JPS61198361A - Kana-kanji converting device - Google Patents

Abstract

PURPOSE:To apply conversion again as to a character string failed in conversion before the character string of conversion object is stored in a storage area by depressing a specified key indicated for re-conversion by the operator. CONSTITUTION:A Kana(Japanese syllabary)-Kanji(Chinese character) candidate character string stored in a clause candidate storage area 27 is displayed on a CRT 5 by a clause candidate display means 3. When the displayed character is not an object character, a Kana-Kanji candidate selecting means 4 selects the next candidate character string and the objective one is found out, a clause candidate confirming means 6 confirms finally the Kanji candidate character string. When paraphrasing is failed or a prescribed candidate is not obtained through the Kana-Kanji conversion in this way, the operator uses the means 4 at first to call a Kana-Kanji candidate string one after another and uses finally the clause candidate confirming means 6 to confirm the Kana character string.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a kana-kanji conversion device that converts a Japanese sentence input in kana code into a sentence containing kana-kanji and outputs the converted sentence.

[Prior art]

In the conventional Kana-Kanji conversion device, which automatically divides and divides Japanese kana sentences input manually, analyzes multiple phrases, and determines candidates for each phrase, this is the data expansion area for analysis. Due to limitations in the number of candidates that can be selected manually, or deficiencies or limitations in the Kana-Kanji conversion analysis, there may be errors in the segmentation of clauses in the Kana-Kanji conversion, or the analysis may be terminated without outputting enough candidates. As a result, it happened that the operator was only able to obtain the conversion content that was not what he was looking for.

For example, an analysis of "peach thermos thermo thermo thermos" is performed, first "peach / peach / lee / lee / li /
Bunsetsu separation is performed as "Momomo/Momomo/Limo/Momouchi", and furthermore, the internal storage area for analysis is full and it is not possible to analyze another bunsetsu separation reverse method, so this candidate is If the process ends with only output, a real problem may arise in which the operator actually wants to convert the third and fourth clauses in the above example to limo/momo, respectively.

In this case, in the conventional technique, the entire kana-kanji conversion process is completed with the corresponding clause determined as kana, and the editing function is used to perform corrections after storing the document.

[Problem that the invention seeks to solve]

However, in the conventional technique described above, if the result of the bunsetsu segmentation processing does not reach the desired state, the character string that has been stored must be edited again.

Once a character string is stored in a storage area, editing operations such as adding letters or adding blank spaces are performed in the storage area to arrange the format of the text. Then, when the stored character string is called up again and reconverted in order to create the desired bunsetsu separation, editing operations such as filling in letters, filling out blanks, etc. are performed again in the storage area based on the result of the conversion. However, for example, filling in just one blank space can affect the entire stored text, so it can take a long time to get the desired phrase separation string. It was taking too much.

[Means for solving problems]

In the present invention, when the bunsetsu delimiter and the analysis of its internal parts fail, the bunsetsu is converted into a new character string, and after specifying the end of the word to be reconverted with a cursor etc., the operator, for example, can perform the reconversion. When a specific key to be instructed is pressed, the re-conversion instructing means recognizes this, and before the string to be converted is stored in the storage area, the string that has failed in conversion is converted again. .

〔Example〕

FIG. 1 is a diagram schematically showing an embodiment of the present invention.

First, 1 is a keyboard for inputting kana character strings and the like. A character string inputted from the keyboard 1 is sent to a kana-kanji conversion section 8, where it is converted into kana-kanji characters.

The internal configuration of the kana-kanji converter 8 is as follows. First, reference numeral 9 denotes a kana character string input means for inputting a kana character string input from the keyboard 1. 10 is a grammar analysis means for grammatically analyzing the imported kana character string. 11 is based on the analysis result of the grammar analysis means 10,
It is a means of developing independent words to extract independent words. Reference numeral 12 denotes a Japanese language dictionary search means for searching the Japanese dictionary $13 for the kanji corresponding to the independent word extracted by the independent word expansion means 11.

Reference numeral 14 denotes an adjunct word dictionary search means for searching an adjunct word dictionary 15 for an adjunct word connected to the extracted independent word.

Further, reference numeral 16 denotes a grammar analysis means 16 that indexes the converted kana-kanji candidate character strings using the phrases separated by the grammar analysis results, and further stores the indexed kana-kanji candidate character strings in the phrase candidate storage area 17. It is.

Reference numeral 2 denotes a kana-kanji conversion start instruction means for instructing the start of kana-kanji conversion for the input kana character string. Reference numeral 3 denotes phrase candidate display means for displaying a plurality of kana-kanji candidate character strings on the CRT 5. Reference numeral 4 denotes a kana-kanji candidate selection means for selecting the next candidate character string when the kana-kanji candidate character string displayed by the phrase candidate display means 3 is not the desired character string. Further, reference numeral 6 denotes phrase candidate determination means that finally determines the kana-kanji candidate character string when the desired one is found.

Further, 7 is a re-conversion instruction means for instructing the Kana-Kanji conversion again for the specified Kana character string (this point will be explained in detail later) when the segmentation of a clause is not successful.

Next, # configured as above! The operation of the embodiment shown in FIG. 1 will be schematically explained.

First, a kana character string inputted from the keyboard 1 is converted into kana-kanji by the kana-kanji converter 8 according to an instruction from the kana-kanji conversion start means 2. The detailed description of the process is as follows.

First, a kana character string is input by the kana character string input means 9. Next, the imported kana character string is analyzed by the grammar analysis means 10, and based on the analysis result, an independent word is extracted by the independent word expansion means 11, and the kanji corresponding to the extracted independent word is extracted. The character string is obtained by the Japanese language dictionary search means 12 searching the Japanese language dictionary 13. On the other hand, the adjunct word dictionary 15 is searched by the adjunct word dictionary search means 14 to obtain an adjunct word that is well connected to the extracted independent word.

Then, the independent words converted into Kanji character strings and the adjunct words connected thereto are stored in the phrase candidate storage area 17 by the phrase indexing means 16 using the individual phrases as indexes. (Figure 2 shows how Kana-Kanji candidate character strings are stored in the phrase candidate storage area 17.)The Kana-Kanji candidate character strings stored in the phrase candidate storage area 1F are , are displayed on the ORT 5 by the phrase candidate display means 5. When the displayed item is not the desired one, the next candidate character string is selected by the kana-kanji candidate selection means 4, and when the desired one is found, the phrase candidate confirmation means 6 finally selects that kana-kanji candidate character string. Confirm.

Incidentally, even if the kana-kanji conversion is performed as described above, there are cases in which the segmentation fails or a predetermined candidate cannot be obtained. In such a case, the operator uses the kana kanji candidate selection means 4 to call up kana kanji candidate character strings one after another, and finally uses the bunsetsu candidate determination means 6 to determine the kana character string.

In other words, the place where bunsetsu separation fails becomes a kana character string. After that, the re-conversion instruction means 7 specifies that the kana character string is to be re-converted from the beginning.

Also, the beginning of the area to be reconverted is the beginning of the area that has been temporarily replaced with a character string as described above, but the end is, for example, the end that is intended for the cursor (the temporary character string shown above). The user can also move up to a certain key (
It may be a key that instructs the start of kana-kanji conversion. ) is pressed, and the kana-kanji conversion start instruction means 2 recognizes this and is confirmed.

The above operation will be explained in more detail using FIGS. 2 and 5.

First, in the phrase candidate storage area 17, candidate kana-kanji character strings are stored as shown in FIG.

That is, clause indexes 101, 102, 103.
. . . is a pointer indicating the start address where a kana-kanji candidate character string corresponding to one clause starting from the corresponding kana digit position is stored. And 201.
202. .......

210 is a group of kana-kanji candidate character strings each corresponding to one clause. That is, for example, the clause inch, Kus101
is the address where the first of the kana-kanji candidate character string group 201 corresponding to a clause starting from the first digit "mo" of the character string "momomomomomomomomo" is stored. It shows.

Then, in the area specified by the address, the first
The second kana-kanji candidate character string is stored, for example, in the form of a character code, and furthermore, an address value indicating the area where the next second kana-kanji candidate character string is stored is stored.

In other words, each kana-kanji candidate character string 201α, 201b, ・
. . . are connected with one pointer indicating the next storage area. (~ in Figures 2 and 3 are
It means omitted. ) By the way, Figure 2 shows a case where the character string ``Momomomomosumomomomomomomomo'' is converted and the result is ``Momomomomomomomomomomomomo''. (Kana-kanji candidate character strings 201α, 204α, 206α, 207α, 208α
, 209α.

) However, suppose that you actually wanted to convert it to ``Momo, Tori, Momo, Momo.''

In this case, first, the kana-kanji candidate selection means 4 and the bunsetsu candidate determination means 6 are operated according to the operator's instructions, and the third.
4th. The fifth clause is tentatively converted to a kana string. (Kana-Kanji candidate character strings 206b, 207 in Figure 2)
See H, 2084. ) Thereafter, when the operator instructs reconversion by pressing a specific key, for example, the reconversion instruction means 7 recognizes this, and the third. 4th. The mode is set to convert the character string of the fifth clause into kana-kanji again. Here, as mentioned earlier, the end of the digit position to be reconverted is determined by the operator, for example, moving the cursor and pressing a specific key to instruct kana-kanji conversion when the cursor reaches the desired location. 2 is confirmed by recognition.

Figure 3 shows 3. 4th. jl! s clause is converted into the desired character string (kana-kanji conversion candidate character string 220α, 2
21α), but as is clear from the figure, when reconverted, the bunsetsu index 106,1
10 and the kana-kanji candidate character string groups 220 and 221 are also rewritten. As a result, the character string ``Momo, Tori, Momo, Momo'' is obtained as originally desired.

Next, the operation of the embodiment shown in FIG. M1 will be explained using the flowchart shown in FIG.

First, Kana characters and Kanji are input from keyboard 1 (<1
>). Next, when the operator instructs kana-kanji conversion for the above-mentioned kana character string by pressing a specific key, the kana-kanji conversion start instruction means 1 recognizes this, and the kana-kanji conversion begins (<2>). .

Next, the key force is checked (<3>), and the following various processes are performed depending on the type of input key. (4).

<4-eL> The operator looks at the kana-kanji candidate character string that corresponds to a phrase,
When you want to see the previously displayed Kana-Kanji candidate character string again, for example, you press a specific key, the Kana-Kanji candidate selection means 4 recognizes this, and the previously displayed Kana-Kanji candidate character string is OR'l''. 5.

<4-b> Basically the same as <4-a), but in this case,
The operator indicates the next kana-kanji candidate character string, and it is displayed on 1ff (!RT5).

4-C> When the operator moves the cursor to the left, the phrase to be selected for the kana-kanji candidate character string becomes the previous phrase.

<4-cl> When the operator moves the cursor to the right, the next clause becomes the target of the kana-kanji candidate character string selection action. (ku4-α>s<a-b>t(a-c),
In either case, after the processing of (a - d) is completed, 3
This will be the process. ) <4-=> When the operator wants to re-convert the part where the bunsetsu segmentation has failed, when a specific key is pressed, the re-conversion instruction means 7 recognizes this, starts re-conversion, and first makes corrections. It is checked whether the desired phrase has been converted to a kana phrase (<40>). Here, if it is already a kana clause, the process goes to step 5>, and after the range is specified with a cursor etc., the kana-kanji conversion start instruction means 2 recognizes the operator's instruction to start conversion and re-converts. The character string to be converted is determined and reconversion begins. After that, the corresponding Kana-Kanji candidate character string groups 201, 202, 205, -
... are rewritten, and the phrase indexes 101, 102, ... of the corresponding portions are also rewritten (<6>), and then the process returns to <3>. Also, if NO in <40>, error processing (<41>)
Return to 5><4-/> When the operator wants to confirm the character string of the phrase to be converted, the operator presses a specific key, for example. Then, the clause candidate determining means 6 recognizes this, stores the determined character string in an output area (not shown) (<7>), and ends the process.

<4-y> If the operator wishes to cancel the Kana-Kanji conversion of the entire input character string, the operator presses a specific key. and,
After this, the kana-kanji conversion for the entire input kana character string becomes invalid, and the process returns to step 1>.

Further, in the embodiment shown in FIG. 1, a CRT5 is used as the display device, but a liquid crystal display device, a plasma display, a KL, etc. may be used instead of the ORT5. Further, in the embodiment shown in FIG. 1, the keyboard 1 is used as a means for inputting a kana character string, but it is also possible to input a voice input device, a handwritten character input device, a light pen, a mouse, etc.

Further, the present invention can be applied to a so-called word processor having a document editing/registration function, but can also be applied to, for example, an operating system having a kana-kanji conversion function.

〔Effect of the invention〕

As explained above, the present invention has the following effects.

(1) If the operator makes a mistake in the bunsetsu break, the short section can be converted again, so corrections can be made at the time of Kana-Kanji conversion, and the correction will not affect the entire stored sentence. You can speed up your work.

(2) By using this function, the number of candidates in the initial conversion can be limited to some extent, so the work area for kana-kanji conversion can be saved and the initial candidate display can be speeded up.

(8) When converting again, use a technical term dictionary, a kanji one-character selection dictionary, a person's name,
The first conversion is to change the dictionary to be processed, such as a place name dictionary, or to use it as a secondary storage dictionary.
− processing, and short-range reconversion can be processed using a medium with slow access speed.

[Brief explanation of drawings]

FIG. 1 is a schematic diagram showing an embodiment of the kana-kanji conversion device of the present invention. FIG. 2 is a diagram showing the relationship between the phrase index and the phrase candidate after the first conversion in the example problem in the sentence. FIG. 3 is a diagram showing the relationship between the phrase index and the phrase candidate after the second conversion in the example problem in the sentence. FIG. 4 is a block flowchart of the processing procedure of the present invention. 2...Kana-kanji conversion start instruction means 7...
・More than reconversion instruction means

Claims (1)

[Claims] A table indicating the location where the first candidate character string of a group of Kana-Kanji candidate character strings containing a Kana character string whose starting digit is the corresponding digit position is stored, and Each of the candidate character strings is instructed to store the storage location one after another, starting from the first one, and furthermore, the operator's instructions are recognized for the clauses that failed in kana-kanji conversion and were converted to kana character strings. A kana-kanji conversion device comprising a re-conversion instructing means for instructing kana-kanji conversion in units of clauses again, and re-converting portions where conversion failed before storing the character string to be converted into kana-kanji. .