JPH0619880A - Kana/kanji converter - Google Patents

Abstract

(57) [Summary] [Purpose] Kana-Kanji conversion that enables Kana-Kanji conversion that matches the co-occurrence relationship, according to the flow of the sentence even when the conversion between the words that have a co-occurrence relationship or between words is confirmed in the middle of a phrase. Get the device. [Configuration] When a kana character string is input by the input means 1,
The kana character string stored in the second character string storage means 7 is read, and the newly input kana character string is connected after the read kana character string, and a series of connected kana character strings. Is stored in the conversion character string storage means 11 and is converted. The character string delimiter 13 divides the character string into phrases, and the converter 15 converts the kana-kanji mixed character string.
At that time, the last clause of the confirmed kana-kanji character string stored in the first character string storage means 5 and the co-occurrence information in the co-occurrence information storage means 9 are referred to. After the conversion, the display unit 19 displays the conversion result excluding the Kana-Kanji character string stored in the first character string storage unit 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a kana-kanji conversion device for converting a kana-character string into a corresponding kana-kanji mixed character string according to a dictionary, and more particularly to a kana-kanji conversion device capable of appropriately connecting kana-kanji characters to a fixed phrase. .

[0002]

2. Description of the Related Art A Japanese word processor inputs word reading information by kana and reads out Kanji corresponding to the input reading information from a dictionary section. In this case, grammatical information, co-occurrence relations, frequency information, etc. It is generally used. The grammar information is connection information between parts of speech and parts of speech, and determines, for example, punctuation and conversion candidates based on the rule that a particle follows a noun but not an auxiliary verb. The co-occurrence relationship is a relationship that represents which word and which word is likely to appear together. For example, “There is no kanbou”
For the same kana character string, "Baby cries" for the kana character string, and "birds cry" for the kana character string. , If the previous word is "baby", "cry",
In the case of "bird", "crowing" tends to appear depending on the co-occurrence relationship. Frequency information is information that shows which candidates are likely to appear for the same reading based on past input situations. The efficiency of can be improved.

A method of automatically segmenting a solid written sentence is described in, for example, the IPSJ Journal Vol. Two
0, No. 4, July 1979, pp. 337-34
It is disclosed in the paper “Divided text and kana-kanji conversion” published in No. 5. FIG. 5 is a flow chart showing the operation of the segmentation and kana-kanji conversion device disclosed in this paper. The outline of the operation based on the figure is
The input sentence is read (step 50), the read input sentence is decomposed into character strings (hereinafter referred to as "divisions") divided by punctuation marks (step 51), and one division is extracted from the decomposed divisions (step 52). ). Prefix processing, independent word processing, and suffix processing are performed on the extracted section to extract the phrase form (step 55). The two-phrase longest matching method is applied to the extracted bunsetsu to make an appropriate delimitation (step 57), and the demarcated bunsetsu is converted to kanji (step 58). If two bunsetsu are not found in succession from the bunsetsu delimiter, the longest bunsetsu form from the bunsetsu delimiter is used as a bunsetsu to find the delimiter of the subsequent character string by appending word segmentation (step 5), and the process is continued. To do.

[0004]

In the conventional kana-kanji conversion device as described above, since the character string whose conversion has been confirmed is not referred to in the subsequent conversion processing, it is not included in the middle of a phrase or in a word having a co-occurrence relationship. When the user confirms the conversion between words, for the kana character string that is input after the confirmed word, a candidate kana-kanji mixed sentence that appropriately corresponds to the confirmed character string may appear. could not.

The present invention has been made in order to solve the above problems, and fits the flow of a sentence even in the middle of a bunsetsu or even when a conversion between words having a co-occurrence relationship is decided.
The purpose is to obtain a kana-kanji conversion device that can convert kana-kanji according to a co-occurrence relationship.

[0006]

A kana-kanji conversion device according to the present invention is a first character string storage means for storing the last phrase of a fixed kana-kanji mixed character string located immediately before an input kana character string. A second character string storage means for storing a kana character string corresponding to the kana-kanji mixed character string stored in the first character string storage means; and a kana stored in the second character string storage means. A character string is read out, a converted character string storage means for connecting and storing the input kana character string after the read kana character string, and a kana character string stored in the converted character string storage means Of the kana character string that is divided by the character string delimiter means, the character string delimiter that divides the kana character string stored in the second character string storage means as a partial character string into one clause First
The phrase is converted into a kana-kanji mixed character string that includes the kana-kanji mixed character string stored in the first character string storage means as a partial character string, and the kana-kanji mixed character string that appropriately follows the first phrase after the second phrase. And a display unit for displaying a kana-kanji mixed character string excluding the kana-kanji mixed character string stored in the first character string storage means from the kana-kanji mixed character string converted by the converting means. It is a thing.

Further, the conversion means, based on the co-occurrence information stored in the co-occurrence information storage means in which the word-to-word co-occurrence information is stored, the second kana character string stored in the converted character string storage means.
It is designed to convert the text after the phrase into a kana-kanji character string.

[0008]

In the kana-kanji conversion device of the present invention,
When the kana character string is input, the kana character string stored in the second character string storage means is read, and the newly input kana character string is connected after the read kana character string,
A series of concatenated kana character strings are subjected to conversion, the kana-kanji character string stored in the first character string storage means after conversion is excluded and the conversion result is displayed.

Further, when converting the second and subsequent clauses of the kana character string stored in the converted character string storage means into a kana-kanji character string, the co-occurrence information is referred to so that a candidate matching the co-occurrence relationship appears.

[0010]

【Example】

Example 1. FIG. 1 is a block diagram showing an embodiment of a kana-kanji conversion device according to the present invention. In the figure, reference numeral 1 denotes an input means such as a keyboard provided with a kana key, a conversion key, a decision key, etc., which inputs a change instruction of a display word other than a kana character string to the control unit 3. The control unit 3 is composed of a microprocessor and performs data processing described later according to a control program written in a memory (not shown).

Reference numeral 5 is a first character string storage means for storing the kana-kanji mixed notation of the last phrase of the confirmed kana-kanji character string, and 7 is a confirmed kana-kanji character string stored in the first character string storage means 5. Second character string storage means for storing kana notation corresponding to, and 9 is co-occurrence information storage means for storing information on the co-occurrence relationship between words. Reference numeral 11 reads the kana character string stored in the second character string storage means 7, combines the kana character string input from the input means 1 after the read kana character string, and combines the kana character strings. It is a conversion character string storage means for storing a series of kana character strings. Reference numeral 13 is a kana character string delimiter that divides the kana character string stored in the converted character string storage means into clauses, and 15 is a kana character string delimiter 1.
It is a conversion means for performing Kana-Kanji conversion for the clauses delimited by 3.

Reference numeral 17 is a collating means, which performs the following operations. That is, the kana character string stored in the second character string storage means 7 is read out, and the kana character string read out by this kana character string and the kana character string first segment kana separated by the kana character string delimiter 13 are read. The kana-kanji mixed character string stored in the first character string storage means 5 is checked and the read kana-kanji mixed character string and the conversion means 15 are converted into kana-kanji mixed sentences. It is to collate with the Kana-Kanji character string in the first bunsetsu of the Takana-Kanji mixed sentence. Reference numeral 19 is a display means for displaying a kana-kanji mixed sentence in which the kana-kanji mixed character string stored in the first character string storage means is excluded from the kana-kanji mixed sentence converted by the converting means 15.

FIG. 2 is a flow chart showing the operation of this embodiment, and FIG. 3 is an explanatory diagram for explaining the relationship between key operations and the display results displayed on the display means 19. In FIG. 3, (a) to (e) show display results at each stage of key operation. The operation of the control unit 3 will be described below with reference to FIGS. First, the conversion of the character string "I am Ito" is taken as an example, and the kana character string "I" is input from the input means 1 (the display result is (a) in FIG. 3).
By operating the conversion key of the input means 1, "I" appears as a kana-kanji candidate ((b) in FIG. 3), and by operating the enter key, "I" is confirmed ((c) in FIG. 3).
The above is the same operation as the conventional one. At this point, the kana character string "I" is stored in the first character string storage means 5, and the kana character string "I" is stored in the second character string storage means 7.

[0014] Next, from the input means 1, "I'm happy".
When a kana character string is input (step 30), the character string "I am my mother" is displayed on the display means 19 ((d) in FIG. 3). At this time, the kana character string "I" stored in the second character string storage means 7 is read out, and the newly input "Kana" character string is newly input "Haitoto". Are concatenated and a series of kana character strings "I am Itou" are stored in the converted character string storage means 11 (step 31).

The kana character string "I am Ito" stored in the conversion character string storage unit 11 is divided into "I" and "Ito" by the character string delimiter unit 13 (step 32). The collating means 13 is the character string delimiting means 1
The first phrase separated by 3 is collated with the first phrase "Iwa" and the kana character string "I" stored in the second character string storage means 7, and the first character string is separated by the character string separation means 13. It is checked whether or not the phrase "Iwa" includes the kana character string "I" of the second character string storage means 7 as a partial character string (step 33), and if it does, the first phrase. The conversion means 13 performs kana-kanji conversion processing for "Iwa" (step 34).

If the first phrase delimited by the character string delimiter 13 does not include the character string in the second character string storage 7 as a partial character string, another delimiter is performed (step 32) and a new character is newly added. It is checked whether or not the delimited first clause contains the kana character string stored in the second character string storage means 7 as a partial character string (step 33). If it is not included as a partial character string, it is determined whether another delimiter is possible (step 33a). If another delimiter is possible, the process returns to step 32 and the same operation is performed. Even if the above operation is repeated to perform all possible delimiters, there is no delimiter such that the first bunsetsu of the delimited bunsetsu contains the kana character string in the second character string storage means 7 as a partial character string. In this case, of the kana character strings stored in the converted character string storage means 11, the same kana character string as the kana character string stored in the second character string storage means 7 is the first character string storage means. The kana-kanji character string identical to the kana-kanji character string in 5 is converted and confirmed (step 35), and conversion processing is performed on the newly input kana-character string (step 3).
7).

By the way, in the case of the present embodiment, the first phrase "Iwa" delimited by the character string delimiter 13 is replaced by the kana character string "I" in the second character string storage unit 7. Since it is included as a partial character string, kana-kanji conversion processing is performed on "Iwa" (step 3).
4). When "Iwa" appears as a candidate as a result of the Kana-Kanji conversion processing, the matching unit 17 is the converted first phrase "Iwa" and the Kana-Kanji character string stored in the first character string storage unit 5. Is "I", which is the first phrase converted by collating "I", included "I", which is the Kana-Kanji character string stored in the first character string storage means 5, as a partial character string? It is checked whether or not (step 36).

If it is not included as a partial character string, it is judged whether or not there is another candidate (step 36a). If there is another candidate, the process returns to step 34 to perform re-conversion,
The same collation is repeated. First of all conversion results
If the kana-kanji character string of the bunsetsu does not include the character string in the first character string storage means 5 as a partial character string, the process returns to step 32 and the first bunsetsu delimitation process is performed again. When it is included as a partial character string or when a corresponding candidate appears by the above-mentioned re-conversion, the kana-kanji character after conversion is performed for the first phrase of the kana character string stored in the converted character string storage means 11. Kana-kanji is delimited by the row, and the kana-character string of the part excluding the first clause is delimited (step 37), and kana-kanji conversion is performed by operating the conversion key to obtain the desired kana-kanji. If the character string appears, it is confirmed (step 38).

In the case of this embodiment, the converted first phrase "Iwa" is the kana-kanji character string "I" stored in the first character string storage means 5 as a partial character string. Since it contains, the first character in the kana character string “I am Ito” stored in the conversion character string storage means 11.
The phrase "Iwa" is fixed as "Iwa", and the kana character string except for this is separated from the kana character string (in the case of this example, This operation is omitted because "Itosu" cannot be further divided into clauses (step 37). Kana-Kanji conversion is performed by the operation of the conversion key, and the desired Kana-Kanji character string "Ito is" appears. If so, it is confirmed (step 38).

At this time, the conversion result in the conversion character string storage means 11 becomes "I am Ito", and the kana-kanji character string stored in the first character string storage means 5 from this kana-kanji character string "I "Is Ito" excluding "is displayed. Therefore, the display result on the display means 19 at this time is "I am Ito" (step 39) (FIG. 3).
(E)).

In the conventional kana-kanji conversion device, the conversion is performed only for the kana character string newly input after the confirmation without referring to the phrase "I" which is the confirmed conversion. The newly entered kana character string "Haitouto" was converted into "dividend." However, according to the present invention, the conversion candidate of the newly input kana character string is made to appear by referring to the fixed character string "I", so that "is Ito" and "I" You can easily convert to the correct Kana-Kanji character string.

Next, conversion of the character string "Torashiri" will be described as an example. FIG. 4 shows key operation and display means 19.
FIG. 4 is an explanatory diagram for explaining the relationship with the display result displayed in FIG. 3, in which (f) to (j) show the display result at each stage of the key operation. A kana character string "toriga" is input by the input means 1 (display result is (f) in FIG. 4), and conversion is performed by operating the conversion key ((g) in FIG. 4). The kana-kanji character string is fixed ((h) in FIG. 4). At this point, the character string "bird" is stored in the first character string storage means 5, and the character string "toriga" is stored in the second character string storage means 7.

After that, when a kana character string "nashi" is input ((i) in FIG. 4), "toriga" which is the character string in the second character string storage means 7 is newly input. "Without"
Are concatenated and the kana character string “Torashiri” is stored in the conversion character string storage means 11 and becomes a conversion target. In this kana character string, the first clause is delimited as "toriga" and the second clause is delimited, and the delimited "toriga" of the first clause is a character in the second character string storage means 7. Check to see if it contains the string "Toriga" as a substring. In the case of this example, since it is included, the conversion for the first phrase "Toriga" is performed, and the Kana-kanji character string "Toriga" appears as a candidate. It is checked whether or not "bird" that appears as a candidate includes "bird" in the first character string storage means 5 as a partial character string. In the case of this example, it is included, so after confirming that the first phrase is "bird",
The conversion for the second clause, “null”, is performed.

At this time, the character string "bird" determined as the first clause is referred to as co-occurrence information, and "crow" having a co-occurrence relationship with this appears as a conversion candidate (FIG. 4).
(J)). Conventionally, when a character string "Kan" is input and converted after "Toriga" is confirmed, "Kana", which is the confirmed Kana-Kanji character string, is not referenced, so "No", "Cry", and "Cry" A conversion candidate appeared according to the frequency of use in the past, and the conversion efficiency was poor. However, according to the present invention, the newly input “n” is converted with reference to the fixed kana-kanji character string “bird”, so that “calling” that has a co-occurrence relationship with “bird” does not occur. It appears first as a conversion candidate and allows efficient conversion.

[0025]

As described above, according to the present invention, a conversion candidate for a newly input kana character string is made to appear by referring to a fixed kana kanji character string, and thus a fixed kana kanji character is determined. You can easily convert to a Kana-Kanji character string that follows the sequence correctly.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a kana-kanji conversion device according to the present invention.

FIG. 2 is a flowchart showing the operation of the first embodiment.

FIG. 3 is an explanatory diagram for explaining a relationship between a key operation and a display result displayed on the display means 19.

FIG. 4 is an explanatory diagram for explaining a relationship between a key operation and a display result displayed on the display means 19.

FIG. 5 is a flowchart showing the operation of a conventional space-dividing and kana-kanji conversion device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 input means 5 1st character string storage means 7 2nd character string storage means 9 co-occurrence information storage means 11 converted character string storage means 13 character string delimitation means 15 conversion means 19 display means

Claims (2)

[Claims] 1. A kana-kanji conversion device for inputting a kana character string, searching a dictionary for a kanji corresponding to the input kana, and converting the kana into a kana-kanji mixed sentence. First character string storage means for storing the last phrase of the kana-kanji mixed character string, and a second character for storing a kana character string corresponding to the kana-kanji mixed character string stored in the first character string storage means A column storage means and a conversion character string for reading out the kana character string stored in the second character string storage means and concatenating and storing the input kana character string after the read kana character string. A storage unit and a character that divides the kana character string stored in the converted character string storage unit into phrases such that the first phrase includes the Kana character string stored in the second character string storage unit as a partial character string. Row And a kana-kanji character string delimited by the character string delimiter means, the first clause is converted to a kana-kanji character string containing the kana-kanji character string stored in the first character string storage means as a partial character string. However, the second and subsequent phrases are converted into a kana-kanji mixed character string that appropriately follows the first phrase, and kana-kanji mixed character strings converted by the conversion means are stored in the first character string storage means. A kana-kanji conversion device comprising: display means for displaying kana-kanji mixed character strings excluding the kana-kanji character strings. 2. The conversion means, based on the co-occurrence information stored in the co-occurrence information storage means storing the word-to-word co-occurrence information, the second clause of the kana character string stored in the converted character string storage means. 2. The kana-kanji conversion device according to claim 1, wherein the subsequent characters are converted into kana-kanji character strings.