JP2595934B2 - Kana-Kanji conversion processor - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a kana-kanji conversion processing apparatus, and more particularly to a kana-kanji conversion result that is output when an operator changes the unit of conversion for the output kana-kanji conversion processing result. The present invention relates to a kana-kanji conversion processing device for learning results.

[Background of the Invention]

In a conventional general kana-kanji conversion processing device, when performing kana-kanji conversion processing for each unit of a single kanji, a word, and a phrase, a delimiter of each unit is inserted using a conversion key or the like. When selecting the homonymous notation word, the next candidate key was used, and when reading the kana as it was, the no-conversion key was only used, and the input could be made almost in the original text.

On the other hand, recently, continuous phrase input or solid sentence input has become widespread, but in the kana-kanji conversion process in this case, the input reading sequence is divided into words or phrases and displayed.

The problem here is that when the kana-kanji conversion processor (hereinafter, also simply referred to as “device”) divides the input reading sequence into words or phrases with the wrong reading length, According to the method described above, the error, that is, it is necessary to correct the division break of the reading arrangement (hereinafter, this is referred to as "change of separation").

Conventionally, when performing the above-described delimiter change, it is common to change the delimiter using a delimiter display and a delimiter position designation cursor or the like, and then select and correct homophones (homophonetic words).

In this case, conventionally, since no learning is performed on the delimiter change operation, a similar description appears many times in the sentence, and if it is a conversion display requiring the delimiter change, the operator Had a very troublesome input correction operation.

[Object of the invention]

The present invention has been made in view of the above circumstances, and an object of the present invention is to solve the above-described problems in the conventional technology, and to eliminate the intention of the operator even if the operator does not perform the delimiter change operation each time. An object of the present invention is to provide an efficient kana-kanji conversion processing device that makes it easy to output a converted result.

[Summary of the Invention]

An object of the present invention is to provide a kana-kanji conversion processing device having kana-kanji conversion control means for converting a character string input by phonetic characters into a sentence mixed with kanji and kana by searching a word dictionary. When the operator modifies the unit division performed by the apparatus using the delimiter change key, at least the reading length before the change, A word delimiter change registering means for storing the connection information with the previous word and the type of change is provided, and the kana-kanji conversion control means is configured to control the display divided into word units as the conversion result by the device. When a change operation is performed by the operator using the delimiter change key, after the change is registered in the registration unit, when the same reading is input next time, the registration unit It is achieved by the kana-kanji conversion processing device and controls to display the conversion results in accordance with the stored contents.

〔Example〕

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a processing flowchart for explaining the outline of the operation of the kana-kanji conversion processing apparatus according to one embodiment of the present invention. The outline of the operation of the kana-kanji conversion processing apparatus according to the present embodiment will be described below with reference to FIG.

Upon completion of the initial processing (step 101), the apparatus displays an input request on the CRT (step 102), and if there is a key input by the operator (step 103), whether the input is character input or other input is performed. Check (step 104). If the latter,
It is checked whether it is a conversion display correction operation or other (for example, a word processor function or another conversion function) (steps 110 and 112). If the operation is an end operation or another operation, the corresponding process is performed. If the operation is a correction operation, a correction process (step 114) is performed. If there is a break change,
In the word break change registration process (step 116), the contents of the operation are stored in the main memory in the form of a table (word break change registration table) 12.

If it is an input character string, preprocessing of the character string is performed (step 105), and a word candidate string is extracted with reference to the word dictionary 10 and the grammar dictionary 11 (step 106). In the non-breaking change candidate priority processing (step 107), the previously registered word break change registration table 12 is referred to, and if there is a corresponding one,
The first conversion candidate is changed to the candidate, and the next conversion character string editing process (step 108) is performed to output as a conversion candidate (step 109). Note that the “candidate not changed by the break” means a candidate different from the display candidate before the change of the break at the time of registration.

As described above, according to the kana-kanji conversion processing device of the present invention, by registering and learning the operator's delimiter change operation, the result can be preferentially used to speed up the processing.

FIG. 2 is an explanatory diagram showing an example of the correction processing (step 114) in FIG. Here, when the "reading" is input for the "input original" and the "conversion sentence" is indicated, the operator exemplifies the notation different from the original by using a delimiter change key and a homophone selection key (conversion key). Modify as has been done. The correction operation is, for example, “Fukusha” → “Fuku / sha” → “Secondary”.
→ "Shachosa" → "President, Sa" ...
"Sa" + "ma" → "sama" → "sama" ... Correct it as in the example of lengthening.

FIG. 3 is a diagram showing the contents of the above-described word segmentation change registration table 12 when the correction operation as shown in FIG. 2 is performed. In the third example, the word break change registration table 12
(A) the length of the reading before the change, (b) the length of the reading after the change, (c) the type of the change, (d) the character string of the changed part, I remember the part of speech of words. Here, as the change type, “1” is registered when the original word is lengthened, and “2” is registered when the original word is shortened. In the change type 1, the reading length after the change is the reading length of the combined word (for example, “ma”), and in the change type 2, the reading length of the first divided word is changed. (Example: "Fuku").

The part of speech of the previous word may be the part of speech of the word before the word, or may be connection information with the previous word. In other words, the “part of speech” or “part of speech” here may be the “part of speech” or “part of speech” in the commonly used national grammar, or may be further finely or roughly classified or expanded. Incidentally, 241,227 exemplified here
Are the parts of speech of the previous word, 241 is a personal name, and 227 is a kind of noun.

FIG. 4 is a configuration diagram of a kana-kanji conversion processing device according to the present embodiment.

In the figure, 1 is an input preprocessing unit, 2 is a candidate extraction unit, 3 is a conversion information management unit, 4 is an output control unit, and 5 is a conversion control unit.
When reading is input, the input pre-processing unit 1 separates a solid sentence and classifies it into kanji and adjunct words to reduce unnecessary processing before converting from a word dictionary to kanji. Search for kana sentence appearing in and remove it from the conversion target. Next, in the candidate extraction unit 2, the conversion control unit 5
Under the control of, the corresponding word is converted into a kanji, and the converted information or the excluded information is transferred to the conversion information management unit 3 and stored and managed. After editing by the conversion control unit 5,
After outputting to the output control unit 4 and performing a return process such as performing backtracking, a sentence mixed with kanji and kana is output.

FIG. 5 is a perspective view of the device shown in FIG. As shown in the figure, the apparatus is divided into a display section and a keyboard section, and the keyboard section has function keys, character keys, conversion control keys, numeric keys, correction keys, candidate selection keys, delimiter change keys, and the like. The display unit displays a text area, a conversion display line, a correction line, a reading input display line, and the like.

FIG. 6 is a diagram showing a detailed configuration example of the candidate extracting unit 2 of FIG. The candidate extraction unit 2 includes a candidate extraction control unit 6, a candidate word string storage unit 7, a dictionary search unit 8, and a conversion dictionary storage unit 9.

When the kana character string signal is input to the candidate extraction control unit 6,
The candidate extraction control unit 6 searches the conversion dictionary storage unit 9 from the head of the conversion dictionary storage unit 9 and extracts all conversion candidates corresponding to the reading. If there is no conversion candidate, the dictionary search unit 8 notifies the candidate extraction control unit 6 of the fact. The candidate word strings are stored in the candidate word string storage unit 77.

FIG. 7 is a processing flowchart of a registration method at the time of changing a word break in this embodiment. Based on FIG.
A registration method at the time of changing a word break will be described.

When there is a correction request from the operator, it is first determined whether or not the registration condition is satisfied (step 201). If the condition is satisfied, the registration position of the registration table is set first (step 202). ), And register the reading length before changing the break (step 203). Second, the length of the reading after the change of the break is registered (step 204), the type of the change of the break is registered third (step 205), and the reading character string of the change unit is registered fourth. (Step 206) Finally, the part of speech of the word located before the change unit is registered (Step 20).
7).

The registration condition in step 201 is considered to be a candidate for a word delimiter that is a candidate by referring to the dictionary in the correction mode and the normal kana input. In the case of the correction operation, as in the example of “sha”, if the substring of the word remaining in the previous delimiter change correction operation is delimited again, it is not registered.

For setting the registration position in step 202, for example, a fixed-length table is set as shown in FIG. 3 and registered in the order of serial numbers. When the table becomes full, it is overwritten and registered again from the beginning of the serial number. The reference operation is performed from the latest registered position, and the reference is performed in the reverse order of the serial numbers (ring method).

FIG. 8 is a processing flowchart of the non-separation change candidate priority processing shown in step 107 of FIG. Referring to FIG. 8, the operation of the non-separation change candidate priority processing will be described.

In the non-separation change candidate priority processing, first, after extracting a word candidate string, it is determined whether or not the reference condition of the registration table is satisfied (step 301). If the condition is satisfied, the registration table is searched. (Step 302). Then, it is determined whether or not there is a corresponding one (step 303). If there is a corresponding one, the first candidate of the conversion display is corrected to the corresponding candidate searched above (step 304). Step 3
It is assumed that the reference condition in 01 is the input mode and the normal kana-kanji conversion.

FIG. 9 is a processing flowchart for searching the registration table shown in step 302 of FIG. The operation of the registration table search process will be described with reference to FIG.

When searching the registration table by the non-separation change candidate priority processing, first, the reference position is set to the latest registered position (step 401), and it is determined whether or not the part of speech of the previous word matches (step 402). ). If they do not match, it is determined whether or not there is something to refer to (step 406). In step 407, the reference position is set to the immediately preceding registered latest position.

If the part of speech of the previous word matches, then it is determined whether or not the reading matches the condition at the time of registration (step 403). If they do not match, it is determined whether there is something else to be referred to (step 406), and the reference position is set to the previous registered latest position (step 407). If they match, it is determined whether or not the candidate is suitable as a non-change candidate (step 404), and if so, the position of the candidate is set (step 405). If the candidate is not appropriate as a candidate other than a change in partition, it is determined whether or not there is another reference (step 406).

FIG. 10 is a processing flowchart for judging whether or not the reading shown in step 403 of FIG. 9 matches the registration condition. The processing operation of reading comparison will be described with reference to FIG.

In step 403 in FIG. 9, “the reading matches the conditions at the time of registration” is as follows. That is, at the time of reference, the reading of the input of the part currently being converted must match the reading in the registration table from the break position after the previous converted word. For example, in the case of "Fukusha" in the registration example shown in FIG.
Radiation-outs ...... ", weigh in rather than" ...... radiation ...
... ", in the case of such as" ...... radiation butterfly is ...... ", is a matching portion of the reading is part of the underline.

Therefore, it is determined whether or not the character string matches the read character string before the change (step 501), and if they match, it is determined whether or not the change type is 1 (lengthened) (step 502).
In step 503, it is determined whether or not it matches the remaining read character string portion after the change. As a result, if they do not match,
It is set that the reading does not match the condition at the time of registration (step 505). If they match, it is set that the reading matches the conditions at the time of registration (step 504).

FIG. 11 is a processing flowchart for judging whether or not it is suitable as a candidate other than a delimiter change shown in step 404 of FIG. Based on FIG. 11, a description will be given of a processing operation as to whether or not the candidate is suitable as a candidate other than a break change.

As described above, the non-separation-change candidate means a candidate different from the display candidate before registration change at the time of registration. In addition, the intention of the present invention is that at least the change display of the apparatus does not match the display desired by the operator for the change candidates changed by the operator, so that the candidates prepared in the apparatus are not included. If there is a candidate with different divisions, it is displayed with priority.

The above-mentioned candidates having different breaks are selected according to the evaluation of the candidate determined by the apparatus and have different reading lengths, and are located in the next order. However, if there is no candidate having a different reading length, the apparatus displays the first candidate determined in the word candidate string extraction (step 106) shown in FIG.

In the example of FIG. 2, for the input reading "Fukusha ..."
Candidates include “Fukusha”, “Fukushi”, “Fuku”, and “Fu”. As the conversion candidates, “Fukushi” and “Fukushi” are selected because “Fukushi” does not follow “ゃ” and “Fu” does not follow “Kusaku”. In general, long-read words are
Since the evaluation of the apparatus is high, "Fukusha" is first selected as the first candidate, and the candidate having the next different length becomes "Fuku". Therefore, the first suitable condition is "Fukusha ...
In the example of “...”, When the registration as shown in FIG. 3 is made, a candidate having a reading of “Fuku” is selected from the conversion candidate sequence.

On the other hand, in the determination as to whether or not the candidate is suitable as a candidate other than the change shown in FIG. 11, first, the candidate comparison position is set to the head position of the candidate column table (step 601). It is determined whether the length matches the reading length (step 602). If they match, if there is still something to compare,
Set the next comparison position to the next position in the candidate column table and step
The process of 602 is repeated (steps 606, 607).

If the length does not match the reading length in step 602, it is determined whether the change type of the registration table is 1 (step 602).
603) If it is 1, the candidate column table position is set as a candidate not to be changed (step 605). The change type is 1
If not, it is determined whether or not it matches the read length after the change in the registration table (step 604). If not, if there is still something to be compared, the comparison position is sequentially set to the next position. And repeat the processing of step 602 (steps 606 and 60
7).

If there are no more items to be compared, a suitable one is set as a candidate other than a change in segment (step 608). In addition,
In the case of the change type 2 (shortened), the non-separation candidate is assumed to match the read length after the change.

According to the above embodiment, it is possible to provide an efficient kana-kanji conversion processing device in which the conversion result intended by the operator can be easily obtained without the operator having to perform the delimiter changing operation each time. More specifically, in a device according to claim 1 in which the reading unit before the change and the connection information with the previous word are stored in the register unit of the word break change, the conversion intended by the operator is performed with a relatively small storage capacity. In contrast to a device in which a result can be easily obtained, a device according to claim 2 in which reading means before / after the change and information on connection with the previous word are stored in the word break change registering means is stored. Although the capacity is slightly increased, it is possible to realize a device in which the conversion result intended by the operator is more easily obtained. It is also possible to use both of them depending on the operator's preference in some cases.

It should be noted that the above embodiment is an example of the present invention, and it is needless to say that the present invention is not limited to this.

For example, as described above, the "part of speech of the previous word" or "part of speech of the previous word" in the present invention may be the "part of speech" or "part of speech" in the commonly used national grammar. What is necessary is just connection information with a word. By using the connection information with the previous word, the identity at the time of registration and at the time of reference is further improved.

〔The invention's effect〕

As described above in detail, according to the present invention, an efficient kana-kanji conversion processing apparatus that can easily output a conversion result intended by an operator without having to perform a delimiter change operation each time. Can be realized.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a processing flowchart for explaining an outline of the operation of a kana-kanji conversion processing apparatus according to one embodiment of the present invention, FIG.
FIG. 3 is an explanatory diagram showing an example of the correction process (step 114) in FIG. 1. FIG. 3 is a diagram showing the contents of the above-described word segmentation change registration table 12 when the correction operation shown in FIG. 2 is performed. FIG. 4 is a block diagram of the kana-kanji conversion processing apparatus according to the embodiment, FIG. 5 is a perspective view of the apparatus shown in FIG. 4, and FIG. 6 is a detailed diagram of the candidate extracting unit 2 in FIG. FIG. 7 shows a configuration example, and FIG.
FIG. 8 is a flowchart of a registration method at the time of changing a word delimiter in the embodiment. FIG. 8 is a flowchart of a process of the priority change processing outside the delimiter change shown in step 107 of FIG. 1. FIG. 9 is a flow chart of step 302 of FIG. FIG. 10 is a processing flowchart for determining whether or not the reading shown in step 403 in FIG. 9 matches the conditions at the time of registration, and FIG. Step 404
7 is a processing flowchart for determining whether or not the candidate is suitable as a non-separation candidate shown in FIG. 1: Input preprocessing unit, 2: Candidate extraction unit, 3: Conversion information management unit, 4:
Output control unit, 5: conversion control unit, 6: candidate extraction control unit, 7: candidate word string storage unit, 8: dictionary search unit, 9: conversion dictionary storage unit, 10:
Word dictionary, 11: Grammar dictionary, 12: Word break change registration table.

Claims (2)

(57) [Claims] 1. A kana-kanji conversion processing apparatus having kana-kanji conversion control means for converting a character string input by phonetic characters into a sentence mixed with kanji-kana by searching a word dictionary. When the operator modifies the unit division performed by the device using the delimiter change key, at least the reading length before the change and the previous word. And a registration unit for storing a word delimiter change for storing connection information and a type of change, and the kana-kanji conversion control unit operates the kana-kanji conversion control unit for a display divided into word units, which is a conversion result by the device. When a change operation is performed by the user using the delimiter change key, after the change is registered in the registration means, the next time the same reading is input, the contents stored in the registration means Kanji conversion processing device and controlling to display the converted results according. 2. The registration means is configured to store, in addition to the read length before change, connection information with a previous word, and the type of change, the read length after change. The kana-kanji conversion processing device according to claim 1, characterized in that: