentation and segmentation. The segment segmentation writing unit 102 determines a first segment segmentation segmentation candidate from the obtained data.

By the above processing, the first candidate for segmentation and segmentation and the first candidate for kana-kanji conversion for each segment are determined. The phrase segmentation writing unit 102 stores the candidates in the phrase data storage unit 106, and the stored candidates are output to the conversion control unit 42 by the conversion character string output unit 108. The conversion control unit 42 displays this character string as a candidate character string, and since an undesired character string may be a candidate, the conversion control unit 42 receives instructions from the user and performs processing such as display and selection of the next candidate. To do. The results of these instructions and selections are
It is input to the phrase data storage unit 106 and each of the learning units 72 to 78 described above, and is used for partially fixing the phrase and rewriting the priority order by learning. Although not shown, when the character string is confirmed by the user, all the data temporarily stored in the respective units are erased to prepare for the next conversion.

The outline from the input of the kana character string to the output of the converted character string has been described above. Next, the details of each process will be described. First, the kana-kanji conversion processing will be roughly described, and then the word dictionary which is the main part of the present invention will be described. FIG. 3 is a flowchart showing a kana-kanji conversion processing routine in this embodiment. This processing routine is a processing which is started when a conversion key (for example, "space key") is pressed after one or a plurality of kana characters are input from the keyboard 24. Even if the conversion key is not operated, when a predetermined number of kana characters are input or when a delimiter such as “.”, “,” “.” Is input, the kana-kanji conversion of FIG. It does not matter if the process is started. When this process is started, first, a word search process (step S100) and a segmentation process (step S110) are performed. These processes will be described with reference to the flowchart of FIG. FIG. 4 is a flowchart showing an outline of the phrase segmentation / drawing processing by the minimum cost method, which is one of the general segmentation / segmentation processing, and corresponds to step S100 and step S110 in FIG. In addition, FIG. 5 is an explanatory diagram showing a manner of segmentation and segmentation by the minimum cost method. Hereinafter, the flow of the phrase segmentation processing in FIG.
A description will be given according to the specific example in FIG.

When a character string is input from the keyboard 24 and the phrase segmentation writing process is started, first, an initialization process such as erasing the temporarily stored data and initializing the analysis position in the first digit ( After performing step S200),
Processing for obtaining an analysis position is performed (step S210).
Now, if the kana character string "Take care changes" is input from the keyboard 24, the first analysis position is the "to" position of the first digit. As the analysis of the input kana character string proceeds, the analysis position is advanced one by one from the beginning of the input kana character string. That is, after the search processing described below is performed with the position of "to" as the analysis position, the next analysis position becomes "ri",
The next step is "a".

When the analysis position is determined in step S210,
At each analysis position, a process of searching the independent word dictionary 58 and the auxiliary word dictionary 68 stored in the hard disk 32 is performed (step S220). For example, if “to” in FIG. 5 is the analysis position, “door”, “climbing”, “tuple”, “bird”, “topping”
A word such as "handling" is obtained as a search result. The word obtained as a result of these searches is linked to the ending digit position of the word as shown in FIG. When the word "handling" is searched, no word having a reading after "Torii Tsukaiga" can be found in the dictionary, so the dictionary search with "to" as the analysis position ends. When the dictionary search is completed, a process of checking the possibility of connecting the obtained word with the previous word is performed (step S23).
0). In the above example where "to" is the analysis position, since "to" is the character at the beginning of the sentence, the possibility of connection with the previous words is not checked, but for example, just before the particle. When it is determined that the connection is impossible, such as when a particle exists, it is treated as invalid data (step S235).

For words that may be connected, the cost is calculated next, and the minimum total cost of the words is calculated (step S240). This is an independent word =
2. This is a process for obtaining the minimum total cost from the beginning of the input character string to each word, assuming that the cost is 2, adjunct = 0. For example, the total cost of "door" (independent word) is 2,
The total cost of "ri" is calculated as 4 if "door" (independent word) + "ri" (independent word). At this time, in order to find the minimum total cost, the total cost of "A" is
"Bird" + not "door" + "Ri" + "A" cost 6
Adopt cost 4 for "A". Since "ga" is an adjunct, the cost of the word "handling" = 2, which has the lowest total cost of the words linked immediately before it, becomes its own cost. In FIG. 5, the cost of each word is shown at the bottom right of each. It should be noted that, for further detailed analysis, the adjunct word may be divided into two, and the particle = 0 and the auxiliary verb = 1 may be set.

After the above cost calculation is completed, the cost of each word is checked (step S250). This is a process that excludes combinations in which a character string is divided into word combinations and costs are higher than other combinations. For example, the combination of "door" + "ri" is higher than the cost of "bird" and "take", so "door" + "ri"
The combination is determined to be inappropriate and excluded from the phrase candidates (step S255). In FIG. 5, “●” is added to the upper right of a word that is determined not to be adopted as a bunsetsu candidate as a result of such a cost check, and “○” is added to the upper right of a word that remains as a word that may form a bunsetsu candidate. did.

Next, a process of linking the remaining words as valid words is performed (step S260).
That is, with respect to a combination of words for which the connection is valid as a result of the connection check, the pointers are set to connect the words. In FIG. 5, “handling” is linked to “ga”, and “ga” is linked to “alternate”, “change”, and “alternate”. Connection check, cost calculation, cost check,
When processing such as word candidate link is completed for all the words searched at one analysis position, step S210
Return to and advance the analysis position by one digit, and search the dictionary again,
Repeat similar processing such as connection check and cost calculation.

When the analysis position reaches the last digit of the input kana character string and the analysis is completed for all the words of the input kana character string (step S265), next, based on the above analysis result, the minimum cost path is determined. Is searched (step S270). In this step, among the combinations that remain valid words in the processing so far and are linked to each other, the combination that minimizes the total cost given to each word is searched. Among the phrase segmentation candidates in FIG. 5, for example, when searching for a candidate “Tsuriagegai” + “Kawaru” that is “Handling” + “Alternative”, the total cost is 8 and is selected as the phrase segmentation candidate with the minimum cost. To be done. When a phrase segmentation candidate having the minimum cost is created in this way (step S2
80) Next, candidates for phrase segmentation writing candidates in each phrase are created (step S290). "Change" here
The homonyms such as “alternate”, “change”, and “alternate” are prepared as candidates for the word The priorities of the candidates in each of these clauses are determined by a predetermined method such as selecting the word used immediately before with the highest priority. here,
If the priority of “alternative” in the word “alter” is the highest, “alternate” is illustrated in FIG.
If the character string is written into the phrase, "Handling is" + "Alternative" shown by the dotted line B in the figure is obtained as the first candidate.

In the above explanation, for simplification of the explanation, the cost calculation is intended only for the cost of each word itself.
It is easy to obtain the desired conversion result, for example, by providing an inter-clause cost to reduce the cost between phrases that are determined to be easy to connect grammatically, or by providing an inter-word cost that focuses on the ease of connecting between specific parts of speech. It is possible to add various ideas for doing so.

Next, a word replacement process (step S120), which will be described later, is performed on the phrase segment writing candidate created by the above process, and the conversion result after the segment segment writing process is output to the CRT display 26 (display process ( Step S130) is performed and the Kana-Kanji conversion processing routine is ended.

Next, the structure of the word dictionary which is the main part of this embodiment will be described. As described above, in the case where the kana character string "Takeshikaiwakaiwa" illustrated in FIG. Here, the word "toriokai" means
Three types of notation, "handling", "handling", and "handling" are allowed. For "changing", for example, the three homonyms "change", "alternate", and "alternate" are "changing", "changing", Two types of notation are allowed, such as “alternate”, “alternate”, “alternate”, and “alternate”. Therefore,
When performing the phrase segmentation processing, similar processing such as connection check and cost check must be performed for all these notations, and information related to all these notations is stored in the independent word dictionary 58 or the above-mentioned dependency dictionary. Will need to be kept. On the other hand, in the present embodiment, for words for which a plurality of notations are allowed, the phrase segmentation processing is performed using only one of the representative notations. Therefore, in the phrase segmentation processing shown in the example in FIG. 5, not all of the other permissible notations shown in () are read from the word dictionary at the time of dictionary retrieval, and Information about this display notation is read together with one representative notation only for words that need to be replaced, and then the phrase segmentation processing is performed only for this one representative notation. At this stage, for words for which multiple notations are allowed, only one of the representative notations will be read from the word dictionary, and for words that need to be replaced with the notations for display described below, When the replacement is performed, the information about the display notation may be read from a predetermined dictionary.

The structure of the independent word dictionary 58 will be described below. As shown in FIG. 6, the word dictionary 58 includes a header, an index, and a dictionary body. The header is information for managing the dictionary itself. The index and the dictionary body are separately managed in a representative writing area, a derived writing area, and a meaning example area. The representative notation means that if one word is allowed to be expressed in multiple ways, for example, "handling" or "handling" is allowed as notation for "handling", then "handling" is used as the notation. Is a representative word notation. Derived notations are other notations that are allowed for one word, such as “handle” and “handle”. When the phrase segmentation process is performed, the phrase segmentation process is performed for a word having such a derivative notation using a representative notation that is representative of the derivative notation. The area for the meaning example is an area in which the information on the dependency described later is managed.

The representative notation, derived notation, and meaning examples are managed by the B-Tree structure in the dictionary body. The B-Tree structure is a well-known management structure adopted when a large amount of data is stored in the external storage. In the B-Tree structure of the present embodiment, no data is stored in the branch tree node and all the data is stored. The structure is placed at the end of the branch tree. Since the tree structure is probably a tree, the depth of the tree structure is smaller than that of a binary tree, and when using external memory,
It is known to have a characteristic of short access time (sometimes called B + Tree). An example of the B-Tree structure in the dictionary body is shown in FIG. When the B-Tree control block is followed to reach the word block based on the reading of the word (kana character string), the actual word data is placed here. In addition, this type of B-Tre
In the e structure, initially, the number of keys is set to be smaller than the maximum number of keys that can be placed in the section, and it is possible to flexibly deal with addition of data.

Below, based on the reading of words, B-Tre
The way in which the e control block is traced will be described with reference to FIG. For example, it shows a case where a kana character string “Kawaki” is input and the search is started in the representative writing area in the independent word dictionary 58. In the first row, which is the root of B-Tree, keys each having one character are arranged in the order of Japanese syllabary.
Which branch tree to proceed from "Kawaki" to
It can be known by comparing the character string "kawaki" with each key of the section S1 in the first row. In this example,
Since "kawaki" exists at least between the key "ka" and the key "sa", the node S2 in the second row is a search target.
Next, in this section S2, the same judgment is made, and it is judged that the reading of "Kawaki" is in the order of the Japanese syllabary after the key "Karu" and before the key "Kuri", and after the key "Karu" The node S3 in the third row corresponding to the range between "" is searched. When searching for "Kawaki" in the four keys in the third row, the word block corresponding to this range is searched because it is after the key "Kawa" and before the key "Kiji" in the order of the Japanese syllabary. It In the word block corresponding to this range, the actual word data related to the word having the pronunciation "kawaki" is stored.

The word data such as the representative writing area generally has a data structure shown in FIG. That is, the data indicating the record length X of the record that constitutes the word block is present at the head, and then X bytes of actual data follow. At the beginning of the actual data, the length Y of the headword is recorded, and subsequently, the headword of Y bytes is recorded. The actual word data follows. The word data length W is recorded at the beginning of each word data, and immediately after that, a 1-byte flag indicating the presence or absence of kanji data is recorded. After the flag, if there is Kanji data in the word data, the Kanji data is recorded. This Kanji data is composed of a Kanji data length and a Kanji code indicating an actual Kanji character string. After that, word information and part-of-speech information (in some cases, a plurality of pieces of part-of-speech information) are recorded. The word information consists of data indicating the length of the word information and the actual word information.

In this way, whether in the representative notation or the derivative notation,
The B-Tree structure can be used to retrieve information about a desired word based on the word heading string. As shown in FIG. 9, these pieces of word information are each composed of a separator and data following the separator. As the separator, a display Kanji separator that indicates that the data that follows is Kanji data, or a reading information separator that indicates that the data is reading information,
There is a derivative notation separator that indicates that it is a derivative notation. The display kanji is a record of the kanji corresponding to the representative notation in order to display the kanji by default when one word has a representative notation and a derivative notation. Although there are various forms of derivative notation as shown in FIG. 10, the notation itself permitted for each word is stored as derivative notation information. The derivative notation separator is information that exists only in the derivative notation area, and the lower 3 bits of the derivative notation stored after each derivative notation separator are the number of types of the derivative notation shown in FIG. It corresponds to. That is,
In the case of the derivative notation “sales” shown in FIG. 10, since it is “main rule” which is 1 of the derivative notation information and “send” which is 2, the lower 3 bits of the derivative notation separator have two types of derivative notation form. Has been shown to correspond. Further, the reading information gives the reading of a word, and gives the reading when the headword is a kanji. This information is used, for example, when the reading is obtained from the kanji and a word having the same reading is searched.
For example, in addition to inputting characters from a keyboard, when inputting by handwriting or the like, and when a character string including kanji is input together with a kana, it is used for correctly converting this (eg, “Higaku” → “airplane”). It should be noted that one representative notation and the derivative notation corresponding to this representative notation are managed in different areas. However, in this embodiment, the one in which the reading (heading) of a word and the word part of speech information match , I think there is a correspondence.

In this embodiment, the word search process (step S100) in FIG. 3 described above searches the independent word dictionary 58 and the adjunct word dictionary 68 described above. The operation at this time will be described in detail below. FIG. 11 is a flowchart showing details of this word search process (step S100 in FIG. 3). When this word search processing routine is activated, first, processing is performed in which the word search start position M is set to the value 1, that is, the beginning of the input kana character string (step S30).
2). Next, a variable L indicating the reading length in the word search
Is initialized to the value 1 (step S30
4), the word with the reading length L from the word start position is the independent word dictionary 5
8. A process of searching the auxiliary word dictionary 68 is performed (step S306). That is, the word having the first one character of the input character string as a reading is searched. afterwards,
It is determined whether or not a word matching the reading kana character string of length L from the start position M is found (step S30).
8). When the corresponding independent word is found, next, a process of reading the data related to the derived notation out of the data attached to the representative notation is performed (step S310).

When there are various derivative notations with respect to the representative notation and the derivative notations are stored in the derivative notation area shown in FIG. 6, the derivative notation is included in the word information of the representative notation area described above. A flag indicating whether or not replacement of is required is stored. Therefore, with reference to this flag, it is determined whether or not the word is instructed to replace the representative notation with the derivative notation (step S312). If a value indicating the necessity of replacement with the derivative notation is set in this flag, the representative notation retrieved previously is expanded in the expansion buffer, and the first mark MA is added to the word (step S314). ). The expansion buffer is a storage area for expanding all the representative notations and adjuncts that can form this kana character string with respect to the input kana character string, and is a memory area secured on the RAM 23. Further, at this time, if an instruction is given to replace the representative notation with the derivative notation, the derivative notation to be replaced is also stored in the word information in the representative notation area, and the derivative notation to be replaced is also the representative notation. Is expanded in the expansion buffer together with.

When the instruction to replace the derivative notation has not been given, or after the instruction to replace the derivative notation has been given and a mark has been added to the representative notation, the process returns to step S306 and the word of the reading length L is returned. Is repeated from the process of further searching. When it is determined that the word having the reading length L no longer exists in the independent word dictionary 58 and the auxiliary word dictionary 68 (step S308), it is determined whether or not the length of the search word can be extended (step S316). ). If M + L <A with respect to the total length A of the input kana character string, it is determined that the reading length L can be increased, and the reading is lengthened by one character (that is, L is incremented by 1). Processing) is performed (step S320). After increasing the reading length L by the value 1, the above-described processing is repeated from step S306. In the dictionary search, it is possible to determine whether or not a word having a reading with a length of L or more is stored. If there is no word with a reading having a length of L or more, the word It is also possible to cancel the processing after extending the reading and searching the dictionary.

As a result, the start position M is stored in the expansion buffer.
All reading words from length 1 to the maximum length in are expanded. When the reading length is gradually increased and the word length cannot be extended (step S316), the start position M
It is determined that the search for the word beginning with is completed, and then it is determined whether the start position M of the word search can be moved toward the end of the input kana character string (step S3).
18). If the position is movable, the process of incrementing the start position M by 1 is performed (step S32).
2) Return the reading length L to the value 1, and repeat the above-mentioned processing. Therefore, when these processes are performed, the expansion buffer has expanded all the representative notations and attached words that may form the input kana character string,
For the representative notation that needs to be replaced with the derivative notation, the first mark MA is added to the representative notation, and the derivative notation to be replaced is also expanded.

With the above processing, the word search processing (step S100 in FIG. 3) is completed. Subsequently, the above-mentioned segmentation processing (step S110) is performed, but these processings are performed using the above-described representative notation expanded in the expansion buffer, and actually explained using FIG. 4 and FIG. As described above, the word search process (step S10
0) and part of the segmentation process (step S110) are performed in parallel. However, here, as a result of the word search process (step S100), not all of the derived notations shown in () in FIG. 5 are expanded in the expansion buffer, and the representative notation that needs to be replaced with the derived notation. With respect to, the first mark MA is added, and the information on the derivative notation to be replaced is developed together with the representative notation.

When the word segmentation process (step S110) is completed, the word replacement process (step S130) is performed on the first candidate word. In this word replacement process, it is judged whether or not the word selected as the first candidate needs to be replaced with the derivative notation.
It is prepared as a notation that should be replaced in the derivative notation,
It replaces the representative notation with the notation expanded in the expansion buffer. Here, as will be described later, the processing performed when the notation used immediately before is learned as a derivative notation to be replaced will be described. For example, “Handling” + “Alternate” in FIG. 5 is the first candidate, and if the words used immediately before are “Handling” and “Alternate”, respectively, “Handling” is used as it is. However, as for “alternative”, the representative notation that has been consistently used in the word search and the segmentation processing is replaced with the derivative notation “alternative”.

FIG. 12 shows a processing routine for performing word replacement processing. The processing executed here will be described with reference to the above example of “handling” + “alternate”. In this processing routine, first, the word replacement process is started with the word of the first phrase segmented by the segment segmentation process as the target word (step S400). When the phrase is “handling” + “alternate”, the first word “handling” is the target word. Next, it is determined whether the target word has the first mark MA (step S410). If the first mark MA is attached, it can be determined that there is a derivative notation and replacement to the derivative notation is instructed, so processing for replacing the first candidate with the derivative notation that is the display notation is performed (step S4).
20). In step S410, the first mark MA is added to the target word.
If is not attached, the display notation is not replaced. That is, the display notation is the representative notation itself if the representative notation has been used before, and the derivative notation if the derivative notation has been used and learned before. In the target word “handling”, since the notation used immediately before is “handling”, the flag stored in the ancillary data of the representative notation does not instruct replacement to the derived notation and is shown in FIG. In step S312 of the word search processing routine, it is determined that the word does not need to be replaced. Therefore, the representative mark “handling” does not have the first mark MA, and the representative mark is a display mark as it is.

Then, it is checked whether or not there is an unprocessed word (step S430), and if there is an unprocessed word, the target word is shifted backward by one (step S440). In the example of "handling" + "alternate", the next target word is "ga", but this is a particle and is a word that does not have a derivative notation registered in the adjunct dictionary 68, so replacement is performed. Absent. When the process returns to step S430 and the target word is moved back by one (step S440), the next target word is "alternate". As for “Alternate”, since “Alternate” is selected immediately before, the flag saved in the data included in the word information that stores the representative notation indicates replacement with the derivative notation and is expanded in the expansion buffer. The first mark MA is given to the representative representation "replace" thus generated (step S410), and the derivative representation to be replaced is also developed. Therefore, the representative notation "alternate" is replaced with the derivative notation "alternate" (step S420). When the above processing is completed, there are no unprocessed words left (step S430), and the word replacement processing is ended. After that, as described above, in the display process of step S130 in FIG. 3, the conversion result "handling is changed" is displayed on the CRT display 26, and the kana-kanji conversion process is ended.

According to the embodiment described above, when performing the segment segmentation writing process, only the representative notation is processed. Therefore, the dictionary used for the segment segment segmentation processing only needs to store information on the representative segmentation. Has the advantage that the size of can be reduced. Further, by performing the phrase segmentation writing process using only the representative notation, the segment segmentation segmentation process can be simplified and the segment segment segmentation writing process can be speeded up. In the example of "Take care change", if the kana-kanji conversion device of this embodiment is not used, all the derived notations shown in () in FIG. However, in the kana-kanji conversion device according to the present embodiment, the derivative notations in parentheses other than the representative notations are not the target of the phrase segmentation processing. Therefore, the number of words to be subjected to the phrase segmentation processing is reduced by the number of derived representations other than the representative notation, the entire process is simplified, and as a result, the segment segmentation processing can be performed at higher speed. Further, since the number of words to be subjected to the phrase segmentation processing is reduced, the memory used as the expansion buffer can be made small.

In the above example, the representative notations "handling" and "alternative" are represented by "learning" and "alternating" after learning the notation used immediately before.
Was set in advance, but learning which notation to use as a display notation when there is a derivative notation is to write that notation in the word information of the representative notation area when a certain notation is selected, and , Move the order of the word data of each derivative notation in the derivative notation area to the selected notation at the top,
It is realized by shifting the order of the remaining derivative notations one after the other. When the conversion result is displayed next time, the notation stored in the word information of the representative notation area is set as the first candidate, and the remaining candidates are displayed according to the arrangement order of the word data stored in the derived notation area. . At this time, if the representative notation is used as the display notation, the representative notation itself (or derivative notation information corresponding to the representative notation) may be stored in the word information of the representative notation area. As in the embodiment, the flag indicating that the representative notation is used may be stored so that the notation is not replaced after the phrase segmentation processing.

In the above embodiment, the display notation is stored in advance in the word information of the representative notation area, and during the word search process (step S100), this display notation is also stored in the expansion buffer. Then, in the word replacement process (step S120), the representative notation is replaced with this display notation according to the mark given to the representative notation indicating the replacement. Therefore,
In the word replacement process, there is an effect that it is not necessary to refer to the dictionary again. However, in the word search process (step S100), the display notation is not expanded in the expansion buffer, and only the mark indicating the replacement is added to the representative notation, and the mark added in the word replacement process (step S120) In accordance with the instructions, the display notation may be read by referring to the derivative notation area and replaced with the representative notation. In this case, it is necessary to refer to the derived notation area again in the word replacement process (step S120), but it is necessary to expand the display notation in the expansion buffer or store the display notation in the representative notation area. Absent. At this time, the learning of the notation for display may be performed by moving the word data of the derived notation used to the beginning of the word data of the derived notation and replacing the derived notation at the top as the notation for display. It may be stored in the word information in the derived writing area so that it can be referred to during the processing (step S120).

Such a learning function of the display notation may be performed by rewriting the notation selected immediately before as described above, or by rewriting the notation most frequently selected from the notations selected in the past. It is also possible to change the notation that the user prefers and select. As a result, it becomes easier to obtain the desired notation in the separated character string after conversion.

The learning function of the display notation when a certain notation is selected has been described above, but the learning performed when a certain notation is selected is not limited to that performed in the derived notation corresponding to the same representative notation. Absent. That is, when a certain notation is selected, the priority of the representative notation corresponding to this notation in the homonyms increases and the independent word dictionary 5
Learning is also performed in the representative notation area of 8. This makes it easier to obtain a desired word among the homonyms.

In the phrase segmentation writing process, the dependency judgment may be performed by referring to the example dictionary, and the words for which the dependency is established may be prioritized. Regarding homophones, they are prioritized according to a predetermined method, such as using the most recently used word as the highest priority. Among the homonyms corresponding to, the priority of "alternate" was the highest, so it was divided into "handling" + "alternate". At this time, the dependency judgment is also performed in the phrase segmentation process, and if the dependency relations of “handling” and “change” are stored in advance in the example dictionary, regardless of the priority order of the homonyms. , "Handling" + "Change" shown by the solid line G in FIG. 5 can be obtained as the first candidate.
Here, the example dictionary is, as shown in FIG.
It is also possible to store it in the area of the meaning example together with 8.

Further, if such dependency judgment is performed, the effect of learning which notation is selected as the display notation extends to other dependency information related to the same representative notation. That is, if the display notation of the representative notation “change” is “changed”, then the “separate change” is processed into phrase segmentation, and the dependency information of the “rule” and the representative notation “change” is referred to. The sentences are segmented and the first candidate is "rules change" instead of "rules change".

When a new dependency relation is recognized between words, this dependency relation may be learned and stored in the part-of-speech data of the meaning example area. Similar to the above-mentioned derivative notation, the meaning example is also stored in association with the representative notation, and the learned dependency relation is referred to in the phrase segmentation processing using the representative notation, and the desired kana-kanji mixed characters are referred to. Easier to convert to columns.
However, in this case, it is also preferable that when the special derived notation is selected and confirmed in the kana-kanji character string, the modification relation is not learned. For example, in FIG. 10, it is assumed that a document relating to sales proceeds is created by selecting the symbol “¥” corresponding to the form number 15 of the derivative notation for the character string “listening”. At this time, even if the relationship is learned between the character "\" and the character "yen" written in the vicinity of the character "\", even if the relationship is learned, the character "\" It is not necessary to recognize the semantic connection between "" and the character "yen". Therefore, if a special derivative notation is selected, the dependency learning is not performed. As the means, for example, FIG.
Of the types of derivative notation shown in, it is possible to specify in advance that the dependency relations will be learned in advance from “main rule” of form number 1 to “in Japanese grammar” of 8 and it is also possible to specify the special notation At the time of selection, the user may specify with an option switch that learning of the dependency relationship is not performed.

In the present embodiment, as shown in FIG. 6, the reading of each word, one representative notation, and the derivative notation, if any, are stored in the integrated independent word dictionary 58. The word dictionary having the word reading and one representative notation as constituent elements may be separated from the derivative notation dictionary in which the permitted notation is stored in association with the representative notation. At this time, in the phrase segmentation process, the representative notation is read by referring to the word dictionary, and after the segment segmentation process, the word having the derivative notation is referred to the derivative notation dictionary and is rewritten to the display notation and then displayed. Of. With such a configuration, the word dictionary needs to store the reading and representative notation of all the words to be registered, but the derivative notation dictionary allows the derivative notations that are allowed for words having derivative notations. It suffices to store in association with the representative notation.

If the conversion candidate displayed as a result of performing the phrase segmentation processing by the kana-kanji conversion device of the first embodiment described above is not the desired notation, the keyboard 2
It is necessary to display other candidates by inputting a conversion key from 4 and to select a desired notation from these, and a selection method for such a second candidate and below will be described below as a second embodiment. To do. FIG. 13A shows a display method used when an instruction to display conversion candidates is given in the conventional kana-kanji conversion device. If you enter the conversion candidate instruction for "change" in "change", "change"
The word with the reading and its derivative notation are displayed one-dimensionally in a state in which homonyms and derivative notation are mixed according to the order sorted by a predetermined learning function or the like. On the other hand, in the kana-kanji conversion device of the present embodiment, as shown in FIG.
As shown in (B), the homonyms of the word displayed as the first candidate and the derivative notation of each word are expanded in two different directions and displayed two-dimensionally. When the first candidate of "alter" is "alternate", "alternate" which is a derivative notation of "alternate" is displayed in the horizontal direction, and the homonym of "alternate" is displayed in the vertical direction. "Change" and "Change" are displayed. Then, for each homonym, the derivative notation is expanded and displayed in the same manner as in the case of the first candidate on the first line in the horizontal direction. In FIG. 13, the derivative notation is expanded in the horizontal direction and the homonyms are expanded in the vertical direction, but the types of words expanded in the vertical and horizontal directions may be reversed.

The conversion candidate display processing routine executed in the second embodiment will be described below with reference to the flowchart of FIG. This routine is started when the display of conversion candidates is instructed by inputting the conversion key. When the conversion key is input, first, the representative notation of the first candidate word and its homonyms is read from the words expanded in the expansion buffer (step S510). In the above example, "change" and "alternate" are read for "alternate". The memory used for reading at this time is RA
This is a predetermined area secured on M23, and the data that is a conversion candidate is temporarily stored. Then the first
It is determined whether or not all homonyms including candidates have derivative notations (step S520). Here, it is determined whether or not a derivative notation exists in the derivative notation area of the independent word dictionary 58, regardless of the flag indicating the necessity / unnecessity of replacement with the display for display. If the first candidate word and its homonyms have derivative notations, the independent word dictionary 58
Is read (step S530).

When the derived notation is read, all of these notations are rearranged according to the two-dimensional display format shown in FIG. 13B (step S540). In the first line, the first candidate word and its derivative notation are expanded according to a predetermined order rearranged by a learning function or the like, and in the first column, the first candidate word and its homonyms are Similarly, develop according to a predetermined order. In the example of FIG. 13, the first candidate “replace” and the second candidate “replace” are in the first row, and the first candidate “replace”, the third candidate “change” and the fifth candidate “replace” are in the first row. Line up. In the second row and the second column and thereafter, the derivative notation of the notation in the first column of each row is arranged such that the number indicating the transformation form of each derivative notation is the same for each column. In "kawaru", there are two types of derivative notation, that is, the "main rule" of 1 and the "warm" of 2 which sends the "waru" which is the feed, and the "4" which sends "ru" which does not feed the "allowance" shown in FIG. Therefore, it is developed as shown in FIG. 13 (B).
The result of rearranging in this way is output and displayed on the CRT display 26, and this routine is ended.

In the embodiment described above, the homonyms and the derivative notations are expanded in two different directions and displayed two-dimensionally, but FIG. 15 shows an example of a character string "replace". The configuration is also suitable. In FIG. 15 (A), each derivative notation is displayed as a separate group for each homonym. Here, first, the word corresponding to the desired notation is selected from the homonyms, and then the desired notation is selected from the derivative notations. In these operations, the conversion candidate display processing executed by the kana-kanji conversion device is the same as that in the second embodiment, and in steps S540 and S550, the homonyms and the derivative notations are rearranged and displayed only in different formats. Is. Further, as in the above example, the homonyms and the derivative notation are not displayed at the same time, but as shown in FIG. 15B, first, the word to be displayed as the first candidate and its homonyms are displayed. After that, the display notation may be displayed in a selectable state upon receiving a predetermined instruction. With such a configuration, first, when a word corresponding to a desired notation is selected from homonyms, the derivative notation of the word is newly displayed. FIG. 15 (B)
Then, it is indicated that when the second candidate “replace” is selected instead of the first candidate “replace”, a derivative notation of “replace” is displayed. Here, the desired notation is selected.

According to such a display method, it becomes very easy to select the desired notation when the first candidate is not the desired notation. Figure 13 with various notations mixed
Unlike in the case of (A), when you want to unify the way of sending, or when you want to select a way of sending a specific word different from other words, you can select the desired notation from multiple candidates. You can easily find it. here,
In the second embodiment, the display of the conversion candidates by the above display method is displayed when the conversion instruction is input by the conversion key input or the like when the displayed first candidate is not the desired notation. Alternatively, when performing automatic conversion collectively, the words currently being converted may be automatically displayed sequentially.

In the word dictionary storing the derivative notation, if the notation using the different characters is also stored, the notation using the different characters will be used when the display of other conversion candidates is instructed by the conversion key input. It is displayed at the same time and can be selected. As a third embodiment, a kana-kanji conversion device having a configuration capable of easily selecting a notation using different characters will be described below. In the third embodiment, when displaying a kana-kanji mixed character string candidate, it is clarified that a variant character notation exists in a word having a notation using variant characters, and if a variant character notation is desired at this time, it is simple. Different characters can be selected by input operation.

In this embodiment, the word search processing routine shown in FIG. 16 is executed in place of the word search processing routine shown in FIG. 11 as the word search processing in the phrase segmentation processing. In the process of FIG. 16, the process flow of extending the reading length of the word to be searched one by one from the beginning of the input character string and further extending the start position of the word to be searched one by one is the same as in FIG. 11. Therefore, the last two digits of the corresponding step numbers are made the same, and detailed description is omitted. The operation of the part different from the process of FIG. 11 will be described below. In the third embodiment, it is determined in step S612 that replacement with the derived notation is necessary, the first mark MA is added to the representative notation, and the representative notation is expanded in the expansion buffer together with the notation for display (step S613). Alternatively, after it is determined in step S612 that the replacement with the derivative notation is not necessary, it is further determined in step S614 whether the derivative notation includes a notation using variant characters. This judgment can be easily performed by saving the flag indicating that there is a need to change to a derivative notation in the word information in the representative notation area described above as well as the flag indicating that a notation using variant characters exists. can do. Here, the first mark MA is recognized during the word replacement process described below, and the representative notation is rewritten to the display notation.

If it is determined in step S614 that there is a notation using different characters, the second mark MB is added to the representative notation and the representative notation is developed (step S61).
5). If it is determined in step S614 that there is no variant character, or if there is a variant character and the second mark MB is added in step S615, step S6 is performed again.
Return to 06. After searching for a word in the entire input character string, a predetermined grammar check and cost check are performed using these representative notations. After the phrase segmentation process, if the conversion key is input when the displayed first candidate is not the desired notation, the conversion candidate display processing routine of FIG. 14 is executed, and the same sound is output in the format shown in FIG. It is possible to display other conversion candidates such as synonyms and derivative notations of each word, and it is also possible to select a notation with a desired variant character from among these. However, in the present embodiment, especially by providing the different character selection means, the notation of the first candidate can be immediately converted into the notation using the different characters.

The word replacement processing routine executed in the third embodiment will be described below with reference to the flowchart shown in FIG. Here, the last two digits of the steps common to the word replacement processing routine shown in FIG. 12 are the same, and detailed description thereof will be omitted. In FIG. 17, it is determined whether or not the target word has the first mark MA (step S710), and if the first mark MA is present, the first candidate is replaced with the display notation (step S72).
0) Further, it is determined whether or not the target word has the second mark MB (step S723). Second mark MB
If there is, a third mark MC is added to the display notation (step S726), and thereafter, as in FIG. 12, it is checked whether there is an unprocessed word (step S730). The process of is repeated. Step S72
If the target word does not have the second mark MB in step 3, the process proceeds to step S730 to check whether there is an unprocessed word.

When the word replacement processing routine shown in FIG. 17 is finished and the first candidate display notation is displayed, the foreign character display processing routine shown in FIG. 18 is executed. First, in step S800, it is determined whether or not the third mark MC is included in the display notation. If there is the third mark MC, an "*" mark is displayed at a predetermined position on the screen when the first candidate is displayed (step S810) to clearly indicate that there is an inscription in different characters. Next, it is determined whether or not the "shift key" + "conversion key" is input (step S820),
When the input is made, it can be determined that the display using the different characters is selected, so the display notation is converted into the notation using the different characters for display (step S830). If there is no third mark MC in the display notation in step S800, and if the "shift key" + "conversion key" is not input in step S820, the display in different characters is not performed and normal conversion is performed.・ The confirmation process is performed.

For example, when the character string "Ichimanen" is input, when the first candidate "10,000 yen" is displayed,
At the same time, "*" is also displayed at a predetermined position on the screen to clearly indicate the presence of the variant character. If you enter the "conversion key" here, homonyms such as "Position", "1", "City", "Ichi", etc. will be displayed, but you can directly enter by pressing the "Shift key" + "Conversion key". It is possible to convert variant characters into "Ichi". Similarly, for "10,000" and "yen", it is possible to know that there are different characters by the display of "*", so you can easily select the different characters "Man" and "En" by entering "Shift key" + "Conversion key". . Further, when only "Ichi" is to be changed into a different character, the character string "Ichiman Yen" can be obtained by performing the confirmation process at the time when the different characters are converted into "I". In addition, by storing a variant character string "Ichiman-en" in the word dictionary, you can convert from "10,000 yen" to "Ichiman-en" at once by entering "Shift key" + "Conversion key". It may be configured as

As described above, in the present embodiment, by displaying the "*" mark at a predetermined position on the screen, it is made clear that there is a display with different characters, and by inputting "shift key" + "conversion key". By selecting the display with different characters,
It is possible to immediately convert the notation of the first candidate into the notation using different characters.

At this time, if the display notation in the word dictionary is rewritten to the variant character notation when the variant character selecting means is instructed to select the variant character, the conversion result of the same word will be changed to the variant character from the next time. The character notation will be displayed as the first candidate. Therefore, when a user wants to create a document using the notation of different characters, the desired notation can be easily obtained as a conversion candidate. On the other hand, when the different character notation is selected by the different character selecting means, the display notation in the word dictionary may not be rewritten. This is because the variant character notation is a special notation and it may not always be a good idea to rewrite the notation for display.

Next, a fourth embodiment of the present invention will be described. The kana-kanji conversion device according to the fourth embodiment includes, in addition to an ordinary word dictionary referred to when converting kana-kanji, date and time information indicating a predetermined time point such as time and date, and a specific word and date information indicating this date and time information. A date and time term conversion dictionary that stores the headwords and the corresponding words is provided. FIG. 19 shows an example of the structure of this date / time term conversion dictionary. The date-and-time term conversion dictionary is configured with a unit of the reading of a specific headword representing date-and-time information and the type of date-and-time information corresponding to this specific headword. Is also stored as a unit. For example, in FIG. 19, “current time” is stored as the type of the date and time information corresponding to this, which is one of the readings of the entry word representing the date and time information. "% H:% M" is prepared as the format. Therefore, when converting a kana character string by referring to this date and time term conversion dictionary, if the character string "now" is input and conversion processing is performed, the current time is displayed as "10:45". Is possible.

The process of converting a specific entry word representing such date / time information into corresponding date / time information will be hereinafter referred to as date / time conversion. When the phrase segmentation processing is started for the input kana character string and a normal dictionary search is performed, the date / time conversion processing shown in FIG. 20 is executed in parallel with this. When this date / time conversion process is started, the top of the input character string is set as the search start position, as in the normal word reading process (step S900). First, the variable L indicating the reading length in the word search is initialized to the value 1 (step S910), and then it is checked whether or not there is a specific headword whose reading length is L starting from the search position. (Step S920). If a character string corresponding to the reading of a specific entry word representing date and time information such as "now" or "kyo" is found, the date and time term conversion dictionary and the timer 33 shown in FIG. Date conversion is performed according to the entry word (step S930). Timer 3
Since 3 has generated date and time information at the present time, when the type of date and time information to be converted is determined by referring to the date and time term conversion dictionary, execution of date and time conversion is performed based on the current date and time information generated by the timer 33. Is possible. For example, if a specific headword "Kyo" is found, it is known from the date / time term conversion dictionary that the word is converted to today's date, and by referring to the timer 33, the specific today's date ("1 Information such as "1st of the month".

This conversion result is temporarily stored (step S
940), a date / time term mark is attached to the specific entry word representing the date / time information (step S950), and the phrase segmentation processing is continued. Then, it is judged whether or not the word length to be searched can be extended (step S96).
0), L is L + 1 if the word length can be extended
As a result, the process of lengthening the reading by one character is performed (step S970), and then the process returns to step S920 again to continue the search for the specific headword. Step S96
If it is determined that the word length cannot be extended at 0, then it is checked whether the search start position can be moved backward (step S980). When the search start position is the last digit of the input character string and it cannot be moved any further backward, the time conversion process ends. When it is determined in step S980 that the search start position can be moved backward, the search start position is moved backward by one, and the process returns to step S910 to initialize L to the value 1 again, and The process of is repeated.

In step S920, if a specific headword having the reading length L is not found, the process proceeds to step S960 and it is determined whether the word length can be extended. After that, as described above, if the word length can be extended, the reading length L is extended by a value of 1 in step S970, and a specific headword is searched again. If the word length cannot be extended, the search is performed in step S980. It determines if the starting position can be moved backwards.

Next, FIG. 21 shows a flowchart of the date and time rewriting process performed when displaying the conversion candidates after segmentation and segmentation of a specific headword whose date and time has been converted by the date and time conversion process. This date and time rewriting process is
Of the words remaining as conversion candidates after performing the phrase segmentation processing on the input kana character string, the first word is started as the target word (step S1000). First, it is checked whether or not the target word has a date / time term mark (step S1010). If the date / time term mark is attached, it is determined that the word is a specific entry word representing date / time information, and thus the word with the date / time term mark is rewritten to the result of date / time conversion (step S102).
0). This is to display the conversion result once stored after the date and time conversion in the date and time conversion processing described above according to the date and time display format stored in the date and time term conversion dictionary of FIG. 18, and rewrite it with the word with the date and time term mark. Is. For example, "Today" with the date and time term mark
If this word is found, this word is stored after being converted into information that today's date is January 1 by the date / time conversion process, so the conversion result is stored in the date / time display format “% M month”. % D days ". The result of rewriting the word according to the date / time display format is once again stored.

After that, it is checked whether or not there is an unprocessed word (step S1030), and if there is no unprocessed word among the words remaining as conversion candidates, this date / time rewriting process is terminated. If it is determined in this step S1030 that there is an unprocessed word, the target word is moved back by one (step S1040), and the process returns to step S1010 to search for a word with a date / time term mark, and the above-mentioned processing is performed. Is repeated. If the target word does not have the date / time term mark in step S1010, the process moves to step S1030, and it is checked whether or not there is an unprocessed word among the candidate words segmented into segments. If there is an unprocessed word, the target word is moved backward by one in step S1040, and the above processing is repeated until it is finally determined in step S1030 that there is no unprocessed word.

When all the processes are completed and the first candidate kana-kanji mixed character string is displayed, the specific headword representing the above date and time information is first processed as a regular word by segmentation. Is displayed.
The conversion result of the Kana-Kanji character string "Kyou" is:
For example, "today" is displayed as the first candidate. Here, if a date / time conversion result is desired, a normal conversion instruction may be input. By inputting the conversion instruction, "January 1st" which is the date and time conversion result can be displayed together with other candidates such as "Kyoto" and "Kyo", and can be selected.

In the fourth embodiment, only one type of date and time display format is prepared for each specific entry word in the date and time term conversion dictionary, but a plurality of date and time display formats are registered. It may be configured.
A kana-kanji conversion device having such a configuration will be described below as a fifth embodiment. FIG. 22 shows an example of the structure of the date and term conversion dictionary provided in the kana-kanji conversion device of the fifth embodiment. Here, for example, four types of date and time display formats are prepared for a specific entry word that represents the date and time, and these are stored in the dictionary according to the order of use priority when displaying conversion candidates. ing. Now
The format with the highest priority of use is "% M Month% D
Since it is "day", if you enter the character string "today" and perform time conversion, "January 1
Today's date such as “day” is displayed as the first candidate for the conversion result. At this time, if "January 1st" is not the desired conversion result, by inputting the conversion key, it is possible to display the conversion results of "January 1st (Tuesday)" for the second candidate and below.

From the conversion results below the second candidate, "1" is selected.
If you select the candidate "Tuesday, month 1", the format "% M month% D day% W day" used by the selected candidate will be used in place of the "% M month% D day" format. It becomes the highest format, and the ranking in the date and time term conversion dictionary is rewritten. Therefore, if the time conversion of the specific headword "Kyo" is performed thereafter, the result of date and time conversion using the notation according to the format "% M month% D day% W day of the week" as the first candidate is displayed. It will be. According to the kana-kanji conversion device of the fifth embodiment described above, the format for displaying the date and time selected by the user is learned as the format with the highest priority of use, so time conversion is performed according to the format that the user prefers. It becomes extremely easy to display the result of.

At this time, the result of the date / time conversion may be displayed as a derivative notation of a specific headword representing the date and time, or may be displayed as a homonym of the specific headword. If displayed as a derivative notation, the notation indicating the date and time such as "January 1"
This is preferable because it is treated as a notation related to "today" indicating the date and time, not "Kyo" or "mirror". On the other hand, if it is displayed as a homonym, at the time of the display, the format of the first candidate is treated as a homonym and the other prepared formats are treated as derivative notations of the homonym. Therefore, the result of the time conversion can be incorporated into the display method shown in the second embodiment and displayed together with other homonyms and derivative notations.

In the fourth and fifth embodiments, the date / time display format is stored in the date / time term conversion dictionary. However, the date / time display format is stored separately from the date / time term conversion dictionary. It doesn't matter. If the date and time display format is stored separately from the date and time term conversion dictionary, the date and time display format is for displaying the time, for displaying the date, for displaying the calendar year, and for each type. It is possible to store and refer to a format of a type suitable for each specific headword.

With the above arrangement, it is possible to time-format the input numbers. That is, "11
By inputting a number such as "47", the conversion result of "11:47" is displayed. The operation when such time formatting is executed will be outlined below. When the phrase segmentation processing is executed for the input character string, the time formatting processing shown in FIG. 23 is executed in parallel with the dictionary search in the text. First, the head of the input character string is set as the search position (step S11).
00), it is determined whether there is a 4-digit number (step S1110). If a four-digit number is found, a time mark is attached to this number (step S1120). It is used as a time word for the subsequent phrase segmentation processing. After that, it is checked whether or not the search position is moved backward (step S11).
30), if possible, shift the search position backward by one digit and return to step S1110. If a four-digit number is not found in step S1110, step S113
It shifts to 0, and a 4-digit number is searched until the end of the input character string.

When displaying the result of writing the input character string segmented into phrases as conversion candidates, in consideration of the above processing, for the character string with the time mark, the first two digits are "% H" and the second Digit numbers are treated as "% M". Therefore, if the usage priority of "% H hour% M minutes" is the highest in the date and time display format for displaying the time prepared in advance in the dictionary, the conversion after the phrase segmentation processing of "1147" is performed. The result is displayed as "11:47".

In the fourth and fifth embodiments, the date and time conversion result is displayed by inputting a predetermined conversion instruction after displaying the result of the ordinary kana-kanji conversion processing. However, it is also preferable to directly display the result of date and time conversion. For example, when a kana character string "Kyou" is entered, the result of converting it as a normal word is displayed as the first candidate, and if the conversion key is entered, other derivative notations and homophones except the date and time conversion result are displayed. The synonyms are displayed, and the date and time conversion result by inputting the "conversion key + shift key" is displayed in a plurality of prepared formats. At this time, it is not necessary to select a desired notation from the mixture of the date and time display format and the derivative notation, and the date and time display format may be referred to only when the "conversion key + shift key" is input. In the above example of the word "kyo", "today" is obtained as the first candidate, and if the conversion key is input, "kyo", "kyo", "mirror", etc. below the second candidate are displayed. At this time, by inputting the "conversion key + shift key", the date and time conversion result is displayed and can be selected. Of course, in the prepared multiple date and time display formats,
The user's selection situation is learned, and the format selected immediately before is displayed as the first candidate. With such a configuration, the date / time conversion result does not appear as a conversion candidate when the date / time conversion is not desired, and the desired notation can be easily selected even when the date / time conversion is desired.

In the above-described embodiment, the case where the kana character string input from the keyboard is written into the phrase is described. It is also preferable that the conversion device is capable of handwriting input, and that the character string “handwritten” is recognized and converted into “airplane”.

The embodiments of the present invention have been described above.
The present invention is not limited to these embodiments, and for example, a configuration using another phrase segmentation method such as the two-segment longest matching method instead of the minimum cost method is within the scope not departing from the gist of the present invention. It goes without saying that the above can be implemented in various modes.

[Brief description of drawings]

FIG. 1 is a block diagram showing hardware for realizing a kana-kanji conversion device according to an embodiment of the present invention.

FIG. 2 is a functional block diagram showing an implementation form of a kana-kanji conversion function in a kana-kanji conversion device as an example.

FIG. 3 is a flowchart illustrating a kana-kanji conversion processing routine executed in an embodiment of the present invention.

FIG. 4 is a flowchart showing a phrase segmentation writing process executed in a segmentation segment writing unit 102.

FIG. 5 is an explanatory diagram showing a state of phrase segmentation writing processing by the minimum cost method.

FIG. 6 is an explanatory diagram showing an internal configuration of an independent word dictionary 58.

FIG. 7 is an explanatory diagram showing how a representative notation area is managed.

FIG. 8 is an explanatory diagram showing a structure of word data.

FIG. 9 is an explanatory diagram showing details of word information together with each separator.

FIG. 10 is an explanatory diagram showing an example of derivative notation.

FIG. 11 is a flowchart showing a word search process in the first embodiment.

FIG. 12 is a flowchart showing a word replacement process in the first embodiment.

FIG. 13 is an explanatory diagram showing a method of displaying conversion candidates in the second embodiment.

FIG. 14 is a flowchart showing the conversion candidate display processing.

FIG. 15 is an explanatory diagram showing another example of the conversion candidate display method.

FIG. 16 is a flowchart showing a word search process in the third embodiment.

FIG. 17 is a flowchart showing word replacement processing in the third embodiment.

FIG. 18 is a flowchart showing a different character display process in the third embodiment.

FIG. 19 is an explanatory diagram showing an example of a date and time term conversion dictionary in the fourth embodiment.

FIG. 20 is a flowchart showing date / time conversion processing in the fourth embodiment.

FIG. 21 is a flowchart showing date / time rewriting processing in the fourth embodiment.

FIG. 22 is an explanatory diagram showing an example of a date and time term conversion dictionary in the fifth embodiment.

FIG. 23 is a flowchart showing a time formatting process.

[Explanation of symbols]

 21 ... CPU 22 ... ROM 23 ... RAM 24 ... Keyboard 25 ... Keyboard interface 26 ... CRT display 27 ... CRTC 28 ... Printer 29 ... Printer interface 31 ... Bus 32 ... Hard disk 33 ... Timer 40 ... Character input section 42 ... Conversion control section 44 ... Converted character string output unit 50 ... Character string input unit 52 ... Character storage unit 54 ... Independent word candidate creation unit 56 ... Independent word analysis position management unit 58 ... Independent word dictionary 64 ... Adjunct word creation unit 66 ... Adjunct word analysis Position management unit 68 ... Adjunct dictionary 72 ... Independent word learning unit 74 ... Auxiliary word learning unit 76 ... Affix learning unit 78 ... Character conversion learning unit 80 ... Word data creation unit 82 ... Connection verification unit 84 ... Connection verification table 86 ... Cost Calculation unit 100 ... Word data storage unit 102 ... Phrase segmentation writing unit 106 ... Phrase data case Part 108 ... conversion character string output unit

Claims (20)

[Claims] 1. A kana character string is input, a dictionary is referred to,
A kana-kanji conversion device that writes out the input kana character string by segmenting and generates kana-kanji mixed character string candidates, including a word dictionary storing word reading and one representative notation as constituent elements, and a plurality of word dictionaries. For a word for which notation is allowed, a derivative notation dictionary in which the allowed notation is stored in association with the representative notation, and for the input kana character string, the word dictionary is referred to, and the representative notation is used. If there is any other notation allowed for the words constituting the bunsetsu obtained by the bunsetsu segmenting process, and the bunsetsu segmenting unit for performing the segment segmenting process,
A word notation selecting unit that determines a predetermined notation as a display notation, and a display that displays the word represented by the determined display notation as the first candidate of the words that constitute the kana-kanji mixed character string candidate A kana-kanji conversion device equipped with means. 2. The kana-kanji conversion device according to claim 1, wherein the word notation selecting means includes means for determining the display notation by referring to the derived notation dictionary. 3. The kana-kanji conversion device according to claim 1, wherein the word notation selecting means stores a corresponding notation from the derivative notation dictionary in the word dictionary based on the use of past derivative notations. A kana-kanji conversion device comprising means and means for determining the display notation by referring to information stored in the word dictionary. 4. The kana-kanji conversion device according to claim 1, wherein the word dictionary and the derivative notation dictionary are integrally configured. 5. The kana-kanji conversion device according to claim 1, wherein the display means allows homonyms of the word displayed as the first candidate and at least the word of the first candidate. A kana-kanji conversion device having division display means for displaying in a form distinguishable from notation. 6. The kana-kanji conversion device according to claim 5, wherein the split display unit has a homonym of a word displayed as the first candidate, a word of the first candidate, and a homonym thereof. A kana-kanji conversion device that is a means for expanding the notation that is allowed for two different directions and displaying other word notations for the first candidate word notation two-dimensionally. 7. The kana-kanji conversion device according to claim 5, wherein the split display unit displays the homonyms together with the word displayed as the first candidate, and then receives a predetermined instruction, A kana-kanji conversion device comprising means for displaying an acceptable notation for at least one of the first candidate word and / or the homonym. 8. The kana-kanji conversion device according to any one of claims 1 to 4, wherein, among the allowed notations including the first candidate, there are notations using different characters, the display means displays the first When displaying one candidate word, a variant character existence display means for displaying that a variant character exists for the word of the first candidate, and converting the word notation of the first candidate into a notation including a desired variant character A kana-kanji conversion device having means for selecting different characters. 9. The kana-kanji conversion device according to claim 1, wherein the word notation selecting means learns a situation in which a user selects a word notation from the plurality of allowed notations, A kana-kanji conversion device that determines subsequent display notations based on the result of the learning. 10. The kana-kanji conversion device according to claim 9, wherein, based on the result of the learning, the order of a plurality of allowed notations stored in the derived notation dictionary is rearranged and stored,
A kana-kanji conversion device equipped with a derivative notation dictionary learning means for preparing the plurality of allowed notations so that the rearranged order can be displayed as a conversion candidate rank. 11. The kana-kanji conversion device according to claim 9, wherein the display notation determined based on the result of the learning is stored in association with the representative notation to form a kana-kanji mixed character string candidate. A kana-kanji conversion device equipped with a display notation learning means that is prepared to be displayed as the first candidate of a word to be reproduced. 12. The kana-kanji conversion device according to claim 9, wherein when one word notation is selected by the user from the plurality of allowed notations, it corresponds to the selected word notation. Regarding a representative notation, a kana-kanji conversion device comprising word dictionary learning means for changing the priority order to be used as a candidate of a word that constitutes the kana-kanji mixed character string candidate within a homonym in the word dictionary. 13. The kana-kanji conversion device according to claim 9, wherein the word notation selecting means selects a notation using variant characters from the plurality of allowed notations,
Kana-to-Kanji conversion device that prohibits learning the selection status of word notation. 14. The kana-kanji conversion device according to claim 1, wherein in the process of generating the kana-kanji mixed character string candidates, it is learned that two or more words have been converted with a predetermined dependency relationship. And a learning method for prohibiting the learning of the dependency relationship when a predetermined derivative with low frequency of use is selected in at least one of the two or more words and is determined in a kana-kanji mixed character string. A kana-kanji conversion device equipped with means. 15. A kana-kanji conversion device for inputting a kana-character string, referring to a dictionary, and writing out the input kana-character string for each phrase, and generating a kana-kanji mixed character string candidate. Date and time information indicating a predetermined time such as month and day, and a date and time term conversion dictionary that stores a specific entry word representing the date and time information in association with each other, and current date and time information generating means for generating the date and time information at the present time, When the phrase segmentation processing is performed in accordance with predetermined grammatical information, a date / time term searching means for checking the presence / absence of a particular headword representing the date / time information, and a particular headword representing the date / time information are found by the date / time term searching means. At this time, referring to the date / time term conversion dictionary and the current date / time information generating means, date / time term conversion for converting a specific entry word representing the date / time information into the corresponding date / time information And a date / time display format storage unit for storing a plurality of date / time display formats, and a date / time display unit for displaying the conversion result in a predetermined format determined from the plurality of date / time display formats, When the used format is undesired, a date and time display format selection means for displaying the conversion result in the desired format selected from the rest of the plurality of date and time display formats, and one of the stored plurality of formats. Kana-kanji conversion provided with a date and time display format changing means for changing the date and time display format to be used when the user learns that the user has selected and the time and date information is displayed based on the learning result. apparatus. 16. The kana-kanji conversion device according to claim 15, wherein a result of converting a specific entry word representing the date and time information into corresponding date and time information is displayed, or a result of conversion as a normal word is displayed. A kana-kanji conversion device equipped with date / time conversion switching means for selecting whether or not to perform. 17. The kana-kanji conversion device according to claim 1, wherein date and time information indicating a predetermined time point such as time and date is stored in association with a specific entry word representing the date and time information. A date and time term conversion dictionary, a current date and time information generating means for generating the date and time information at the present time, and a date and time term searching means for checking the presence or absence of a specific headword representing the date and time information when performing the phrase segmentation processing according to predetermined grammatical information When a specific entry word representing the date / time information is found by the date / time term search means, the specific entry word representing the date / time information is referred to by referring to the date / time term conversion dictionary and the current date / time information generation means. To a corresponding date / time information and a conversion candidate of a specific entry word representing the date / time information is displayed. Date / time candidate mixed display means for displaying the conversion result to the date / time information in a predetermined format as one of the phonetic synonyms or as one of the allowed notations of a specific headword representing the date / time information. Kana-Kanji conversion device. 18. A kana-kanji conversion method for inputting a kana character string, referring to a dictionary, and writing out the input kana character string for each phrase, to generate a kana-kanji mixed character string candidate, wherein When performing segmentation and segmentation processing on a kana character string, the segmentation and segmentation processing is performed using the representative representation by referring to a word dictionary that stores word reading and one representative representation as constituent elements. If the words that compose the phrase obtained by the segmentation processing are words for which multiple notations are allowed, the predetermined notation is determined as the display notation, and the word represented by the display notation is mixed with the Kana-Kanji. A kana-kanji conversion method for displaying as a first candidate of a word constituting a character string candidate. 19. A kana-kanji conversion method for inputting a kana character string, referring to a dictionary, writing out the input kana character string for each phrase, and generating a kana-kanji mixed character string candidate. When performing the phrase segmentation processing for a kana character string, a specific headword representing time and date information indicating a predetermined time such as time and date is searched, and a specific headword representing the date and time information is found. When referring to the date and time term conversion dictionary that stores the date and time information and a specific entry word representing the date and time information in association with each other, the specific entry word representing the date and time information is referred to. To the corresponding date / time information, and the conversion result is displayed in a predetermined format determined from a plurality of stored date / time display formats, and the displayed format is undesired. At this time, the conversion result is displayed in a desired format selected from the rest of the plurality of date and time display formats, and it is learned that one of the plurality of stored formats is selected by the user. A kana-kanji conversion method for changing the date and time display format used when displaying the date and time information based on the above. 20. For a word in which a plurality of notations allowed in Japanese exist, the plurality of allowed notations are organized into one representative notation and a derivative notation which is the notation permitted except for the representative notation. A kana-kanji conversion device that uses the representative notation in the processing of segmentation and segmentation in kana-kanji conversion, and uses one of the representative representation and the derivative representation in the representation after segmentation and segmentation.