JP3099457B2 - 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 processor such as a Japanese word processor.

[0002]

2. Description of the Related Art Conventionally, a kana-kanji conversion device of this kind has been disclosed in Japanese Patent Application Laid-Open No. 3-142658.
Performs Kana-Kanji conversion based on a basic dictionary that stores notation such as Kanji for reading words, stores the Kana-Kanji conversion result in the conversion result storage unit, and stores a set of word string patterns and their rewriting information as a set. The rules stored as rules are used as a rule dictionary, and when a rule that matches a rule in the rule dictionary is searched, the contents of the Kana-Kanji conversion result are rewritten and displayed on a display device. Such a series of actions was called rule conversion.

[0003] For example, when "Do not wear kimono" is input, "Kimono within regulation" may be output as a conversion result for the first time in normal conversion.

[0004] Next, when performing rule conversion based on the rules in the rule dictionary, when the <U-kimono: N><
If there is a rule of U-: N><D-ki-te: wear, te, ii>, an erroneous conversion of "within a kimono within a rule" is rewritten as "within a kimono". .

[0005]

However, when performing the rule conversion process, if the conversion result is erroneously converted to "Kimono within the rules", the portion rewritten by the rule conversion is "Kimono Although the conversion result is correct when "wearing", the conversion result "inside" immediately after the part rewritten by the rule remains erroneous conversion without any correction. Therefore, although the user wants to convert to "without kimono", an erroneous conversion is output as "within kimono". Also, in some cases, the rule conversion always converts "within a kimono in" even though the user wants to use the kimono in the form of "within a rule".

The present invention has been made in order to solve the above-mentioned problem. When rule conversion is performed using a rule dictionary, if there is an erroneous conversion immediately after rewriting, It is an object of the present invention to undo the rewriting by the rule conversion and improve the conversion accuracy of the correct kanji conversion device.

[0007]

In order to solve this problem, a kana-kanji conversion apparatus according to the present invention, as shown in FIG. 1, stores an input means for inputting a kana reading character string and a notation for reading a word. Kana-Kanji conversion means for performing Kana-Kanji conversion based on the basic dictionary and the connection table between the two words of the basic dictionary, Kana-Kanji conversion based on the basic dictionary and the connection table, and conversion results obtained by the Kana-Kanji conversion means A conversion result storage means for storing a pattern having a plurality of word strings in the basic dictionary and rules having rewrite information thereof, and a pattern matching the contents of the conversion result storage means from the rule dictionary. A rule search means for searching for a rule possessed by the user, and a conversion result storage means that matches the searched rule when the rule matched by the rule search means is searched. And a rule rewriting means for rewriting the contents.

Then, a connection judging means for judging, based on the connection table, whether or not the part rewritten by the rule rewriting means and the part not rewritten are connected correctly, and the connection judging means judges whether the part is rewritten. When it is determined that the connection is not correct, there are provided rule rewriting recovery means for restoring the rewriting by the rule rewriting means and output means for outputting the contents of the conversion result storage means.

[0009]

According to the kana-kanji conversion device of the present invention having the above-described configuration, the kana-reading character string input from the input means is converted by the kana-kanji conversion means with reference to the basic dictionary and the connection table.

[0010] Thereafter, the result of the Kana-Kanji conversion is stored in the conversion result storage means, and a rule matching the conversion result storage means is searched from the rules in the rule dictionary by the rule search means.

Next, according to the instruction of the searched rule, the matching part is rewritten in the conversion result storage means. Therefore, the connection determining means determines whether or not the last word of the portion rewritten according to the rule and the word immediately after the last word can be connected by referring to the connection table. Undo rule rewriting by means. As a result, the contents of the conversion result storage means converted into the Kana-Kanji character string are displayed by the output means.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings.

The overall structure of the kana-kanji conversion device will be described with reference to FIG.

An input device 10 such as a keyboard for inputting a character string to be subjected to kana-kanji conversion is connected to a central processing unit (CPU) 12 for controlling the entire device.
RAM 20 as storage means is connected to CPU 12,
A conversion result storage area 22 for storing a result of Kana-Kanji conversion, a reading input buffer area 24 for storing an input kana reading character string, and a conversion result storage area 22
An output buffer area 26 for storing the contents of the character string as a Kana-Kanji character string, a work area 28 for recording pointer information, and an auxiliary storage area 29 for storing the storage contents of the conversion result storage area 22 before rule rewriting is performed. Is stored.

As shown in FIG. 3, the conversion result storage area 22 stores the entire conversion result attached with information in word units, and identifies each word stored in a basic dictionary 42 described later. ID2 which is a unique numerical value for performing
2a, word reading length 22b, word notation 22c,
The homophone head ID 22d and the homophone end ID 22e are stored as data for each word.

Note that the homophone head ID 22d is stored in the input device 1
It is the same as the character string (word) input from 0, and indicates the word ID 42a stored in the basic dictionary 42 with the smallest numerical value. Also, the homophone end ID 22e
Is the same as the character string (word) input from the input device 10 and indicates the word ID 42a stored in the basic dictionary 42 with the largest numerical value.

ROM 30 for storing programs, dictionaries, etc.
Is connected to the CPU 12, the program unit 32 and the dictionary unit 4
It consists of 0. The program unit 32 includes a kana-kanji conversion program 34, a rule search program 36, a rule rewriting program 38, and a rule rewriting recovery program 39
And are stored.

The rule search program 36, the rule rewriting program 38, and the rule rewriting recovery program 3
9 is referred to as rule conversion processing.

The dictionary section 40 comprises a basic dictionary 42, a connection table 44, and a rule dictionary 46. As shown in FIG. 4, the basic dictionary 42 includes a word ID 42a which is a unique identification number of the word, a reading 42b of the word,
The kanji 42c of the word and the part of speech 42d of the word
And the meaning code 42 representing the category to which the word belongs
e is stored.

The connection table 44 stores data for determining whether or not there is a connection relationship between words based on part of speech information. For example, information that a noun and a case particle can be connected and information that a verb and a case particle cannot be connected are stored as table data.

As shown in FIG. 5, the rule dictionary 46 stores a set of rule items 46a to 46d as one rule. In general, rule items include <search factor-word pattern:
Rewriting information>, and one rule is created by combining these items. The search factor indicates the type of search method for rule conversion.
There are a U factor, a D factor, a Y factor, and an M factor. The U factor means reading and notation of a word stored in the conversion result storage area 22 and a perfect match, and the D factor is stored in the conversion result storage area 22. This means that the word that is being read is the same as the word. Further, the Y factor means that it matches the reading of one or more word strings stored in the conversion result storage area 22. M
The factor means that they match with the meaning code of the word stored in the conversion result storage area 22. The word pattern indicates a word that matches the rule item with the item in the conversion result storage area 22, and contains a word reading or notation or a meaning code.

The rewriting information is any one of the unchanged marker N, the deleted marker D, and the word string. When the unchanged marker N is stored, the conversion result storage area 22 corresponding to the item of the rule dictionary 46 is stored. Means that the word is not rewritten. When the delete marker D is stored, it indicates that the item is to be deleted.
In the case of a word string, it indicates that a word in the conversion result storage area 22 that matches an item of the rule dictionary 46 is rewritten to the word string. Here, the word string is a plurality of words (only one word may be stored) stored in the rule dictionary 46 by a word ID.
2, the word reading length 22b, the homophone head ID 22c, and the homophone tail ID in the conversion result storage area 22
Settings such as 22d are possible. In FIG. 5, for easy understanding, the word ID portion is represented by a reading or a notation instead of a numerical value.

Next, the operation of the kana-kanji conversion device thus constructed will be described with reference to the flowchart of FIG.

For example, when a reading character string "Do not wear kimono" is input from the input device 10, the kana character code is stored in the reading input buffer area 24 of the RAM 20 (S10). Thereafter, the kana character code stored in the reading input buffer area 24 is converted into a kanji mixed kana sentence based on the basic dictionary 42 and the connection table 44 by the kana / kanji conversion program 34 in the ROM 30, and the word ID is stored in the conversion result storage area 22. Etc. are stored. For example, in the basic dictionary, the words “prescription” and “inside” are selected and converted into a sentence “Kimono within the provisions” mixed with kanji or kana. Then, the word ID is stored in the conversion result storage area 22 (S12). Then, the rule conversion process is performed by the rule search program 36 and the rule rewriting program 38 in the ROM 30 with reference to the contents of the conversion result storage area 22 and the rule dictionary 46 (S14).

Next, the specific processing of the rule conversion processing will be described with reference to the flowcharts of FIGS. 7, 8 and 9.

First, a pointer P1 is set to the first word "Kimono" of "Kimono within rules" stored in the conversion result storage area 22, and the pointer information of the pointer P1 is stored in the RAM2.
0 is stored in the work area 28 (S30).

Next, it is searched whether or not a rule matching the word pointed by the pointer P1 exists in the rule dictionary 46 (S3).
2). In the rule search subroutine shown in FIG. 9, first, the pointer P5 is set to the first rule 5-1 in the rule dictionary 46.
And the pointer information is recorded in the work area 28 (S320). Next, the pointer P6 is set to the first item 46a of the rule 5-1 indicated by the pointer P5, and the pointer information is stored in the work area 28 (S322). Next, the pointer P7 is set to the first word “kimono” of the kana-kanji conversion result in the conversion result storage area 22 pointed to by the pointer P1 (S324).

Here, the rule pointed by the pointer P6 <U-
Listening: N> and conversion result storage area 2 pointed by pointer P7
Then, the word “kimono” is compared with the word “kimono” (326). In this case, the rule does not match the word (S326 / NO), and the rule 5-1 indicated by the pointer P5 is not the last rule in the rule dictionary 46 (S328 / NO). 2 and the pointer information is stored in the work area 28.
(S330), and returns to S322. next,
The pointer P6 is set in the first item 47a of the rule 5-2 indicated by the pointer P5, and the pointer information is set in the work area 28.
(S322). Next, the pointer P7 is set to the first word “kimono” of the kana-kanji conversion result in the conversion result storage area 22 pointed to by the pointer P1 (S324).

Here, the rule item pointed by the pointer P6 <
U-Kimono: N> 47a is compared with the word "Kimono" in the conversion result storage area 22 indicated by the pointer P7 (326).
In this case, the rule and the word match (S326, YES),
Since the pointer P6 is not the last item (S332 · N
O), the pointer P7 is moved to the next word "" of "kimono" stored in the conversion result storage area 22, and the pointer P6 is also moved to the next item <U-: N> in rule 5-2. (47
b) and stores the pointer information in the work area 28.
(S334), and returns to S326.

Here, the rule item pointed by the pointer P6 <U
-A: N> 47b is compared with the word "wo" in the conversion result storage area 22 indicated by the pointer P7 (326). In this case, the rule and the word match (S326, YES), and since the pointer P6 is not the last item (S332, NO), the pointer P7 is moved to the next word "regulation" in the conversion result storage area 22, and P6 also moves to the next item in the rule 5-2, <D-Iki: wear, te, i> (47c), and stores the pointer information in the work area 28 (S33).
4), and return to S326.

Here, the rule item pointed by the pointer P6 <D
-Coming: wearing, te, ii> (47c) is provided with a factor D and means a match in reading. Therefore, the pointer P7
(S326: YES), and the item 47c indicated by the pointer P6 is the last item (S33).
2 ・ YES), the rule flag ON is stored in the work area 28 (S336), and the process proceeds to S34 shown in FIG.

If no matching rule is found (S326: NO), the pointer P5 is sequentially moved to the next rule until the rule indicated by the pointer P5 becomes the last rule in the rule dictionary (S328: NO). Then, a matching rule is searched (S322 to S330). Pointer P
When the rule indicated by No. 5 is the last rule in the rule dictionary (S328: YES), a rule flag OFF indicating that there is no matching rule is stored in the work area 28 (S338).

Next, the rule rewriting process shown in FIGS. 7 and 8 will be described.

Since the rule flag of the work area 28 is ON in the rule search process of S32 (S34.
YES), and enters a rule rewriting process by the rule rewriting program 38. First, the pointer P2 is set to the first word “Kimono” of the word string “Kimono within the rules” stored in the conversion result storage area 22 pointed to by the pointer P1, and the pointer P2 is set.
Is stored in the work area 28 (S3).
6). Next, in the word string “Kimono within rules” stored in the conversion result storage area 22 pointed to by the pointer P7, the word “rule” at the end of the part that matches the rule
And the information of the pointer P3 is stored in the work area 28 (S38). Subsequently, the pointer P4 is set in the rule dictionary 4
6 is set as the first item of the matched rule 5-2, and the pointer information of the pointer P4 is stored in the work area 28 (S40).

Next, of the conversion results stored in the conversion result storage area 22, the range from the word indicated by the pointer P2 to the word indicated by the pointer P3 is stored in the auxiliary storage area 29 (S42), and the flow proceeds to S44. .

Here, in the rule 5-2 in the rule dictionary 46, the head item <U-kimono: N> indicated by the pointer P4.
(47a) is provided with the invariable marker N (S4).
4. YES), so that it is left as it is without rewriting, and the process proceeds to S46.

If the deletion marker D is standing at the location of the rewriting information in the item 46b indicated by the pointer P4 (S44 / NO, S48 / YES), the word "kimono" indicated by the pointer P2 in the conversion result storage area 22 is displayed. Is deleted (S
50), and proceed to S46. Further, when there is a word string at the location of the rewriting information in the item 47a indicated by the pointer P4 (S44 / NO, S48 / NO, S52 / YES), the word "kimono" indicated by the pointer P2 in the conversion result storage area 22.
Is replaced with the word string (S54), and the process proceeds to S46.

In S46, it is determined whether the word pointed by the pointer P2 is the same as the word pointed by the pointer P3.
That is, when the word indicated by the pointer P2 and the word indicated by the pointer P3 are the same, it means that the pointer P2 has advanced to the last word of the word string that matches the rule. In this case, the “kimono” indicated by the pointer P2 is stored in the conversion result storage area 2
Since the word is not the last word of the conversion result stored in No. 2 (S46: NO), the word pointer P2 is moved to the next word "" stored in the conversion result storage area 22 and the pointer P4 is changed to the rule 5 -2: <U-: N> (47
The process moves to b) (S56), and returns to S44.

Here, the item pointed by the pointer P4 <U-
: Since there is no change marker N in N> (S44 • YE
S) The process proceeds to S46 without rewriting. In this case, since "" indicated by the pointer P2 is not the last word of the conversion result stored in the conversion result storage area 22 (S46 / NO), the word pointer P2 is moved to the next word "prescribe". Then, the pointer P4 is moved to the next item in the rule 5-2, <D-Iki: wear, te, i> (47c) (S56), and the process returns to S44.

Next, the word string “wear” is added to the rewrite information of the item <D-kite: wear, te, i> indicated by the pointer P4.
(S44.NO, S48.NO, S52.Y)
ES), the word “regulation” indicated by the pointer P2 stored in the conversion result storage area 22 is rewritten as “wearing” (S
54), and proceed to S46.

Here, since the word pointed to by the pointer P2 is the last word of the part of the word string stored in the conversion result storage area 22 that matches the rule indicated by the pointer P3 (YES in S46). ), And proceed to the rule rewriting recovery process (S57).

FIG. 10 shows the rule rewriting recovery process.
This will be described with reference to FIG.

The pointer P2 indicates the part of the word "wearing kimono" that matches the rule in the word string "wearing kimono" stored in the conversion result storage area 22 by the rule conversion processing. The word string “wearing” at the end is shown. Here, the word string “wearing” indicated by the pointer P2
Is not the last word in the word string stored in the conversion result storage area 22 (S400: NO), the pointer P8 is changed to the pointer P2 in the word string stored in the conversion result storage area 22. Is set to the word “in” immediately after the word string “wear”, and the pointer information of the pointer 8 is stored in the work area 28 (S410). Next, the rule rewriting recovery processing program 3 determines whether the word string “wearing” indicated by the pointer P2 and the word “inside” indicated by the pointer P8 are connectable.
9 by referring to the connection table 44 (S42).
0).

Here, since the word string "wearing" indicated by the pointer P2 is not connectable to the word "inside" indicated by the pointer P8 (S420: NO), the word in the conversion result storage area 22 indicated by the pointer P1 is not detected. From the first word of the portion that matches the rule in the column to the last word that matches the rule indicated by the pointer P2, the contents stored in the auxiliary storage area 29 are rewritten to the contents before being rewritten by the rule rewriting program 38. Return (S430). Specifically, what is converted to "within the kimono" by the rule conversion process is returned to the original conversion result of "within the kimono".

When the pointer P2 is the last word of the words stored in the conversion result storage area 22 in S400 (S400: YES), or S420.
When the word indicated by the pointer P2 and the word indicated by the pointer P8 can be connected (YES at S420),
The process proceeds to S58 shown in FIG. 8 without performing the rule rewriting recovery process (S430).

Here, if the pointer P1 is not at the last word of the word string stored in the conversion result storage area 22 (NO in S58), the pointer P1 is moved to the next word one by one (S60). Again, the same rule search processing (S3
2) is performed by the rule search program 36. In this case, the pointer P1 is moved to the second word “O” of the character string “Kimono within the rules” stored in the conversion result storage area 22. Then, the processing of S32 to S60 is repeated. Finally, when the pointer P1 has become the last word “at” in the conversion result storage area 22 (S58: YES), the rule conversion process ends.

When the rule flag stored in the work area 28 is OFF (S34 / NO), the flow proceeds to S58. Here, if the pointer P1 is not at the last word of the word string stored in the conversion result storage area 22 (S58 / NO), the pointer P1 is moved to the next word one by one (S60), and again A similar rule search process (S32) is performed by the rule search program 36. In this case, the pointer P1 is moved to the second word "" of the character string stored in the conversion result storage area 22 (S58.
NO, S60). Then, the processing of S32 to S60 is repeated. Finally, when the pointer P1 is the last word in the conversion result storage area 22 (S58: YES), the rule conversion processing ends.

As described above, the kana-kanji conversion apparatus of the present embodiment converts the kana-kanji conversion program 34 into "within the kimono within the rules" by the rule conversion processing. When the conversion is performed, by returning the kimono to the original candidate of “Kimono within the rules”, it is possible to prevent the conversion that the user does not intend by the rule conversion.

[0049]

As is apparent from the above description, the kana-kanji conversion apparatus of the present invention matches the rules by the rule conversion process, avoids the connection mismatch between the rewritten portion and the non-rewritten portion, and applies an appropriate rule. Only kana-kanji conversion accuracy can be improved by applying only rules and not applying rules at inappropriate positions.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of the present invention.

FIG. 2 is a block diagram of a control unit of the kana-kanji conversion device according to the embodiment.

FIG. 3 is a conceptual diagram of the contents of a conversion result storage area according to the embodiment.

FIG. 4 is a conceptual diagram of the contents of a basic dictionary according to the present embodiment.

FIG. 5 is a conceptual diagram of the contents of a rule dictionary according to the present embodiment.

FIG. 6 is a flowchart of the kana-kanji conversion device of the present embodiment.

FIG. 7 is a flowchart of rule conversion according to the embodiment.

FIG. 8 is a flowchart of rule conversion according to the embodiment.

FIG. 9 is a flowchart of a rule search subroutine of the embodiment.

FIG. 10 is a flowchart of a rule rewriting recovery processing subroutine of the embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Input device 12 CPU 20 RAM 22 Conversion result storage area 24 Reading input buffer area 26 Output buffer area 28 Work area 29 Auxiliary storage area 30 ROM area 32 Program part 34 Kana-Kanji conversion program 36 Rule search program 38 Rule rewriting program 39 Rule rewriting recovery Program 40 Dictionary unit 42 Basic dictionary 44 Connection table 46 Rule dictionary 50 Output device

Claims (1)

(57) [Claims] An input unit for inputting a kana reading character string; a basic dictionary storing notations for reading words; a connection table storing a connection relationship between two words of the basic dictionary; Kana-Kanji conversion means for performing Kana-Kanji conversion on the basis of the Kana-Kanji conversion means, a conversion result storage means for storing a conversion result of the Kana-Kanji conversion means, A rule dictionary storing rules; a rule search unit for searching a rule having a matching pattern from the rule dictionary with respect to the contents of the conversion result storage unit; and a search when the rule search unit finds a matching rule. Rule rewriting means for rewriting the contents of the conversion result storage means that matches the set rule, and the rule document based on the connection table. The connection determining means and the connection determining means for determining whether the part rewritten by the replacement means and the part not rewritten are correctly connected, when the connection between the words is determined to be incorrect, A kana-kanji conversion device, comprising: rule rewriting recovery means for restoring the contents rewritten by the rule rewriting means; and output means for outputting the contents of the conversion result storage means.