JP3768577B2 - Character processing method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a character processing method, and more particularly to a character processing method for generating a kana-kanji mixed character string by kana-kanji conversion from a kana character string input to a document processing apparatus.
[0002]
[Prior art]
A kana-kanji conversion processing method of the conventional document processing apparatus will be described with reference to the flowchart of FIG.
[0003]
First, in the document processing apparatus, a kana character string (reading) is input (step s1), kana-kanji conversion is performed on the kana character string, and conversion candidates are displayed and output. More specifically, a kana-kanji dictionary is referred to, and a notation corresponding to the text is extracted as a conversion candidate (step s2). If the conversion candidate is not the result desired by the user (step s3), the user operates the keyboard to change the input character string delimiter and instruct reconversion, or other homophones The conversion candidate is corrected depending on whether the candidate is designated (step s4). When the word (reading and notation) created by the correction process of s4 does not exist in the dictionary (step s5), this word is stored in the learning data storage area by the automatic registration process (step, s6). At the same time, a word learning process is performed for the usage frequency so that the word is reflected in the next conversion process (step s7).
[0004]
A specific example of Kana-Kanji conversion processing will be described with reference to FIGS.
[0005]
In FIG. 17, the input character string [Daijinsho] is converted, and [Ministerial Chapter] is output as a conversion candidate. However, since the conversion candidate is different from the desired result [Minister Award], by changing and selecting the candidate, [Chapter] is changed to [Award] and [Prize] is learned by words.
[0006]
When the above operations are executed, the next time [Daijinjo] is entered, conversion candidates for [Minister Award] can be obtained.
[0007]
Next, as shown in FIG. 18, when an input character string [deta-sho] is given, [data award] is output as a conversion candidate because the use frequency of [award] is high. However, since the conversion candidate is different from the desired result [data show], the user must convert [prize] to [show] by selecting the candidate. Since the changed [Show] is a word that does not exist in the dictionary, it is automatically registered as an automatically registered word in the learning data. In addition, learn words automatically registered words.
[0008]
Next, when [Daijinjo] is input again and converted as shown in FIG. 19, [Show] exists as an automatically registered word, and the word is learned at the time of the previous conversion. Show].
[0009]
[Problems to be solved by the invention]
Thus, when a word that was not originally in the dictionary in the conventional character processing method is registered in the dictionary, it takes precedence over the word in the normal dictionary, and as a result, a desired conversion candidate cannot be obtained immediately, Conversion rate may drop.
[0010]
In view of the above, an object of the present invention is to provide a character processing method capable of performing a suitable conversion process for automatically registered words.
[0011]
[Means for Solving the Problems]
In order to solve such an object, the invention of claim 1 includes a dictionary storage means for storing word readiness , notation, and frequency in association with each other, and an input means for inputting the word readiness. The dictionary storage means is referred to, and a high-frequency notation is preferentially used as a conversion candidate for the word read from the input means, and a word not yet registered in the dictionary storage means when it is created as a conversion result of the Gayomi, the character processing method for the possible document processing device to automatically register denoted word reading in the conversion result to the dictionary storage means, the document processing device , A first determination means, a second determination means, a learning means, and an adjustment means, and when the word input from the input means is converted into a notation, the conversion target word is automatically registered in the dictionary storage means Is A first determination step of determining by said first determining means for determining words that are words or pre-registered, if it is determined that the word is automatically registered by the first determination step, is stored in the dictionary storage unit A second determination step in which the second determination means determines whether or not the frequency of the word has reached a predetermined value, and when it is determined in the first determination step that it is registered in the dictionary in advance, or When it is determined by the second determination step that the predetermined value has been reached, the learning unit performs word learning for storing the conversion history in the history storage unit in order to reflect the conversion result in the next conversion. and learning step, for the word, which is the automatic registration stored in the dictionary storage means when it is determined not to have reached the predetermined value by the second determination step The frequency is increased by only 1 wherein adjusting means, by said learning step of 1 only the adjusting means for the word that has been automatically registered frequencies stored in the dictionary storage unit having the same reading as the words and word learning after completion And an adjusting step for decreasing.
[0012]
The invention according to claim 2 is characterized in that the dictionary storage means has an area for storing automatically registered words and a kanji conversion normal dictionary area.
[0013]
A third aspect of the present invention includes a dictionary storage means for storing in association with each title and frequency and word reading, and input means for inputting a word of readings, with reference to the dictionary document storage means When the word input from the input unit is preferentially used as a conversion candidate with high frequency notation, and a notation of a word not registered in the dictionary storage unit is created as a conversion result In addition, in the character processing method for a document processing apparatus capable of automatically registering the word reading and notation in the conversion result in the dictionary storage means , the document processing apparatus includes a determination means, a first adjustment means, and a second adjustment means. The method further comprises adjusting means, and when the word input from the input means is converted into notation, the determination is made as to whether the word to be converted is a word automatically registered in the dictionary storage means or a word registered in advance. hand A first step of increasing the frequency stored in the dictionary storage means by 1 by the first adjusting means for the word to be converted when it is determined that the word is automatically registered by the determining step. When it is determined that the word is registered in the dictionary in advance in the adjustment step, the same word as the word is learned by storing the conversion history in the history storage means in order to reflect the conversion result in the next conversion. And a second adjustment step for reducing the frequency stored in the dictionary storage means by 1 by the second adjustment means for the automatically registered words having the following meaning.
[0014]
The invention of claim 4 is characterized in that the dictionary storage means has an area for storing automatically registered words and a kanji conversion normal dictionary area.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[0016]
FIG. 1 shows a system configuration of a document processing apparatus to which the present invention is applied. In FIG. 1, a character string input from an input device (such as a keyboard) 1 is processed by a central processing unit (such as a CPU) 2 and output to an output device (such as a CRT) 3. The conversion history (word learning or the like) is stored in the storage device (hard disk or the like) 4 and the change history is reflected in the conversion result at the time of the next analysis. The storage device 4 is provided with a kana-kanji conversion normal dictionary 4A and an automatically registered word storage area 4B constituting a dictionary.
[0017]
FIG. 2 shows a processing procedure related to the character processing of the CPU 2.
[0018]
The CPU 2 processes the character string input from the input device 1 by Kana-Kanji conversion processing and outputs conversion candidates (steps s101 → s102). The CPU 2 determines whether or not the conversion candidate is desired based on the presence / absence of an instruction to determine the conversion candidate (step s103). If it is the desired conversion result, it is determined whether or not the word of the conversion result is automatically registered (step s107). If it is not the desired homophone, the conversion candidate that has not been subjected to the candidate correction process is corrected by the user's instruction and finally confirmed. It is determined whether or not the conversion candidate after correction exists in the dictionary (step s105). If the word does not exist, a new automatic registration process (step s106) is performed.
[0019]
If the confirmed candidate is a homophone registered in the dictionary, it is determined whether or not the conversion result word is automatically registered (step s107). If it is not automatic registration, word learning processing (step s109) is performed. When the converted word is automatically registered, the frequency of the automatically registered word is checked. If the frequency is sufficiently high, that is, if the predetermined number of times of use is exceeded, word learning processing (step s109) is performed. Further, in the case of an automatically registered word, an automatic registered word frequency adjustment process is performed (step s10).
[0020]
Specific processing contents when the processing procedure of FIG. 2 is executed will be described with reference to FIGS.
[0021]
In FIG. 3, when the user inputs “deta-sho”, the document processing apparatus performs kana-kanji conversion based on the dictionary 4A to obtain a conversion candidate “data certificate”. When this conversion candidate is not the conversion result desired by the user, the user operates the keyboard to convert it to the desired conversion candidate [data show].
[0022]
As shown in FIG. 4, when the selected [Show] is a word that does not exist in the dictionary A4, the document processing apparatus performs automatic registration processing.
[0023]
When the user later inputs [deta sho] again, the document processing apparatus performs kana-kanji conversion based on the dictionary 4A to obtain a conversion candidate [data certificate]. Since it is not the desired conversion result for the user, the user converts it into a desired conversion candidate [data show] (see FIG. 5).
[0024]
Since [Show] selected by candidate selection is an automatically registered word, the document processing apparatus looks at the frequency of the automatically registered word, and if the frequency is sufficiently high (in the present embodiment, the determination value is set to 4 or more). ) Perform word learning processing. In this case, since the frequency is 0, the word learning process is not performed, and the frequency of the automatically registered word is increased from 0 to 1 by the automatic registration frequency adjustment process (see FIG. 6).
[0025]
Every time the user selects / determines this word, the frequency of [Show] increases, and when it becomes 4 or more, [Show] is output as the first conversion result as shown in FIG. .
[0026]
As shown in FIG. 8, when the user inputs [Daijinsho], performs Kana-Kanji conversion and obtains a conversion candidate called [Minister Show], the user desires when the conversion candidate is not the conversion result desired by the user. Change the conversion result [Minister Award] by selecting a candidate. In addition, [word] is learned as follows. That is, since the user changed and confirmed the automatic registration [show] that became the first candidate to another candidate, the document processing apparatus reduces the use frequency of [show] by adjusting the automatic registration frequency (see FIG. 9).
[0027]
Suppose that the user inputs [deta-sho] as shown in FIG. 10, performs kana-kanji conversion, and obtains a conversion candidate called [data award]. Since this conversion candidate is not the desired conversion candidate, the user corrects [prize] to [show]. Since [Show], which is automatic registration, has been selected, the document processing apparatus determines whether the frequency of this word is sufficiently high. Since this word has a frequency of 3 and is smaller than the judgment value 4, the word learning process is not performed, and the frequency is increased by automatic registration frequency adjustment (frequency 4, see FIG. 11).
[0028]
Here, as shown in FIG. 12, when the user inputs [Daijinjo] again, [Show] is not learned, and the word learning of [Prize] is effective, so the conversion result is converted to [Minister Award]. Is done. When the user again inputs [deta-sho] as shown in FIG. 13, the word learning is also applied, so that it is converted to [data award]. However, since this candidate is not a candidate desired by the user, the user converts [prize] to [show] by candidate selection processing. Since the [Show] selected at this time is an automatically registered word, the document processing apparatus checks whether the frequency is sufficiently high. As shown in FIG. 14, since the frequency is 4 and the determination value 4 is reached, the document processing apparatus treats this automatically registered word in the same way as an ordinary dictionary registered word, performs word learning processing, and further increases the frequency.
[0029]
(Other examples)
In the above-described embodiment, when the conversion candidate determined by the user is an automatically registered word, it is determined whether or not to learn the word according to the frequency. However, it is possible to perform word learning for words described in the dictionary 4A and not perform word learning for automatically registered words. The character processing procedure for this is shown in FIG.
[0030]
In FIG. 15, the same processes as those in FIG. Steps s200 to s210 in FIG. 15 show a procedure for processing for detecting automatically registered words and not learning words.
[0031]
According to this processing procedure, when the user inputs [deta-sho] and performs kana-kanji conversion, a conversion candidate [data certificate] is obtained (steps s101 → s102 in FIG. 15). Here, since it is not the conversion candidate desired by the user, the conversion candidate “proof” is corrected to “show” by operating the keyboard device. A desired conversion candidate “data show” is acquired, and the conversion candidate is instructed (see step s104 in FIG. 15 and FIG. 3).
[0032]
The document processing apparatus determines whether the confirmed conversion candidate “show” exists in the dictionary 4A or the automatic registration area 4B (step s105 in FIG. 15).
[0033]
When this “show” does not exist in the dictionary 4A and the automatic registration area 4B, the read “sho” and the notation “show” are registered as automatic registration words, and the frequency is set to 0 (step s106 in FIG. 15).
[0034]
When the user newly inputs “detasho”, the document processing apparatus executes kana-kanji conversion.
[0035]
Since the above-mentioned “show” has not been word-learned, “data certificate” is obtained as a kana-kanji conversion candidate as in the previous case (steps s101 → s102 in FIG. 15).
[0036]
The user corrects “certification” in the conversion candidates to “show” and determines “data show” (step s104 in FIG. 15). In the document processing apparatus, the determination in step s105, that is, the determination whether the conversion candidate “show” exists in the dictionary is executed. Since “Show” is automatically registered at this time, a positive determination is obtained. Subsequently, since an affirmative determination is also obtained in the determination in step s200 in FIG. 15, the word learning in step S201 is not performed, and the frequency related to “show” is adjusted from “0” to “1” (step s202 in FIG. 15). .
[0037]
In this way, since the frequency of the automatically registered words increases every time the user uses the words, the display order is earlier than other words of the same sound. Moreover, since the word learning is not performed about the automatically registered word, the display order of the word learned in the normal kana-kanji conversion dictionary 4A is earlier.
[0038]
In this way, after using the automatically registered word, when the word described in the Kana-Kanji conversion dictionary 4A having the same sound as “Show” is selected by the user, the execution procedure of FIG. The route is followed, and after the word learning of the word, the frequency of the word “show” is decremented by 1 (step s202 in FIG. 15).
[0039]
【The invention's effect】
As described above, in the present invention, the automatic registration word is distinguished from the word previously described in the dictionary, and the automatic registration word that is not used much lowers the rank as a conversion candidate. This eliminates the negative effects of registering auto-registered words. Further, the conversion rate is improved, which can contribute to the improvement of the operability of the document processing apparatus.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a system configuration of a document processing apparatus to which the present invention is applied.
FIG. 2 is a flowchart showing a processing procedure of the document processing apparatus.
FIG. 3 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 4 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 5 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 6 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 7 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 8 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 9 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 10 is an explanatory diagram illustrating processing contents of the document processing apparatus.
FIG. 11 is an explanatory diagram illustrating processing contents of the document processing apparatus.
FIG. 12 is an explanatory diagram illustrating processing contents of the document processing apparatus.
FIG. 13 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 14 is an explanatory diagram showing processing contents of the document processing apparatus.
FIG. 15 is a flowchart illustrating a processing procedure according to another embodiment.
FIG. 16 is a flowchart showing a conventional processing procedure.
FIG. 17 is an explanatory diagram showing conventional processing contents;
FIG. 18 is an explanatory diagram showing conventional processing content.
FIG. 19 is an explanatory diagram showing conventional processing content.
[Explanation of symbols]
1 Input device 2 CPU
3 Output device 4 Storage device 4A Kana-Kanji conversion dictionary 4B Automatic registration word area

Claims (4)

Includes a dictionary storage means for storing in association with reading and notation and word frequency, and input means for inputting a word reading,
With reference to the dictionary document storage means, as well as with priority frequent representation against reading of the input word conversion candidates from said input means, notation unregistered word to the dictionary storage means for reading In a character processing method for a document processing apparatus capable of automatically registering the word reading and notation in the conversion result in the dictionary storage means when created as a conversion result,
The document processing apparatus further includes first determination means, second determination means, learning means, and adjustment means,
When the word input from the input unit is converted from reading to notation, the first determination unit determines whether the conversion target word is a word automatically registered in the dictionary storage unit or a word registered in advance . 1 determination step;
The second determination means determines whether the frequency of the word stored in the dictionary storage means has reached a predetermined value when it is determined that the word is automatically registered in the first determination step . 2 determination steps;
When it is determined in the first determination step that it is registered in the dictionary in advance, or when it is determined that the predetermined value is reached in the second determination step, the conversion result is used for the next conversion. A learning step in which the learning means performs word learning for storing the conversion history in the history storage means for reflection ;
When the second determination step determines that the predetermined value has not been reached, the adjustment unit increases the frequency stored in the dictionary storage unit by 1 for the automatically registered word , and the learning step A character processing comprising: an adjusting step for reducing the frequency stored in the dictionary storage means by 1 for the automatically registered words having the same reading as the word learned after completion by the adjusting means. Method. 2. The character processing method according to claim 1, wherein the dictionary storage means includes an area for storing automatically registered words and a kanji conversion normal dictionary area. Includes a dictionary storage means for storing in association with reading and notation and word frequency, and input means for inputting a word reading,
With reference to the dictionary document storage means, as well as with priority frequent representation against reading of the input word conversion candidates from said input means, notation unregistered word to the dictionary storage means for reading In a character processing method for a document processing apparatus capable of automatically registering the word reading and notation in the conversion result in the dictionary storage means when created as a conversion result,
The document processing apparatus further includes a determination unit, a first adjustment unit, and a second adjustment unit,
A determination step of determining, by the determination means, whether the word to be converted is a word automatically registered in the dictionary storage means or a word registered in advance when the word input from the input means is converted into a notation; ,
A first adjustment step for increasing the frequency stored in the dictionary storage means by 1 by the first adjustment means when the word is automatically registered by the determination step ;
If it is determined that the word is registered in the dictionary in advance, the word learning for storing the conversion history in the history storage means is performed to reflect the conversion result in the next conversion, and the word has the same reading as the word. character processing method characterized in that comprises a second adjusting step of decreasing the above for words that are automatically registered by one the frequency stored in said dictionary storage means a second adjusting means. 4. The character processing method according to claim 3, wherein the dictionary storage means has an area for storing automatically registered words and a kanji conversion normal dictionary area.

Images