JPH05174011A - Kana/kanji conversion processor - Google Patents

Abstract

PURPOSE:To improve the KANA (Japanese syllabary)/KANJI (Chinese character) conversion accuracy for the close batch conversion processing by adding and registering the andividable flags of paragraphs to the words which are registered in a dictionary like an ordinary dictionary, a user dictionary, etc. CONSTITUTION:A key input means 101 is provided together with a display device 1.02 which has a display screen, an input character storage means 103 which stores the inputted KANA character strings, a paragraph dividing means 104 which divides the KANA character string into the KANA character paragraphs, a KANA/KANJI conversion dictionary 105 which stores the KANJI to be converted into KANA characters, a converter means 106 which converts the KANA characters into KANJI or the KANA-KANJI sentences, a specific paragraph conversion dictionary 107 which registers a specific KANA character paragraph and its corresponding KANJI or KANA-KANJI sentence, a paragraph control means 108 which controls the means 104 so as not to divide the specific KANA character paragraph if registered in the dictionary 107 among those KANA character strings end also converts the specific paragraph into KANJI or a KANA-KANJI sentence, and an output means 109 which outputs the character strings stored in the means 104 to a display screen.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a Japanese word processor and a kana-kanji conversion processing device capable of kana-kanji conversion processing.

[0002]

2. Description of the Related Art Conventionally, in this type of kana-kanji conversion processing device, a kana-kanji conversion dictionary (hereinafter referred to as a normal dictionary) for converting kana characters input from a keyboard into kanji is provided. However, since the number of words registered in the ordinary dictionary has a limit such as capacity, the ordinary dictionary cannot sufficiently perform the conversion.

Therefore, a specific phrase conversion dictionary (hereinafter referred to as a user dictionary) in which a user can newly register a word is provided to improve the operability of kana-kanji conversion. That is, the words registered in the user dictionary are generally used by the user, and thus the priority of the conversion processing output is higher than that of the words registered in the normal dictionary, and frequency information is added. As a method of increasing the conversion efficiency by controlling the conversion processing output in this way, there is a method disclosed in Japanese Patent Laid-Open No. 55-143648.

[0004]

However, in the conventional kana-kanji conversion processing device, when the input kana character string is a long sentence, the program processes which segment is most appropriate. With the conventional word information, it may be divided by a segment having a different meaning rather than a segment of a phrase having a meaning required by the user. Therefore, even if a word required by the user is registered in the user dictionary, it may not be reflected in the Kana-Kanji conversion process, and the initial effect may not be exhibited.

[0005] For example, when "I went to the sword to make a story" was entered, the desired kana-kanji conversion was "I told the Prince Shotoku who built Horyuji."
There are cases where the meanings are changed differently, such as "I said to the Ambassador of Houdoji who established Horyuji" or "I went to the small and painful ambassadors who set the release time". This is because the conversion candidates (words) for the input character string differ depending on how the segment delimiter positions are processed.

Example 1: When "Shoutokutai" is used as one phrase, it is converted to "Shutoku Taishi". Example 2; When "Shoutoku / Taishi" has two clauses,
"Otoku (independent word), Shode (name), Chengde (city), Masanori (year name), Shotoku (university name)" + "Taishi (Yamana), Prince (Town name), Ambassador (Independent word), Taishi (independence word), Thailand (country) +, other meaning +, Tai +, Tai +, against, sea bream +
There are conversion candidates, such as "shi, soldier, body + shi". Example 3; If you use "Shouto / Kutai / Shi" as three clauses, you can read "Merchandise (independent language), Sho and (name), award, title, chapter, testimony, small, and quotient. , Sho, Sex, Akira, Akira, Akira, Akira, Positive, Ministry, Still, SHOW and "+" Ward "+
"Shi, Shi, Shi, Shi, Shi, Shi, Shi, Shi, Shi, Shi, Shi, Shi, Shi,
There are conversion candidates such as "child".

Example 4; When "sho / tokutai / shi" has three clauses, "quotient, sho, award, title, chapter, testimony, small, quotient,
"General, sex, Akira, Akira, Akira, Tadashi, Ministry, Sho," + "special treatment" + "shi, city, samurai, poetry, teacher, history, four, Mr, Shi, death,
There are conversion candidates such as "word, visual, child". Example 5: If "Sho / Utoku / Taishi" is used as three bunsetsu, then "sign, calligraphy, cord, first" + "virtues" + "Taishi (Yamana),
Prince (town name), ambassador (independent word), ambition (independent word), Thailand (country) +, other meaning +, Tai +, Tai +, opposite, sea bream +
There are conversion candidates, such as "shi, soldier, body + shi".

Therefore, a large number of conversion candidates are inferred according to the way of segmenting the clauses. Therefore, if the program cannot process "Shoutokutaishi" as one phrase, even if the word "Shotoku Taishi" is registered in the user dictionary, it is converted by a conversion candidate having a different meaning.

Here, as another method, there is a method of registering "Shotoku (university name)" + "Taishi (town name)" in the co-occurrence dictionary (composite example dictionary) using the co-occurrence (example) relationship. .. However, although "Shotoku" + "Taishi" can be registered, since it does not become a person's name from the meaning of the word in the preceding clause, inconvenience may occur in the conversion processing of the succeeding clause. For example,
It is impossible to distinguish whether "I went to Prince Shotoku" from "I said to the person +" or "I went to the place +". However, if "Shotoku Taishi" has a meaning code of a personal name,
It is possible to convert from "I went to Shotokutai" to "I told Prince Shotoku".

That is, when registering in the user dictionary, in order to accurately demarcate the position of demarcation (delimitation) of the phrase, a flag for instructing not to perform the demarcation of the phrase if such a kana character string is present. Put on. For this reason, words registered in the user dictionary can be less likely to be confused with clauses having different meanings when converting into kana-kanji.
Of course, such non-dividable flags are also attached to words stored in the normal dictionary other than the words registered in the user dictionary.

The present invention has been made in consideration of the above circumstances, and if a word registered in a dictionary (ordinary dictionary or user dictionary) has the above kana character string, it will not be shared. Register with a flag to instruct. By this flag, a kana-kanji conversion processing device capable of improving the accuracy of kana-kanji conversion during a solid (kana character string before segmentation) batch conversion process is provided.

[0012]

FIG. 1 is a block diagram showing the basic configuration of the present invention. As shown in FIG. 1, a kana-kanji conversion processing device having a continuous phrase conversion function of the present invention has a key input means 101 and a display means 102 having a display screen.
And an input character storage means 103 for storing the kana character string input from the key input means 101, and a phrase segmentation means 104 for recognizing this kana character string into a kana character phrase.
And a kana-kanji conversion dictionary 1 that stores the kanji for converting this kana character clause into kanji or kanji-mixed sentences.
05, conversion means 106 for converting kana characters stored in the input character storage means 103 into kanji or kanji mixed sentences based on the kana-kanji conversion dictionary 105 according to an instruction from the key input means 101, and the key input means. 101
The Kana-Kanji conversion dictionary 105 is used to specify a particular Kana character clause and the corresponding Kanji or Kanji-mixed sentence.
And specific phrase conversion dictionary 1 registered by the conversion means 106
07, and if there is a specific kana character phrase registered in the specific phrase conversion dictionary 107 in the kana character string stored in the input character storage means 103, do not know this specific kana character phrase. A phrase control unit 108 for controlling the phrase segmentation means 104 and for converting into a kanji character or a kanji mixture sentence corresponding to the specific kana character segment.
The kana-kanji conversion processing device further comprises an output unit 109 for outputting the character string stored in the input character storage unit to the display screen of the display unit 102.

In the present invention, the key input means 10
As 1, a keyboard device, a tablet device, or an input device such as a pointing device is used.
As the display means 102, a CRT display device, L
A CD (liquid crystal display device) or an EL display device is used. Input character storage means 103, phrase segmentation means 104, kana-kanji conversion dictionary 105, conversion means 106,
As the specific phrase conversion dictionary 107, the phrase control unit 108, and the output unit 109, it is convenient to use a microcomputer system including a CPU, a ROM, a RAM, and an I / O port. In particular, a RAM is usually used as the input character storage means 103, and the kana-kanji conversion dictionary 105,
An external storage device is used as the specific phrase conversion dictionary 107.

[0014]

According to the present invention, as shown in FIG. 1, in a kana-kanji conversion processing device having a continuous phrase conversion function, the key input means 101 inputs a kana character string for document creation and performs specific phrase conversion. The kana-kanji conversion dictionary 105 and the conversion means 106 register a specific kana character clause and a kanji or kanji mixture sentence corresponding thereto in the dictionary 107. The display means 102 displays the character string stored in the input character storage means 103 on the display screen. The input character storage means 103 stores the kana character string input from the key input means 101.

The bunsetsu wording means 104 passes the kana character string stored in the input character storage means 103 to a kana character bunsetsu. The kana-kanji conversion dictionary 105 has an input character storage unit 1
Kanji for converting each kana character clause of the kana character string stored in 03 into a kanji or kanji mixture sentence is stored. The conversion means 106 converts the kana characters stored in the input character storage means 103 into kanji or kanji mixed sentences based on the kana-kanji conversion dictionary 105 according to an instruction from the key input means 101.

In the specific phrase conversion dictionary 107, a specific kana character phrase and the corresponding kanji or kanji mixed sentence are operated by operating the key input means 101, and the kana-kanji conversion dictionary 1
05 and the conversion means 106 to register. If there is a specific kana character clause registered in the specific phrase conversion dictionary 107 in the kana character string stored in the input character storage means 103, the phrase control means 108 will know the specific kana character clause. In addition to controlling the bunsetsu wakachi means 104, it converts into a kanji or kanji mixed sentence corresponding to this particular kana character bunsetsu. The output unit 109 outputs the character string stored in the input character storage unit 103 to the display screen of the display unit 102.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail based on the embodiments shown in the drawings. The present invention is not limited to this.

FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention. In the figure, reference numeral 1 is a keyboard provided with kana character keys, numerical keys, conversion keys, and function keys, and an input operation for kana character string and kana-kanji conversion processing is performed from this keyboard 1. Reference numeral 2 denotes a control unit, which comprises a microprocessor for executing the control program, a control program for executing the kana-kanji conversion processing of the present invention, a ROM (read only memory) such as a table, and an external storage device (not shown). Has been done.

Reference numeral 3 denotes a display unit for displaying the data converted into the inputted kana character string or kana-kanji mixed character string,
A CRT display device or LCD (liquid crystal display device) is used. Reference numeral 4 is a normal dictionary including independent words, proper nouns, affixes, etc., including a meaning code. Reference numeral 5 is a co-occurrence dictionary (composite example dictionary) that stores the co-occurrence (example) relationship between clauses. Reference numeral 6 is a user dictionary in which the reading of the words registered by the user, the kanji notation, the part of speech / classification, and the indivisible flag are registered. ROMs / RAs for 4, 5 and 6 dictionaries
M and an external storage device (not shown).

Reference numeral 7 is a buffer, which is composed of a character buffer 71, a learning buffer 72, and a homophone word buffer 73. The character buffer 71 is a buffer for storing a character string input from the keyboard 1. Learning buffer 72
Is a buffer for learning words. Homophone buffer 7
Reference numeral 3 is a buffer for storing homophones in the Kana-Kanji conversion process.

The kana character string input from the keyboard 1 is stored in the character buffer 71. Next, when the user gives a conversion instruction or when it comes to the conversion timing, the control unit 2 refers to the normal dictionary 4, the co-occurrence dictionary 5, the user dictionary 6, and other tables to perform the kana-kanji conversion. The conversion result can be stored in the homophone word buffer 73.
Candidates other than the first candidate can be displayed on the display unit 3 and selected by operating the keyboard 1. When a homophone is selected by operating the keyboard 1, the result is stored in the learning buffer 72.

FIG. 3 is an explanatory view showing a dictionary creation screen of the user dictionary. In the dictionary creation screen as shown in FIG. 3, an example of registering in the user dictionary, such as headline = “Shoutokutaishi”, post-registration = “Shotoku Taishi”, part of speech = noun, classification = 1 (person) ... ing. Here, in order to set “Shoutokutai” = “Shotoku Taishi” (person) (undividable), an indivisible X (flag) is specified after the noun (unique) + classification (person).

FIG. 4 is an explanatory view showing the storage area of the user dictionary. As shown in FIG. 4, the reading, notation, part of speech, classification, and division of the words registered in FIG. 3 are stored in the storage area of the user dictionary.

FIG. 5 is a flowchart of the kana-kanji conversion processing of this embodiment. As shown in FIG. 5, the input kana character string (solid) and the reading “Shoutokutaishitai” will be described as an example along the flow of the processing.

The kana character string "Shoutokutai ni ita" was fetched into the character buffer 71 (step 501).
Next, the preceding clause is extracted from the kana character string of "Shoutokutai ni ita" (step 502). Step 5
In 02, "Shoutokutai", "Shoutokutai", "Shoutoku", "Shoto", "Sho",
“Sho” is extracted as the preceding clause. Following the preceding clause extracted in step 502, the next clause (successor clause) is extracted (step 503).

If the preceding phrase is "Shoutokutaishi ni",
Subsequent clauses are "ta". If the preceding phrase is "Shoutokutaishi", the subsequent phrase will be "Nitani". If the preceding clause is "shoutoku", the look-ahead clause is "taishi" or "tai". If the preceding phrase is "Shouto", the subsequent phrase is "Kutai". If the preceding phrase is "sho", the subsequent phrase will be "Tokutai". If the preceding clause is "sho", the following clause is "utoku". In this way, when the subsequent clause is extracted according to the preceding clause, the process proceeds to step 504.

Next, the relationship between the preceding clause and the succeeding clause is examined, and the optimum position is determined by the delimiter position between the preceding clause and the succeeding clause (step 504). Since the connection between two Kanji characters was emphasized, there was a possibility that the words "Shoutoku" and "Taishi" would be broken. In the present invention, the character string starting with "Shoutokutai" has a flag attached to it so that the phrase is not compulsorily cut. There is.

Therefore, the phrase is "Shoutoku taishi ni" +
There are two ways of sharing the word, "I went" or "I went to" + "I went to", but I can't find a candidate for "I went" in the dictionary, so "I want to go to" + "I" is selected. In step 505, it is judged whether or not the phrase segmentation processing of the input kana character string is completed. If not completed, the process returns to step 503, and if completed, the process proceeds to step 506.

By repeating the process of LOOP1, the phrase segmentation process of the input kana character string is completed. The candidates divided into phrases are stored in the homophone buffer 73 as a result for each phrase. The homophone buffer 73 has a separate buffer for each phrase. In this case, the subsequent clause "I"
"I said,""Ineeded,""Iwent,""Entered,"
Candidates such as "roasted", "roasted", "passed", and "said" are stored.

The homophone selection process of step 506 narrows down homophone candidates divided for each phrase. In this process, although there are multiple candidates for "I", "said" is selected based on the co-occurrence relationship (person + + said). LOOP
By repeating the processing of 2, the kana-kanji conversion processing of the input kana character string is completed (step 507). In this way, "I went to Shotokutai" is converted to "I told Prince Shotoku."

[0031]

According to the present invention, a word that is registered in a dictionary (normal dictionary or user dictionary) is attached with a non-dividable flag that is a segment break, and is registered. The kanji conversion accuracy can be improved.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a dictionary creation screen of a user dictionary.

FIG. 4 is an explanatory diagram showing a storage area of a user dictionary.

FIG. 5 is a flowchart of kana-kanji conversion processing according to this embodiment.

[Explanation of symbols]

 1 keyboard 2 control unit 3 display unit 4 normal dictionary 5 co-occurrence dictionary 6 user dictionary 7 buffer

Claims (1)

[Claims] 1. A kana-kanji conversion processing device having a continuous phrase conversion function, key input means, display means having a display screen, and input character storage means for storing the kana character string input from the key input means. Kana-kanji conversion dictionary that stores this kana character string into kana-character bunsetsu and kana to convert this kana-character bunsetsu into kanji or kanji-mixed sentence, and instructions of the key input means. A conversion means for converting the kana character stored in the input character storage means into a kanji or kanji mixed sentence based on the kana-kanji conversion dictionary, and a specific kana character phrase corresponding to this by operating the key input means. A kanji or a mixture of kanji characters is stored in the kana-kanji conversion dictionary and a specific phrase conversion dictionary registered by the conversion means, and in the input character storage means. If there is a specific kana character phrase registered in the specific phrase conversion dictionary in the specified kana character string, the phrase segmentation means is controlled so as not to know this specific kana character segment, and A clause control means for converting into kanji or a mixture of kanji characters corresponding to a kana character clause, and an output means for outputting the character string stored in the input character storage means to the display screen of the display means are provided. Kana-Kanji conversion processing device.