JP3329476B2 - Kana-Kanji conversion device - Google Patents

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 for inputting a kana character string and outputting a kanji-kana-kana character string.

[0002]

2. Description of the Related Art Kana-Kanji conversion devices occupy an important position in a Japanese interface of a computer. The kana-kanji conversion device aims to convert a sentence reading into a kanji-kana-mixed sentence.

However, in the conventional kana-kanji conversion device, the emphasis is on converting kana to kanji, so that the conversion result contains many kanji, and words that are usually written in hiragana are written in kanji. Tend to be. For example, reading information "at school" may be converted to "at school". If the user intends to write “at school”, the word “at” must be changed to “put” by holding down the “next candidate key” or “another candidate key”. In this case, if "OKA" is not registered in the Kana-Kanji dictionary, it cannot be converted by this method. In particular, since words written in kana are rarely registered in the kana-kanji dictionary,
This method often cannot be converted. In order to solve this problem, a method of adding a word in Kana notation as a homophone candidate when selecting a homophone has been performed. With this method, it is possible to convert even if a word consisting only of hiragana is not registered, but it must be pressed the required number of times, such as a "next candidate key" or "another candidate key", which is very time-consuming. . This is because a word in Kana notation is not registered in the Kana-Kanji conversion dictionary, and is therefore often the last candidate of the same phonetic word.

[0004] In order to solve such a problem, a method is conceivable in which the fact that the user has corrected the word in Kana notation is stored, and the result is reflected in the result of subsequent Kana-Kanji conversion.

[0005] For example, Japanese Patent Application Laid-Open No.
No. 0764 is known. This device includes an input unit having a non-conversion key for correcting the converted Kanji kana mixed notation into a kana notation, and a non-conversion correction registration unit for storing the operation content corrected by the non-conversion key. When the same reading is input after registration in the registration means, the conversion display is corrected and displayed in advance in accordance with the stored correction operation content.

In this type of apparatus, for example, by modifying "at" of "at school" to "at",
After that, as a conversion result of the reading "put",
It is possible to convert to "put". However, this type of device has a problem that if the corrected kana notation is a conjugation form of a word, it does not affect a different conjugation form of the word. For example, if the user modifies "wearing" to "complied" for the conversion result of "wearing clothes", the modified result is valid for the same On the other hand, the correction result is not valid, and "wear clothes".

In addition, since only the character information such as the reading and the length of the reading is used as the content of the correction operation, the kana-kanji conversion method using the grammatical information of the part of speech of the word, the word of the kana notation acquired as the correction content, There is a problem that words cannot be handled in the same way as words described in the kanji conversion dictionary.
For example, Japanese Patent Laying-Open No. 57-155668 proposes a kana-kanji conversion system using syntactic analysis, but this requires a word part of speech. However, in the conventional device, there is no means for storing the part of speech and grammatical information of the word in the kana notation, so that the above-described kana-kanji conversion method cannot be applied to the word in the kana notation, which hinders an improvement in kana-kanji conversion efficiency. Would.

However, if a user registers words in kana notation in a dictionary, complicated operations such as designation of registered words are required, which is extremely time-consuming and requires the user to determine grammatical information. Therefore, there is a problem that knowledge such as grammar is required of the user.

[0009]

As described above, in the conventional Kana-Kanji conversion device, if a word in Kana notation obtained as a result of correction is an inflected form of a certain word, it is corrected to a different inflected form of the word. There was a problem that the result was not affected. Also, since only character information such as reading and reading length is used as the content of correction operation, the Kana-Kanji conversion method that uses various conversion information such as grammatical information such as part of speech of words and semantic information, requires There is a problem that the acquired kana-notation words cannot be handled in the same way as words described in the kana-kanji conversion dictionary.

[0010] The present invention has been made in view of the above points, and an object of the present invention is to provide a kana-kanji conversion apparatus for automatically registering a word corrected to kana notation, including conversion information, in a kana-kanji conversion dictionary. To provide.

[0011]

According to the present invention, there is provided a kana-kanji conversion device for converting an input unit for inputting a reading of a Japanese sentence, a reading, a kana notation or a mixture of kanji and kana, and conversion information. Kana-Kanji conversion dictionary that is attached and pre-stored,
A kana-kanji conversion unit that converts readings entered using this kana-kanji conversion dictionary into words mixed with kanji and kana, and displays words that have been converted to kana-kanji, including words with the same reading as kana. Display unit, an instruction unit for instructing selection of one word from a plurality of displayed words having the same reading, and a word of the selected kana notation when a newly included kana notation is selected. A conversion information determination unit that determines the conversion information of the character string, and a hiragana word registration unit that registers a word in the selected kana notation with the determined conversion information in the kana-kanji conversion dictionary. is there. Further, the conversion information determination unit is for estimating the conversion information of the phrase is the same as the conversion information of the homophone, if the homophone that is the basis of the estimation is selected, A hiragana word management information updating unit for deleting a kana-notation word having conversion information estimated based on the same homophone word from the kana-kanji conversion dictionary.

[0012]

According to the present invention, when a word in kana notation is registered in the kana-kanji conversion dictionary including grammatical information, a word in kana notation is selected once to automatically convert the same phonetic word of this word. Along with the information, the kana-notation word is registered in the kana-kanji conversion dictionary. Therefore, from the next time, if the user inputs the word of the kana notation or the reading information of the inflected form of the word, the kana notation intended by the user can be converted as the first candidate. Furthermore, if a homophone in kanji notation is selected from the kana-notation words registered above and its homophone conversion information, a kana-notation word having the same conversion information as the selected homophone By deleting from the Kana-Kanji conversion dictionary, a more reliable Kana-Kanji conversion dictionary can be automatically created. Regarding the registered kana notation words, various learning methods,
Kana-Kanji conversion processing using syntactic information and co-occurrence of words can be performed.

[0013]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a schematic configuration of a kana-kanji conversion device according to the present embodiment. In FIG. 1, an input unit 1 includes a keyboard for inputting a kana character string to be subjected to a kana-kanji conversion process, or for inputting a command for editing such as homophone selection.
The kana-kanji conversion unit 2 converts the kana character string received from the input unit 1 into a kanji-kana mixed sentence using information stored in the kana-kanji conversion dictionary 9 and converts the result into a conversion result memory 6
Write to.

The grammatical information judging section 3 judges grammatical information for a selected word (hereinafter referred to as a hiragana word) among the words written in the conversion result memory 6 and selects the hiragana word. Is transmitted to the hiragana word registration unit 4. Further, the same homonym that is the basis for determining the grammatical information is stored in the hiragana word management information storage unit 10. The hiragana word registration unit 4 receives the hiragana word and the grammar information of the word from the grammar information determination unit 3,
The information is registered in the hiragana word dictionary 93.

The hiragana word management information updating section 5 receives words other than the hiragana words from the editing control section 8, and updates the hiragana word management information storage section 10. The output unit 6
A conversion result after the Kana-Kanji conversion processing, a homophone list used for correcting the conversion result, and the like are displayed. The editing control unit 8 sends the result converted by the kana-kanji conversion unit 2 to the output unit 7, receives an editing command from the input unit 1, performs an operation according to each command, and according to the editing result, the content of the conversion result memory 6. To update.

The kana-kanji conversion dictionary 9 stores the correspondence between the reading of words and the corresponding notation of kana-kana kana, and stores a general dictionary 91 prepared in advance, a new word dictionary 92 registered by the user, and hiragana words. There is a hiragana word dictionary 93 for registration.

FIG. 2 shows an example of the contents of the general dictionary 91 stored in the kana-kanji conversion dictionary 9. This dictionary is "reading" 2
1, “Heading” 22, “Grammar Information” 23, “Heading number” 24
The word “Yomi” (eg, “Haru”) is stored in the “Yomi” 22 part, and the “Heading”
In the part 22, “kana” kana mixed notation (for example, “spring”) of the word having the pronunciation of the part “Yomi” 21 is stored. The “grammar information” 23 includes grammar information of a word (for example,
"Noun") is stored. "Heading number" 24
A recognition number (for example, “1”) assigned to each heading is stored.

FIG. 3 shows an example of a storage format of the Hiragana word management information stored in the Hiragana word management information storage unit 10. One set in this format is:
It consists of “Hiragana word heading number” 31, “Homophone number” 32, and “Homophone heading number list” 33. The “hiragana word heading number” 31 stores a heading number of a hiragana word registered in the kana-kanji conversion dictionary 9. The “same word number” 32 indicates the number of heading numbers stored in the “same word heading number list” 33. "Homophone heading number list"
Consists of N items of "same word heading numbers" 321-32N. “Homophone heading numbers” 321 to 32N store heading numbers (see FIG. 2) of words having the same reading and the same grammatical information as the hiragana word indicated by “Hiragana word heading number” 31. Hiragana words that have the same reading but different grammatical information are stored in different sets with different “Hiragana word heading numbers”. FIG. 4 shows an example of hiragana word management information classified in this format.
The meaning of this information is that "Hiragana word heading number" is 2001
(Here, “Haru”, grammatical information is “Noun”), the Hiragana word is a homophone heading number 1 (“Spring”) in Hiragana notation, and the “Hiragana word heading number” is 20.
02 (here, "Haru", grammatical information is "verb-la-five-row-conjugation"), and the hiragana words are those with a homonym heading number of 2 ("stretch") or 3 ("paste"). It represents that. Therefore, for example, if the word notation “spring” that is currently being input is used, the user does not intend to write the word “haru” as “haru”, and “hiragana word heading number” is 2001. It is determined that the hiragana word is incorrectly registered. Details of this processing will be described later.

FIG. 5 shows an example of a storage format of the conversion result stored in the conversion result memory 6 for storing the result converted by the Kana-Kanji conversion unit 2. The conversion result is separated for each clause.
1, "Total number of candidates" 52, "Selected candidate number" 53, and "Homophone data" 54. “Phrase number” 51 stores the number of the phrase from the beginning of the record with this item as a result of the Kana-Kanji conversion.
The “total number of candidates” 52 holds the number of candidates of the “same word data” 54. The “selection candidate number” 53 stores the number of the candidate currently selected from the N “phrase candidate data” of the “same word data” 54. The “homophonetic data” 54 is composed of N “phrase candidate data” 541 to 54N. The numbers 1 to N at this time are called “phrase candidate numbers”. The phrase candidate data 541 to 54N includes a “kana-kanji conversion result” 55 for a kana character string and a “heading number” 56 of a word included in the conversion result. The “heading number” is the heading number of the word in the general dictionary (see FIG. 2), and is 0 for the hiragana word in the hiragana word dictionary. Next, the operation of the embodiment configured as described above will be described.

First, the overall operation will be described with reference to the flowchart of FIG. When the input unit 1 receives the kana character string to be converted in step A1, the kana character string input by the kana-kanji conversion unit 2 is converted to kana-kanji in step A2. As for the kana-kanji conversion process, for example, a method disclosed in Japanese Patent Application Laid-Open No. 60-72014 or the like may be used, and therefore will not be described in detail here. Next, in step A3, the editing control unit 8 adds a hiragana word to a phrase in which a hiragana word does not exist as a candidate, and causes the user to select a phrase candidate. In step A4, if the hiragana word added in step A3 is selected,
The grammar information determination unit 3 determines the grammar information of the Hiragana word and registers it in the Hiragana word dictionary 93 (step A).
5). If any other than the added Hiragana word is selected in step A4, the Hiragana word management information updating unit 5 rewrites the Hiragana word management information, and deletes unnecessary Hiragana words from the Hiragana word dictionary 93 based on the result. (Step A6).

Next, the operation of the editing control section 8 (step A)
3) will be described with reference to FIG. First, the conversion result memory 6
Of the first candidate is displayed based on the contents of (Step B)
1). The first candidate is the “kana-kanji conversion result” 55 of the “phrase candidate data” 541. For example, assume that the contents of the conversion result memory 6 are in the state of FIG. This state is the result of the conversion for the kana input of “Haru-ga-Kita”. In this case, the display of the first candidate refers to the selection candidate number 53 and refers to the first candidate of each phrase (the first phrase “Haruga”).
For the first phrase candidate data "Haruga" and the second phrase "Kita" for the second phrase candidate "Kita". Next, 1 is assigned to a variable BN representing a phrase number (step B2). In step B3, the presence or absence of a hiragana candidate in the phrase candidate data having a phrase number of BN is checked, and if it does not exist, it is registered in a dictionary. Then, a phrase candidate having a phrase number of BN is displayed including the above-mentioned hiragana candidates (step B4), and the user can select one of them (step B5). For example, the selection is made by the user as shown in FIG. Next, the number of the selected phrase candidate is written in the “selection candidate number” 53 of the conversion result memory 6 (step B6). Step B7
If there are no more unprocessed clauses, the process ends. If there is a clause that has not yet been processed, the variable BN
Is incremented by 1 (step B8), and the processing after step B3 is performed.

Further, step B3 will be described with reference to FIG. First, 1 is set to a variable KN representing the number of a phrase candidate.
And 1 is substituted for a variable F representing a flag for processing (step C1). It is checked whether the phrase candidate indicated by the variable KN is a kana notation, that is, a hiragana word (step C2). If it is a hiragana word, the flag F is set to 0 (step C3). Next, step C4
, The variable KN is set to “the total number of candidates” in the conversion result memory 6.
If it does not match 52, the variable KN is incremented by 1 (step C7), and the processing of steps C2 to C3 is performed until it matches. When the processing is completed for all the phrase candidates, the flag F is checked (step C5). If F is 1, that is, if there is no Hiragana word in the homophone candidate, A hiragana word is added to the end of the “phrase candidate data” 55 of the “same word data” 54. At this time, the “header number” 56 is set to 0 to indicate that it is an added word. At the same time, the “number of all candidates” 52 is increased by one. For example, if the contents of the conversion result memory 6 are as shown in FIG.
If it is (a), the hiragana word is added in step B3, and the number of the sentence candidate selected by the user in step B6 is written in the "selection candidate number" 53. Become like

The operation of the grammar information judging section 3 (step A5) will be described with reference to FIG. First, 1 is substituted for a variable BN representing a phrase number (step D1). It is checked whether the "heading number" 56 of the candidate indicated by the "selection candidate number" 53 in the "same word data" 54 having the phrase number BN is 0 (step D2). If the index number is not 0, that is, if the added hiragana word is not selected in the editing control unit 8, the process is skipped to step D7. In contrast, a word with a heading number of 0 is
Since the hiragana word to be registered in the hiragana word dictionary 93 (a part of this content is shown in FIG. 12), the grammatical information of the hiragana word is determined. Variable KN indicating the number of a phrase candidate
(Step D3). Next, the hiragana word dictionary 93 is searched to determine whether the grammar information of the candidate for the variable KN is already registered together with the hiragana word to be registered (step D4). That is, looking at the heading number 56 of the variable KN candidate, the reading 21-1 and the grammatical information 23-1 of the heading number 24 in the general dictionary 91 that matches the heading number 56 are read.
And the reading 21-2 of the Hiragana word dictionary 93 is
The grammar information 23-2 of the grammar information 23-2 that matches the -1 is subtracted. If there is a grammar information 23-1 that matches the grammar information 23-1, the registration is determined to be already registered. If not registered, in step D5, the same grammatical information 23-1 as the candidate for the variable KN is registered in the Hiragana word dictionary 93 as the Hiragana word grammatical information 23-2. At the time of registration, a heading number 24 is assigned. For example, a heading number in a general dictionary is 1 to 1000, and a heading number in a Hiragana word dictionary (this is “Hiragana word heading number” in Hiragana word management information) is 1001 to Similarly, a new heading number is assigned to the general dictionary and the hiragana word dictionary so that the heading numbers do not overlap and do not overlap with the already registered heading numbers. Then, in step D6, the heading number of the registered Hiragana word is changed to the Hiragana word management information (see FIG. 3).
1 in “Hiragana word heading number” 31
In 32, the heading number of the candidate pointed to by the variable KN is registered in “homological heading number 1” 331. If it is determined in step D4 that the entry is registered, a record having the entry word number of the hiragana word registered in the grammatical information in the “hiragana word entry number” 31 of the hiragana word management information is searched for. The heading number of the candidate pointed to by the variable KN is stored in the “homophone heading list” 33. Further, the "same word number" 32 is increased by 1 (step D9). In step D7, if the variable KN has not reached the “total number of candidates” 52, the variable KN is increased by 1 (step D7).
9) Perform the processing of step D4 and subsequent steps for the next candidate. When the processing is completed for all the phrase candidates, it is checked whether there is a phrase that has not been processed yet (step D).
7) If it still remains, the variable BN is incremented by 1, and the process from step D2 is performed on the next phrase. For example, when the contents of the conversion result memory change from FIG. 6A to FIG. 6B, the contents of FIG. 12 are written in the hiragana word dictionary 93 by the operation of the grammar information determination unit. The contents of the hiragana word management information storage unit 10 are as shown in FIG.

Next, the operation of the hiragana word management information updating section 5 (step A6) will be described with reference to FIG. First,
Substitute 1 for a variable BN representing a clause number (step E
1). If the phrase "selection candidate number" indicated by the variable BN
Homophone heading number with 51. That is, if the selected word is not a hiragana word (step E
2) Check whether this heading number exists in the “homophone heading number list” 33 of the hiragana word management information (step E).
3) If it exists, the headword number is deleted (step E4), and the “same word number” 32 is reduced by 1 (step E5). As a result, if the “same word number” 32 becomes 0 (step E7), that is, if the Hiragana word represented by the record is not a Kana notation of any Kanji notation,
The Hiragana word indicated by the “Hiragana heading number” of the record is deleted from the Hiragana word dictionary 93. And
The above operation is performed for all the phrases in the conversion result memory 2 (steps E8 and E9).

Next, an example of actual processing in this embodiment will be described. For example, it is assumed that there is a kana input such as "Haruga came." At this time, the contents of the conversion result memory 6 by the kana-kanji conversion unit 2 are as shown in FIG.
First, it is converted to "Spring has come."
Assuming that the user wants to use the hiragana notation of “coming” as “coming”, the conversion result is corrected to “spring comes” using a method as shown in FIG. The contents of the conversion result memory at this time are as shown in FIG. The grammar information determination unit 3 operates on the contents of the memory, and registers the hiragana word “Kita” in the hiragana word dictionary. The homonyms of “Kita” are “North”, “Kita”, and “Weared” from the contents of the conversion result memory 2, and therefore, the homonyms of “Kita” are “Noun”, “Verb "," One verb above verb ". The hiragana word dictionary 93 at this time is shown in FIG. 12, and the contents of the hiragana word management information storage unit 10 are as shown in FIG. Next, if there is a kana input such as “Natsu ga kuru.”, The hiragana word dictionary 93 (FIG. 1)
The hiragana word "Kuru" is registered in 2), so if you prefer to use the Hiragana word as the first candidate, it will be converted to "Summer will come." There is no need to do.

Furthermore, for the conversion "Aim for Kita.", "Aim for Kita." Is converted as a first candidate, but once corrected to "Aim for North." The contents of the information storage unit 10 (FIG. 14A) are changed to FIG. 14B by the hiragana word management information updating unit 5, and since the “same word number” of the index number 1001 is 0, the hiragana word dictionary is used. , The hiragana word "Kita" (noun) whose heading number is 1001 is deleted, and thereafter "Kita" (noun) no longer appears in the conversion result.
As described above, in the present embodiment, once a hiragana word is selected, not only that word but also its inflected form can be used in hiragana. In addition, once a misregistered Hiragana word is corrected, the word is automatically deleted from the dictionary.

It should be noted that the present invention is not limited to the above-described embodiment, but can be carried out by appropriately modifying it without changing the gist. In this embodiment, operations such as registration / deletion are frequently performed in the Hiragana word dictionary. Therefore, the Hiragana word dictionary is provided separately from the general dictionary, but this embodiment is free.
For example, a Hiragana word dictionary may be included in a general dictionary, or a Hiragana word may be registered in a new word dictionary that can be registered by a user. Further, the present invention can be implemented even if the format of these dictionaries is changed or dictionary items are added.

Further, in the above example, the word “north” (noun) is changed to “north”
If you want to write the word "Kita" (verb ka) in Hiragana as "Kita", first register both "Kita" in the Hiragana word dictionary and write the word "Kita" in "Kita". Deletes the noun “Kita” from the Hiragana word dictionary when is selected by the user. Then, when the user selects a word with the notation “Kita” in the Hiragana notation meaning “Kita”, “Kita” meaning “North” is not registered in the Hiragana word dictionary again, as shown in FIG. Step A
No. 4 is a condition that "is the hiragana word added in step A3 determined?" This may be changed to the condition of “whether a hiragana word that is not the first candidate is determined at present”.

In the above example, the timing at which a hiragana word is added to a phrase candidate is before the user selects a candidate. In this case, when the device presents the candidate of the kanji notation to the user, and the user converts the kanji notation into a hiragana notation using a non-conversion key or the like and fixes it, a hiragana word may be added.

In the above example, among the phrase candidates in the conversion result memory, only the first candidate is indicated as the selection candidate number, but the priorities (presentation order) of other phrase candidates are not given. Alternatively, the priority number 57 may be stored in the phrase candidate data in addition to the kana-kanji conversion result 55 and the index number 56 without the selection candidate number 53.

In the above example, step D14 in FIG.
When it is determined whether grammatical information of a word having the same reading as the determined Hiragana word is already registered in the Hiragana word dictionary, a search is performed by subtracting grammatical information from the reading of the Hiragana word dictionary. Using the Hiragana word management information section 10 that stores the heading number of the word that is the source of the grammar information of the Hiragana word, the same reading as the Hiragana word determined in all the homophone heading numbers 53 is used. The presence or absence of registration may be determined by examining whether the heading number of the word possessed exists.

Further, in the present embodiment, only grammatical information judgment is performed when a hiragana word is registered in a dictionary, but various information possessed by a homophone may be used as hiragana word information. For example, in a kana-kanji conversion method that uses co-occurrence information about which words are used at the same time,
If there is information that "sakura" and "blooming" are used at the same time, it is also possible to add information that "sakura" and "blooming" will be used simultaneously when registering the hiragana word "sakura" in the dictionary It is. In the present invention,
Such grammatical information, co-occurrence information, etc. are all referred to as conversion information. If the character of the word is represented by the conversion information in more detail than only the grammatical information, for example, the “Haru” and “Paste” corresponding to “Haru” classified into the same index number in FIG. You can register in the dictionary. That is, the number of homophones in the hiragana word management information is mostly one. Then, for example, the notation "Haru" for "Tension",
More detailed kanji conversion control can be performed, such as using the notation "paste" for "paste".

Further, in the present embodiment, the word registered as a hiragana word is preferentially set as the first candidate. This is done by using various learning such as long-term learning, kana-kanji conversion means using syntax information and semantic information. Thus, the first candidate can be determined. At this time, it is useful to combine the above-described modification example regarding step A4 and the modification example regarding the priority order.

[0035]

As described above, according to the present invention, once a hiragana word is selected, the conversion information including the grammatical information of the word is determined and registered in the kana-kanji conversion dictionary. Since then, the time and effort of re-selecting the word and its conjugation that you want to write in Hiragana is reduced, and the selected Hiragana word and its conjugation can be treated in the same way as other words registered in the dictionary. In addition, by deleting a hiragana word that has been erroneously registered based on the word selection status, a kana-kanji conversion device capable of improving the reliability of the registered hiragana word can be realized, which has a great practical effect. Can be done.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of a kana-kanji conversion device according to an embodiment.

FIG. 2 is a diagram showing an example of the contents of a general dictionary 91 in a kana-kanji conversion dictionary 9.

FIG. 3 is a diagram showing an example of a format of Hiragana word management information stored in a Hiragana word management information storage unit 10.

FIG. 4 is a diagram showing an example of hiragana word management information in the storage format of FIG.

FIG. 5 is a diagram illustrating an example of a storage format of a conversion result stored in a conversion result memory 6;

FIG. 6 is a view showing an example of conversion result contents in the storage format of FIG. 5;

FIG. 7 is a flowchart showing the overall processing flow of the kana-kanji conversion device;

FIG. 8 is a flowchart showing the operation of the editing control unit 8;

FIG. 9 is a screen diagram showing an example of an interface for homophone selection.

FIG. 10 is a flowchart showing a process of adding a hiragana word (step B3 in FIG. 8).

FIG. 11 is a flowchart showing the operation of the grammar information determination unit 3;

FIG. 12 is a diagram showing a part of the contents of a hiragana word dictionary 93;

FIG. 13 is a flowchart showing the operation of the Hiragana word management information updating unit 5;

FIG. 14 is a diagram illustrating the contents of hiragana word management information for conversion of “Harugarita”.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Input part 2 ... Kana-kanji conversion part 3 ... Grammar information judgment part 4 ... Hiragana word registration part 5 ... Hiragana word management information update part 6 ... Conversion result memory 7 ... Output part 8 ... Edit control part 9 ... Kana-kanji conversion dictionary 91 ... General dictionary 92 ... New word dictionary 93 ... Hiragana word dictionary 10 ... Hiragana word management information storage 21 ... Reading 22 ... Heading 23 ... Grammar information 24 ... Heading number 31 ... Hiragana word heading number 32 ... Homophone number 33 ... Homophone heading number list 51… Sentence number 52… Total number of candidates 53… Selected candidate number 54… Syllable data

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-3-137765 (JP, A) JP-A-63-276166 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06F 17/21-17/24

Claims (1)

(57) [Claims] 1. A kana-kanji conversion dictionary in which reading information, kana-kana mixed character information and conversion information are stored in advance in association with each other, and the kana-kanji conversion dictionary is used to convert input reading information to kanji-kana mixed characters. Kana-kanji conversion means for converting to information, output means for outputting character information converted by the kana-kanji conversion means, and kana-only characters in which the same reading information is written in kana only in the character information output by this output means If the information is not included, the character information is added, and a selection means for selecting one character information from the character information is selected. If the character information of only the kana added by the selection means is selected, the character information is selected. The kana-kanji conversion dictionary associates the kana-kanji character information, the reading information, and the conversion information corresponding to the character information output by the output unit with each other. Registration means for registering, and when character information other than kana-only character information is selected by the selection means, the selected character information and the kana-only character information sharing the reading information and the conversion information are ,
A kana-kanji conversion device comprising, if the kana-kanji conversion dictionary has already been registered by the registration unit, deleting the kana-only character information from the kana-kanji dictionary.