JPS613268A - Kana and kanji conversion processor - Google Patents

Abstract

PURPOSE:To prevent an improper candidate from being generated by learning as to every composite word in addition to each word. CONSTITUTION:An input part 1 is equipped with a KANA (Japanese syllabary) key group of the Japanese syllabary, a conversion key for making a KANJI (Chinese character) conversion indication, etc. A CPU2 sends information on a KANA character string inputted from the input part 1 to a KANA-KANJI conversion part 3 according to a specific processing program. The KANA-KANJI conversion part 3 retrieves a KANA-KANJI conversion dictionary 4, word learning buffer 5, and composite word learning buffer 6, and sends information on the KANA character string to the CPU2 in the form of information on a KANJI or KANA and KANJI mixed character string. Then a display part 7 displays the KANA character string inputted to the input part and the KANJI or KANA and KANJI mixed character string obtained by converting said KANA character string into KANA and KANJI.

Description

[Detailed Description of the Invention] Technical Field> The present invention relates to the kana-kanji conversion process of Japanese word processors, sessa, etc., which inputs kana characters from a kana keyboard according to the pronunciation of the characters and outputs kanji or sentences containing kanji and kana. Regarding equipment.

<Prior Art> In a kana-kanji conversion processing device, homophone and allograph candidates are generated when an input kana character string is converted into a kanji character string or a character string containing kana-kanji characters. Conventionally, methods for registering (hereinafter referred to as learning) specific words for homophones during conversion include ``frequency learning'' and ``latest usage learning.''

However, in these studies, learning word by word has a negative effect on the kana-kanji conversion of compound words.Here, compound words are words that are broken down into two different words depending on how they are separated in a specific kana character string. If you perform kana-kanji conversion based on the learning results, meaningless kanji or combinations of character strings containing kana-kanji will be created. For example, when searching for the kana-kanji character string ``Itadakimasta'' for the kana character string ``Itadakimasta'', if ``Mashita'' has been learned, meaningless combinations such as ``Top'' and ``Mashita'' will be generated. I can do it. On the other hand, when converting words from kana to kanji, words learned during compound word conversion become inappropriate candidates.

<Object of the Invention> The present invention has been made in view of the above circumstances, and its purpose is to develop kana that avoids the generation of inappropriate candidates by learning compound words separately from word units. An object of the present invention is to provide a kanji conversion processing device.

<Structure of the Invention> The kana-kanji conversion processing device according to the present invention is provided with a learning buffer in which learning results of compound words consisting of a combination of a plurality of words are registered, and in kana-kanji conversion, the combination of words is treated as one set. Learning is performed, and in subsequent kana-kanji conversion, a kanji string corresponding to the information registered in the learning buffer is selected as a candidate, giving priority to word-by-word learning.

<Example> An embodiment of the present invention will be described below.

FIG. 1 shows the configuration of a kana-kanji conversion processing device. The input unit 1 includes a group of kana keys of 50 syllables, and a conversion key for instructing kanji conversion. The CPU 2 sends information on the kana character string manually entered by the input unit 1 to the kana-kanji conversion unit 3 according to a predetermined processing program. The kana-kanji conversion unit 3 is
The input kana character string is searched in the kana-kanji conversion dictionary 4, the word learning buffer 5, and the compound word learning buffer 6, and the information of the kana character string is converted to the information of kanji or a character string containing kana-kanji. and sends it to CPU2. The display unit 7 displays the kana character string input into the input unit 1 and the kanji or character string containing kana-kanji characters obtained by converting the kana character string into kana-kanji characters.

FIG. 2 shows the contents of the word learning buffer 5, which stores the number assigned to the learned word in the kana-kanji conversion dictionary 4.

FIG. 3 shows the contents of the compound word learning buffer 6.
This compound word learning buffer 6 stores the numbers assigned in the kana-kanji conversion dictionary 4 to the words constituting the learned compound word. The numbers of a plurality of words constituting this compound word are separated by a separator in units of compound words, and in the example shown in FIG. 3, the separator is "0".

Now, the kana character string entered manually from the input section 1 is displayed on the display section 7. Then, when the conversion key of the input section 1 is pressed, the CPU 2 converts the input kana character string to the kana-kanji conversion section 3.
The kana-kanji conversion unit 3 searches the kana-kanji conversion dictionary 4 to create a kanji or a character string containing kana-kanji.

The kana-kanji conversion unit 3 checks whether the received kana character string can be converted into a kanji string with one word (including inflections and attached words), and if it can be converted into a kanji string with one word, it converts the kanji string into a kanji string. are candidates. In this case, if there is a homophone word and there are two or more candidates, the word learning buffer 5 is referred to and kanji or kana-kanji mixed character strings are learned for the sent kana character string. In this case, that kanji or character string containing kana-kanji is given priority over other kanji or character strings containing kana-kanji.

The kana character string sent cannot be converted into a kanji string in one word6)
In this case, the kana-kanji conversion unit 3 checks whether a kanji string can be formed using two or more words (suffixes and prefixes are counted as one word). When a kanji character string can be formed using two or more words, the compound word learning buffer 6 is referred to and a check is made to see if a combination of two or more words has been learned for the kana character string.

If the kanji string has been learned, that kanji string is given priority over other kanji strings as a candidate. If a combination of two or more words has not been learned, refer to the word learning buffer 5 word by word,
If the word has been learned, that word is given priority over other words as a candidate.

Candidate kanji or character strings containing kana and kanji are sent to CPU2.
The CPU 2 displays these kanji characters or character strings containing kana and kanji characters on the display unit 7 in descending order of priority.

The word learning buffer 5 and the compound word learning buffer 6 are referred to by searching from before the learning end pointer EP to the learning start pointer SP, as shown in FIGS. 2 and 3.

When an unlearned kanji or a kana-kanji mixed character string is selected for the input kana character string, the selected character string is registered in the word learning buffer 5 or the compound word learning buffer 6. Also, when the input part 1 is selected as the next candidate and the next kana character is entered manually, ■ If it is a word, it goes to the word learning buffer 5, and if it is a combination of two or more words, it goes to the compound word learning buffer 6. be registered. When the word learning buffer 5 and the compound word learning buffer 6 are full, the oldest word or word combination is erased, the learning start pointer SP is advanced, and a new word or word combination is registered at that position.

<Effects of the Invention> As explained above, in the present invention, learning is performed in compound word units in addition to word units, so that inappropriate candidates are not generated by learning in kana-kanji conversion. Furthermore, if one of the word-based and compound word-based learning is applicable, the other is given priority, so the learning results of the other are not wasted.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing the contents of a word learning buffer, and FIG. 3 is a diagram showing the contents of a compound word learning buffer. 1... Input unit 2... CPU 3... Kana-kanji conversion unit 4... Kana-kanji conversion dictionary 5... Word learning buffer 6... Compound word learning buffer

Claims (1)

[Claims] (1) Kana that has a kana-kanji conversion dictionary that stores in advance kanji corresponding to kana character strings or character strings containing kana-kanji mixtures, and converts character strings input in kana to kanji or character strings containing kana-kanji mixtures. The kanji conversion processing device is equipped with a learning buffer in which the learning results of compound words made up of combinations of multiple words are registered, and in the kana-kanji conversion, the combination of words is learned as one set, and in the subsequent kana-kanji conversion. A kana-kanji conversion processing device characterized in that a kanji string or a kana-kanji mixed character is selected as a candidate in accordance with information registered in the learning buffer, giving priority to word-by-word learning.