JPS60239865A - Retrieving device of dictionary - Google Patents

Abstract

PURPOSE:To process conveniently a long sound in Japanese sound information and simply by converting a long vowel code out of a character string including the long vowel code into a character with priority to retrieve a dictionary. CONSTITUTION:When a certain character string is inputted as sounds and then inputted to a character string conversion part 1, the character string is converted on the basis of the following regulations. Namely, characters other than long vowel codes are not converted. Long vowel codes following the ''a-stage'' are converted into ''a''. Long vowel codes following the ''i-stage'' are converted into ''i''. Long vowel codes following the ''u-stage'' are converted into ''u''. Long vowel codes following the ''e-stage'' are converted into ''i'' as the 1st proposed character and ''e'' as the 2nd proposed character. Long vowel codes following the ''o-stage'' are converted into ''u'' as the 1st proposed character and ''o'' as the 2nd proposed character. Consequently, plural proposed characters are formed and sent to a dictionary matching par 3. The matching part 3 outputs the correct proposed character through a dictionary 2. Consequently, long sounds in the Japanese sound information can be processed conveniently and simply.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a dictionary search device in a kana-kanji conversion device, and more particularly to a dictionary search device that performs appropriate conversion processing on a long vowel character string. be.

Conventional configuration and its problems In recent years, text creation devices such as Japanese word processors have become popular. This type of device generally creates a Japanese sentence by inputting kana characters into a created sentence, and converting this into a character sequence containing kanji for each word. This kana-kanji conversion is performed by searching for character strings that are registered in the word dictionary in advance and include kanji that respectively correspond to the kana character strings. In other words, character strings containing kanji are given corresponding to kana character strings. Therefore, if a kana character string is inputted incorrectly, the kanji conversion will not be performed or an incorrect kanji conversion character string will be output.

Incidentally, recently, a method has been proposed in which Japanese information is input and processed by inputting voice and mode information that specifies the type of character corresponding to the voice.

Many problems arise when inputting languages by voice.

Especially if you enter a string containing long vowel symbols.

The type of kana character to convert the long vowel symbol to when searching the word dictionary has a significant impact on the search time. For example, if the input voice contains the character string ``e'', even if you search a word dictionary using the kana character string in which the long vowel symbol is converted to the kana character ``e'', such a kana character string will not be found. In many cases, it was not available in the dictionary, and it was necessary to search again using a new kana character string converted to a different kana character "i".

OBJECTS OF THE INVENTION An object of the present invention is to provide a dictionary search device that can conveniently and easily process long sounds in Japanese audio information.

Structure of the Invention The dictionary search device of the present invention takes as input a character string including a long vowel symbol, and searches the long vowel symbol after "Adan" for "A" and the long vowel symbol after "Idan" for "A". "stomach".

The long vowel symbol after "Udan" is "U", and the long vowel symbol after "Edan" is "I" or "E".

A character string converter that converts the long vowel symbol next to "odan" into "u" or "o" and outputs it, and outputs it as is when the long vowel symbol is not included, a dictionary, and the character string converter. and a dictionary matching section that searches the character string obtained from the dictionary in the dictionary.This section appropriately converts character strings containing long vowel symbols into written and written expressions, improving the efficiency of kana-kanji conversion. It is possible to raise the value to A.

The explanatory diagram of the embodiment shows a configuration diagram of the dictionary search device in the embodiment of the present invention, and 1 indicates that a character string is input, and if a long vowel symbol is included, it is input according to the rules defined in the present invention. A character string conversion section 2 converts into a character string, a dictionary in which 8 ml of written words are registered, and a dictionary collation section 3 searches the dictionary 2 for the character string obtained from the character string conversion section 1.

The operation of the dictionary search device of this embodiment configured as described above will be described below. For purposes of explanation, long vowel symbols are expressed as long.

If you input "high school student" by voice and it is recognized correctly,
"Kochokochocho" is obtained. When this character string is input to the character string converter 1, the character string is converted according to the following rules.

(1) Do not convert anything other than the long vowel symbol (2) Convert the long vowel symbol following “A” to “A” (3) Convert the long vowel symbol following “I” to “i” (4) ) The long vowel symbol following "tsu-dan" is converted to "u" (four long vowel symbols following "two-dan" are converted to "i" as the first candidate and "e" as the second candidate (6) The long vowel symbol following "o-dan" is converted to "u" as the first candidate and "o" as the second candidate. Therefore, for the input character string "ko-cho-ko-naga-se-cho", the first candidate is "koukosei". ” 2nd candidate ``Kookoosei'' 2nd candidate ``Koukoosei'' 2nd candidate ``Koukousee'' 6th candidate ``Kookoosei'' 5th candidate ``Kookoosee'' 6th candidate ``Koo The eighth candidate "Ko-ko-osee" is generated and sent to the dictionary collation unit 3. Since one Japanese written word is registered in the dictionary 2, the dictionary search unit 3
Then, the first candidate is regarded as the correct answer and the corresponding dictionary contents (for example, a kanji code string) are output.

In the above example, there are eight possible combinations of long vowels, so this ranking greatly changes the dictionary search time. However, in the dictionary search device of the present invention, the search time is minimized by having the above-mentioned long vowel symbol conversion rule. The basis for this is the uneven existence of Japanese words, as shown below.

First, words containing continuous vowels were extracted from 30,000 Japanese words. Here, continuous vowels refer to two or more consecutive vowels in one word or in a combination of words. In this case, it is not limited to the continuation of one independent vowel syllable, but also the continuation of a vowel included in the previous syllable and the vowel syllable that immediately follows it.

For example, "love (ai)" and "nephew (oi)."

``Fish (Uo)'' etc. are continuous vowels, [shellfish (kai month).

r, 11 (: hui)'' and ``sui'' also contain continuous vowels. When there are continuous vowels, there are cases in which each vowel is pronounced in isolation, and cases in which the following vowel is closely connected to the previous vowel and is pronounced as if it were a long vowel. The latter ones are especially called "diphthongs."

As a result of researching words containing continuous vowels among the 30,000 Japanese words mentioned above, the following data was obtained.

■ The number of words in which the character 71J of "edan" + "i" can be pronounced as "echo" in a long consonant manner is approximately 1400% The character string "edan" + "e" can be pronounced as "echo" in a long consonant manner Only four words can be uttered.

■ The number of words in which the character string "Odan" + "U" can be pronounced as "Onaga" is approximately 3,960. The number of words in which the character string "Odan" + "O" can be pronounced long as "Onaga" is about 200.

In the natural language Japanese, there are biased rules such as ■ and ■ above, and the present invention uses the rules (5) based on these rules.
, which has the characteristics of (6).

As described above, according to this embodiment, by using the conversion rule for long vowel symbols, it is possible to effectively and correctly convert a character string including long vowel symbols into kana-kanji.

Effects of the Invention As is clear from the above explanation, the present invention is configured such that, for character strings including long vowel symbols, the long vowel notation t is prioritized, converted into characters, and dictionary search is performed. Because of this, the excellent effect of being able to efficiently obtain appropriate kana-kanji conversion output even for character strings that include long vowel symbols can be obtained.

BRIEF DESCRIPTION OF THE DRAWINGS The figure is a schematic configuration diagram showing the configuration of a dictionary search device according to an embodiment of the present invention. 1...Character string conversion section%2...Dictionary, 3
...Dictionary checking section. Name of agent: Patent attorney Toshio Nakao and one other master swordsman

Claims (3)

[Claims] (1) Input a character string including long vowel symbols, "A" 1
The long vowel symbol next to 0th is "a", the long vowel symbol next to "i-dan" is "i", and the long vowel symbol after "u-dan" is "i".
Change the long vowel symbol after ``u'' or ``edan'' to ``i'' or ``edan''.
Change the long vowel symbol after ``e'' or ``odan'' to ``u'' or ``
a character string converter that converts the character string into "O" and outputs it as is when it does not include a long vowel symbol, a dictionary, and a dictionary matching unit that searches the dictionary for the character string obtained from the character string converter. A dictionary search device consisting of. (2) The dictionary matching section uses the character string conversion section to check the next VC of "edan".
For a character string that has a certain long vowel symbol converted to "i" or "e", first search the dictionary for the character string converted to "i", and only if it does not exist in the dictionary, search for the character string converted to "e" in the dictionary. 2. A dictionary search device according to claim 1, wherein the dictionary search device performs a search using . (3) The dictionary matching unit converts the long vowel symbol next to “odan” into “u” or “o” by the character string conversion unit. 2. The dictionary search device according to claim 1, wherein the dictionary is searched for a character string converted to "o" only when the character string does not exist in the dictionary.