JPS60107171A - Voice kana (japanese syllabary) kanji (chinese character) converting system - Google Patents

Abstract

PURPOSE:To realize a voice Kana/Kanji converting system possible for sounding a sentence as a normal reading by referencing a replaced Kana character storage means and an unreplaced Kana character storage means. CONSTITUTION:A voice input means 1 transmits a Kana character string to a dictionary retrieval means 3. The dictionary retrieval means 3 uses the Kana character string as a key to retrieve the dictionary in the dictionary storage means 2. Thus, partial Kana characters unmatched as the result are left. A Kana character replacing means 6 has a function to replace required Kana character strings left as the result of dictionary retrieval into other Kana characters by referencing the replaced Kana character storage means 4 and the unreplaced Kana character storage means 5. To the replaced Kana character storage means 4, ''Kana characters to be replaced'' and ''Kana character after replacement'' are stored. Moreover, Kana characters not to be replaced among Kana character strings including Kana characters being object of replacement are stored in the unreplaced Kana character string storage means 5.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a phonetic kana-kanji conversion method for obtaining a character string mixed with kanji, kana, and kanji from pronounced Japanese. This article concerns a method for obtaining the correct notation from the pronunciation of the reading, even for single-junctions that have different readings and notations.

The phonetic kana-kanji conversion method is a method that allows Japanese sentences containing thousands of types of kanji to be input into information processing equipment such as computers by simply saying their readings aloud. , users can input all Japanese characters with almost no training. However, there are many Japanese sentences that have different characters and readings. For example, the ``ha'' in ``Washiwa'' is pronounced as ``-ゎ''.

Therefore, conventionally, instead of "wa""taJ rLJ rwa", "wa" [taJ rLJ r is like j, commonly written υ
I had to pronounce it using the reading. Also, in order to avoid this inconvenience, if you use a method that automatically converts "wa" to "ha", "wa", "ta", "shi", "wa" will become "ha".
It ends up being converted into ``ta'', ``shi'', and ``ha''.

SUMMARY OF THE INVENTION An object of the present invention is to provide a phonetic kana-kanji conversion method that eliminates the above-mentioned drawbacks and allows a user to pronounce entire sentences as they are normally read.

According to the present invention, in a phonetic kana-kanji conversion method for obtaining a character string containing kanji and kana from pronounced Japanese, a plurality of voice input means that perform voice recognition on a pronounced Japanese sentence and output a kana character string; a dictionary storage means for storing a dictionary having at least the pronunciation and notation of the word; a dictionary search means for searching the dictionary using the kana character string as a key to obtain the notation; Replacement kana character storage means for storing kana characters that should be replaced with other characters among the kana characters forming the remaining part of the character string;
By referring to the non-replacement kana character string storage means for storing kana character string patterns that are not subject to replacement, the replacement kana character storage means, and the non-replacement kana character H line storage means, all of the remaining portions are stored. Alternatively, by using a kana character replacement means that replaces a part with another kana character string, and a kanji-kana mixed character string forming means that forms and outputs a kanji-kana mixed character string from the above-mentioned notation and the above-mentioned remaining part, the pronunciation of the normal yomitsushi can be achieved. You can get a string of characters mixed with Kanji and Kana.

The present invention will be described below with reference to the drawings and specific examples.

FIG. 1 is a block diagram showing one embodiment of the present invention.

1 is a voice input means for voice recognition of a Japanese sentence and outputs a kana character string; 2 is a dictionary storage means that includes at least the pronunciation and notation of a plurality of words; and 3 is a voice input means for outputting the kana character string. This is a dictionary search method that searches the dictionary using the column as a key to obtain the notation, and 4 should be replaced with another character among the kana characters that make up the remaining part of the kana character string that was not matched. 5 is a replacement kana character string storage means for storing kana character string patterns that are not to be replaced; 6 is the replacement kana character storage means for storing kana character string patterns that are not to be replaced; 7 is a kana character replacement means for replacing all or part of the remaining part with another kana character string (by referring to the means and the non-replaced kana character string storage means); This is a kanji/kana mixed character string forming means for forming and outputting a kanji/kana mixed character string. Here, to get "on the coast," we use "e", "n", and "ga".
Let's take the example of pronouncing ``n'', ``ni'', ``o'', ``ij'', ``te'', and ``wa''.

"E""N""Gaj"N""Ni""O""
"i""te" [Depending on the pronunciation of waj, the voice input means 1 is
The kana character string "Engan tewa" is sent to the dictionary search means 3. The dictionary search means 3 searches the dictionary in the dictionary storage means 2 using this kana character string as a key. Let us assume that we have obtained the notation ``coastal'' which has the reading ``engan''. As a result, the unmatched partial kana character ``deiwa'' is left.

Referring to the replaced kana character storage means 4 and the non-replaced kana character storage means 5, the kana character replacement means 6 replaces necessary kana characters with other kana characters among the kana character strings left as a result of the dictionary search. It has the function of replacing it with The substituted kana character storage means 4 stores "kana characters to be replaced" and "kana characters after replacement". In addition, the non-replacement kana character string storage means 5 stores the kana character among the kana character strings that include the sharp kana character to be replaced (that is, the "kana character to be replaced" in the replacement kana character storage means 4). Those that should not be replaced are remembered.

FIG. 2 is a diagram showing an example of characters corresponding to replacement in the replacement kana character storage means 4, and FIG. 3 is a diagram showing an example of characters in the non-replacement kana character storage means 5. FIG. 4 is a diagram showing an example of the processing procedure of the kana character changing means 6.

When the processing procedure shown in FIG. 4 is applied to the remaining kana character string "Nidewa", the result is as follows.

First, the kana character string in the non-replaced kana character string storage means 5 is
Search for "Nidewa". If the example shown in FIG. 3 is used as the non-substituted kana character string storage means 5, "Ote" will be searched. Next, if the unmatched kana character is the "kana character to be replaced" in the replacement kana character storage means 4, it is replaced with the "replaced kana character". If we use the example shown in Figure 2 as the substituted kana character storage means 4, then ``ni'' is obtained by removing ``te'' from ``deiwa''.
In "wa", "wa" is replaced with "ha". As a result, "dewa" is replaced with "dewa". The kanji-kana-mixed 9-character string forming means 7 forms a kanji-kana-mixed multi-character string from the notation (coastal) obtained by the dictionary search means 3 and the remaining kana character string (inha) that has been replaced by the kana character replacement means 6. and output. As a result, you can get ``on the coast''.

In this way, by using the present invention, even when there is a discrepancy in Japanese pronunciation and writing, you can input Japanese sentences into the normal pronunciation system without being aware of the discrepancy, and create character strings containing kanji and kana. can be obtained. In this example, the dictionary search is performed only once for one kana character string, but for example, for an input such as "Jiyou wa Ushiyo Shiwa", it corresponds to "Jiyouhou" and "Shiyoshi". do"
It is also possible to search for the two words "information" and "processing" and replace the remaining part "wa" with "ha" to obtain "information processing." Furthermore, the processing procedures, drawings, etc. used to explain the present invention are merely examples, and do not limit the scope of the claims.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment for realizing the present invention, FIG. 2 is a block diagram showing an embodiment of replacement compatible characters in a replacement kana character storage means, and FIG. A fourth diagram showing an example of characters in a character string storage means.
The figure is a flowchart showing an example of the processing procedure of the kana character replacement means. In the figure, 1 is a voice input means, 2 is a dictionary storage means, and 3 is a voice input means.
t,,c; 'f-book search means, 4 is fm kana character storage means, 5 is non-replacement kana character storage means, 6 is kana character replacement means, and 7 is kanji-kana mixed υ character string forming means. Figure 1 Otsu 5 72 Figure 3 Figure 1

Claims (1)

[Claims] ■In the phonetic kana-kanji conversion method for obtaining a kanji-kana-mixed character string from Japanese words that have been combined, a voice input means that performs voice recognition on a pronounced Japanese sentence and outputs a kana character string;
A dictionary storage means for storing a dictionary including at least readings and spellings of a plurality of words; and a dictionary search means for searching the dictionary using the kana character string as a key to obtain the spelling. Replacement kana character storage means for storing kana characters that should be replaced with other characters among the kana characters constituting the remaining part of the kana character string, and for storing unusual kana character string patterns to be replaced. Kana characters that replace all or part of the remaining part with another kana character string by referring to the non-replaced kana character string storage means, the replaced kana character storage means, and the unreplaced kana character string storage means. A phonetic kana-kanji conversion method comprising a replacement means and a means for forming a character string containing kanji and kana from the notation and the remaining portion.