JPH0429286A - Japanese pronunciation training machine - Google Patents

Abstract

PURPOSE:To obtain this Japanese pronunciation training machine suitable for the learning of correct Japanese pronunciation by imparting reading in another language to KANA (Japanese syllabary) characters to register them and confirming KANA characters for which reading is given and displaying them on a display screen. CONSTITUTION:The function is provided which displays sentences in HIRAGANA (cursive form of Japanese syllabary) and sentences in another language like Roman character on the display screen as reading of Japanese- language sentences where KANJI (Chinese characters) and KANA characters are mixed, and a designated KANA character is selected from a KANA character dictionary 35 by a KANA character selecting means 3a, and reading in another language is given to this KANA character and they are registered. A registered reading confirmation means 3d confirms the reading of registered HIRAGANA by comparison and reference on the display screen between selected HIRAGANA and HIRAGANA having resembling sounds easy to mistake. Thus, the objective Japanese pronunciation is easily trained.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a Japanese pronunciation training machine, and particularly to a Japanese pronunciation training machine suitable for learning correct pronunciation of Japanese.

[Conventional technology] Traditionally, foreigners whose native language is not Japanese,
When reading and pronouncing sentences written in Japanese, it is often difficult to know how to pronounce the various characters such as kanji, hiragana, and katakana.

Therefore, in order to deal with such cases, the applicant first listed sentences containing kanji and kana, and displayed their pronunciations in hiragana and romaji, thereby contributing to Japanese pronunciation practice by foreigners. We proposed the configuration of a Japanese pronunciation practice machine that would do the following.

[Problem to be solved by the invention] However, in the previously proposed Japanese pronunciation practice device, even if the reading of a sentence containing kanji and kana is displayed in hiragana and romaji, it is relatively troublesome to display it in romaji. There are also concerns that even though the Roman alphabet is the same, it is read differently in countries such as the United States, Australia, and China, or people in regions who do not use the Roman alphabet as their native language. In some cases, he was unable to practice pronunciation because he did not understand the Roman alphabet itself.

This invention was made to improve these conventional problems, and its purpose is to help users who want to practice their pronunciation, i.e. foreigners and other users (hereinafter simply referred to as "users"). ) automatically assigns the native language reading to hiragana, and allows this to be output and displayed. The purpose of this invention is to provide this kind of Japanese pronunciation practice machine.

[Means for Solving the Problem] In order to achieve the above-mentioned object, the Japanese pronunciation training device according to the present invention changes the pronunciation display of Japanese sentences containing kanji and kana to the pronunciation display sentences in hiragana and Roman characters, etc. In a Japanese pronunciation training machine that has the function of displaying pronunciations and sentences in other languages on the display screen, there is a dictionary of kana characters that has at least information on headings for hiragana, phonetic characteristics, and the display screen position. , a kana character selection means for selecting the designated kana character from the kana character dictionary; and a kana character selected by the kana character selection means to be registered with a "yomi" in the other language. a reading registration means for confirming the reading, a registered reading confirmation means for checking the kana characters to which the reading registration means has assigned a reading, and a reading confirmation means for outputting the reading confirmed by the registered reading confirmation means and displaying the reading on the display screen. Further, the output means has a function of outputting the sound of the "yomi" of the confirmed kana character in conjunction with the display of the "yomi" of the confirmed kana character. .

[Function] Therefore, in the Japanese pronunciation training machine of the present invention configured as described above, each information on the heading of the hiragana, the phonetic characteristics, and the position on the display screen of the hiragana, such as the 50-on diagram, is displayed. A dictionary of kana characters is prepared, and the user can use the kana character selection means to display the hiragana table, such as a 50-on diagram, on the display screen. When adding , the pronunciation registration means first registers the "pronunciation" of the hiragana in the Japanese vowel set, and then the vowel part is displayed in other lines, leaving room for the corresponding consonant, and then, By registering the selected consonants, it is possible to easily register the "reading" of the hiragana using these consonant and vowel parts.

The registered pronunciation confirmation means allows users to compare and refer to the selected hiragana and hiragana that have similar sounds that are easy to confuse on the display screen and, if necessary, through voice output. You can check the "reading" of the registered hiragana using your eyes and ears.

[Embodiment] Hereinafter, an embodiment of the Japanese pronunciation training machine according to the present invention will be described in detail with reference to FIGS. 1 to 7.

FIG. 1 is a perspective view showing the overall structure of a Japanese pronunciation training machine to which an embodiment of the present invention is applied, and FIG. 2 is a functional block diagram showing the detailed structure of the central processing section.

In the embodiment configuration shown in FIG.
are input device 17 display device 2. and a central processing section 3.

The input device l is a keyboard on which the user manually presses the keys;
It consists of a voice input device, a character recognition device, etc.

The display device 2 has a display screen 21 such as a CRT,
It also has a built-in audio output device. On the display screen 21, a table of hiragana such as a 50-on diagram is displayed, and the "reading" given by the user can also be referred to at the same time. Displays hiragana reading character strings, user reading character strings, etc.

As shown in FIG. 2, the central processing unit 3 includes a hiragana reading registration unit 31. Reading generation unit 32. and each processing system of the display/output unit 33, and a corresponding user registration memory 34. Hiragana dictionary 35. and compound clause conversion dictionary 3
It consists of 6.

Here, in the hiragana reading registration unit 31, when the user wants to add "yomi" and selects a hiragana consisting of After extracting the hiragana, an area for "reading" corresponding to the hiragana is secured in the user registration memory 34.

Further, the reading registration means 3C has this user registration memory 3.
4. Write the input user's "reading" in the set area above.

Furthermore, the registered pronunciation confirmation means 3d performs various functions such as checking and confirming the already registered "pronunciation" and the newly registered "pronunciation", and comparing the "pronunciation" of hiragana with similar sounds. Checks whether a "yomi" suitable for the characteristics of the character has been registered.For example, during the reading registration,
If you try to give the same reading to hiragana with different phonetic attributes, a confirmation message will be displayed.

The pronunciation generation unit 32 generates a hiragana character string from the kanji-kana mixed sentence using the multi-phrase conversion dictionary 36, and then generates a hiragana character string from the hiragana dictionary 35 to which the user's “yomi” has been assigned. is used to generate the user's reading string from the hiragana string.

The display/output section 33 displays a table of hiragana and "yomi" such as a 50-on diagram from the hiragana dictionary 35 and the user registration memory 34, and outputs the sounds of the selected hiragana. , the processing results of the pronunciation generation unit 32, that is, the kanji/kana mixed sentences, the hiragana pronunciation character strings, and the user pronunciation character strings are displayed.

As described above, for any kanji/kana mixed sentence,
It is possible to automatically add the user's reading.

Note that detailed explanation of the reading generation section 32 will be omitted.

The structure of the hiragana dictionary 35 will be described with reference to FIG.

The hiragana dictionary 35 includes hiragana headings 511 position information 52. Audio information 53. and a registered reading 54 registered in the user registration memory 34.

The hiragana heading 51 is literally a hiragana heading.

The position information 52 is used when displaying a table such as a 50-on diagram (hereinafter referred to as a 50-on diagram, although it is not necessarily accurate from a Japanese linguistic standpoint) on the display screen 21. "line"
and “column” information. That is, 5. For example, the hiragana ``ka'' is a ``power row array''. In this case, the row information is 521 rows, 7 columns, and 522 columns.

The audio information 53 includes output data 531 for outputting the reading of each hiragana to the audio output device, and consonant information 532 for phonetic pronunciation classification. It consists of vowel information 533. The consonant information 532 includes voiced and unvoiced classifications 5321. Articulation point 5322. and articulation method 5323. Further, the vowel information 533 is [a
One of the following information is entered: l, riJ, lu], re], (o].For example, if the hiragana is "ka", the consonant information 532 is "voiceless/velar/stop sound", vowel The information 533 becomes [al.

Here, the position information 52. and the audio information 53 serve as search conditions for extracting hiragana by the hiragana selection means 3a. In other words, 1. For example, when the user selects "power running", "ka, ki, (, ke, ko)" which has the position information 52 of "power running" in the hiragana dictionary 35 is extracted.Also, similar sounds are extracted. When extracting, to select that line, use the audio information 5
Hiragana that differs in only one attribute are extracted from among the 3r/I4 consonants in 3. For example, "force running" (consonant information 5
32; voiceless, soft palate, stop consonant), there is only a difference between voiceless and voiced, and there is a similar sound, ``G line J (consonant information 532;
This includes selecting voiced, soft palate, and stop sounds. Further, when selecting one hiragana, consonant information having the same vowel (different in only one attribute out of three attributes) is extracted.

The registered reading 54 is a “reading” registered by the user, and is recorded in the rewritable user registration memory 34.

Next, the process when the user registers a "yomi" in the hiragana dictionary 35 will be explained with reference to the flowchart shown in FIG.

When the user attempts to add a "yomi" to a hiragana and selects a hiragana (step S-101, hereinafter referred to as S), the hiragana selection means 3a first selects the hiragana and hiragana selected by the user. Perform matching with the hiragana heading 51 of the dictionary 35 and extract the corresponding hiragana code (S-102), and set a "yomi" area corresponding to the hiragana in the user registration memory 34 (S-103)
.

Next, when "reading" is input (S-104), the reading registration means 3c stores this in the area set in the user registration memory 34 (S-105).

The registered reading confirmation means 3d searches the user registration memory 34 to see if the "reading" registered here overlaps with the already registered "reading" (S-
106, S-107). If we have the same “reading”,
The corresponding hiragana is displayed (S-108) to confirm whether or not to be corrected (S-109). If the input "yomi" is correct, the registration is completed.

Then, in this registered reading confirmation means 3d, the registered “
It is also possible to check with your ears whether the notation of "Yomi", that is, the registered reading, matches the sound of the hiragana, and this process uses the output data 531, which is audio information in the hiragana dictionary 35. This can be realized by outputting it to an audio output device.

Furthermore, the registered pronunciation confirmation means 3d can compare and confirm hiragana that have similar sounds. This is relatively easy for users to make mistakes when practicing pronunciation.
This process is used to check hiragana that have very similar sounds (that is, have similar sounds). It should be noted that the similar sounds referred to here are the three attributes of consonant information among the hiragana in the hiragana dictionary 35, that is, voiced and unvoiced categories 5321. Articulation point 5322. and articulation method 5323 indicates a different hiragana.

Next, the process for comparing and confirming hiragana having similar sounds will be explained with reference to the flowchart shown in FIG.

The user selects hiragana character by character or line by line (see the character cursor and line cursor in FIG. 6, which will be described below). When Hiragana is selected (S-201)
, starts the process of searching the hiragana dictionary 35 (S-20
2).

If there is no consonant, in this case, the hiragana is "A, I, U, E, O" which has only vowels, so in this case, it is displayed that there is no similar sound (S-203) .

Also, if there is a consonant, it is checked whether the selected character is one character (S-2043゜If it is one character,
Match the vowel information (S-205), extract hiragana with the same vowel (S-206), proceed to match the consonant information (S-207), and if it is not one character, it will be one line. Therefore, the consonant information is matched without checking the vowel information (S-2071, 3
Among the attributes, hiragana that has only one different attribute are extracted (S-208) and displayed (S-2
09).

Next, a specific example of registering the "reading" will be described.

FIG. 6 shows an example of the user's procedure and system processing during pronunciation registration, as well as the display screen. This specific example is the Japanese vowel "A" - "A".

1 reading of ``i, u, e, o'' is ra, i, u.

After registering e and oJ, continue with "power running" - "ka.

This is a case of registering the ``pronunciation'' of ``ki, <, ke, ko''.

Furthermore, as mentioned above, the phonetic characteristics of Japanese hiragana are that each character includes a vowel, and the individual hiragana that have the same column information in the hiragana dictionary 35 have the same vowel. can. Therefore, in this case, this characteristic is effectively utilized to automatically insert each vowel corresponding to the "yomi" on the syllabary into the "yomi" of the other consonants. In other words, for "ka, sa, ta, na, ha, ma,..." which has the position information of "array",
For example, add the "pronunciation""*a" and do the same for the "I column""ki, shi, chi, ni,..."
*i", r columns" (,su,tsu,nu.

By adding r*uJ, ... to "...", the user can input the consonant character corresponding to "*" for each hiragana in each line. This allows you to automatically add the required reading.

In FIG. 6, first, 1. The user attempts to register a "yomi" and displays the hiragana reading registration screen (1-a).

The system displays a range selection cursor menu as a means for selecting hiragana on the 50-on diagram displayed on the display screen 21. The types of range selection cursors are row cursor, column cursor, and character cursor (
1-b).

2. In this case, for example, when the line cursor is selected, a line cursor that moves line by line on the 50-on diagram will appear, and even if it is at the corresponding "yomi" position, the line cursor will move in line with this. A line cursor appears (2-a).

3. Next, when the user selects "A line", the system highlights the hiragana and reading area for "A line" (3-a), and displays the hiragana reading input screen for "A line". is displayed (3-b), and the dictionary search means 3b sets a reading registration area for the code having the position information of "A line" in the hiragana dictionary 35 in the user registration memory 34 (3-c). ).

4. After that, when the user inputs the "yomi", the system registers the hiragana and "yomi" of "A line" in the set reading registration area of the user registration memory 34 (4-
a), and for the “yomi” of each hiragana in other lines,
``*J'' is inserted in the area where the ``yomi'' of ``A'' and the consonant immediately before it are inserted (4-b).

Notte is the Japanese vowel "A row" → "Ah,"
The “pronunciation” of “i, u, e, o” is ra, i.

are registered as u, e, oJ, and at the same time, the other lines "i,
The “pronunciations” of “tsu,...” are respectively [*a, *i
, *u, *e+ *QJ will be inserted. Subsequently, 5. When the user attempts to input the "yomi" of a line and selects "power running", the hiragana and reading area of "power running" are displayed in reverse (5-a).

6. Next, by inputting rkJ, which is the consonant of "force running", r*a, *i, *u are already included in the "force running" - "kaki (, ke, ko)".

*6. Since *QJ is inserted, "*" is replaced with "k" and becomes rka, ki, ku.

The readings for "power running" such as ke and koJ are automatically created (6-a).

In other words, here, as expected, the Japanese vowels "A" → "A, I, U, E, O" are "pronounced" as ra, i,
In addition to being registered as u, e+ oJ, "power running" → "ka,
The "pronunciation" of "ki, (, ke, ko") is registered. By effectively utilizing the phonetic characteristics of hiragana in this way, it is extremely easy to register the hiragana reading, and the user's work is simplified. This can reduce the burden and improve work efficiency.

Also, lines with other points to note regarding pronunciation.

For example, if there are hiragana characters with different consonants in the same line, this will be displayed and confirmed. I will discuss this point.

FIG. 7 is an example of a message screen when there are hiragana characters with different consonants in the same line.

For example, in the case of ``Hego J → ``Ha, hi, fu, he, wa J,''
Each consonant in "ha, he, ha" is a "voiceless, glottal, fricative", the consonant "hi" is a "voiceless, velar, fricative", and the consonant "fu" is a "voiceless, glottal, fricative". unvoiced·
"bilabial fricative" and has different consonants in the same line.

Here, the "Ha line" is rha, hi, hu, he, ho.
After registering J and “reading”, regarding “rhi, hU”,
These have different consonants, but the reading can be left as is, but please check by sending a message. If the user wants to change “fu”-rhul to rfuJ,
This can be easily rewritten. In this case, ``hi'' and ``fu'' have different consonants, but they can be expressed in different ways. You can make this fact known by putting a fence around the area.

In other words, by using the hiragana dictionary with the user's readings created in this way, it is possible to provide a pronunciation practice system that automatically assigns the user's ``readings'' to sentences containing kanji and kana. be.

In the above description, as an example, the alphabet is used as a reading character string (in this case, the Roman alphabet reading), but it is also possible to form a reading character string in other languages such as Hangul.

[Effects of the Invention 1] As detailed above, the Japanese pronunciation training device according to the present invention can display the headings of hiragana, their phonetic characteristics, and the positions of hiragana on the display screen, such as the 50-on diagram. Using a dictionary of kana characters containing information, the user selects the "yomi" of the hiragana using the kana character selection means while displaying a table of hiragana, such as a 50-on diagram, on the display screen.
In order to add the desired hiragana, the pronunciation registration means registers the ``yomi'' of the hiragana in the Japanese vowel set, and the desired consonants are registered in the other rows. The "pronunciation" is obtained, and the registered pronunciation confirmation means compares and references the selected hiragana and hiragana with similar sounds that are easy to mistake on the display screen, and if necessary, compares and references by voice output, Users themselves can check the "pronunciation" of the registered hiragana, and when using a hiragana dictionary with user pronunciations formed in this way,
If the user's "reading" can be automatically assigned to a sentence containing kanji and kana, and the user pronounces it in accordance with the "reading," the user can easily practice pronunciation in Japanese.

[Brief explanation of drawings]

Fig. 1 is a perspective view showing the overall structure of a Japanese pronunciation training machine to which an embodiment of the present invention is applied, Fig. 2 is a functional block diagram showing the detailed structure of the central processing section of the same, and Fig. 3 is the hiragana of the above. FIG. 4 is a flowchart showing the process for registering a user's reading in the hiragana dictionary, and FIG. 5 is a process for comparing and confirming hiragana with similar sounds. Figure 6 is an explanatory diagram showing the user's procedure and system processing during reading registration, and an example of the display screen. Figure 7 is the message screen when there are hiragana with different consonants in the same line. It is an explanatory view showing an example. 1... Input device, 2... Display device, 21... Display screen, 3... Central processing unit, 31... Hiragana reading registration unit, 32... Reading generation. 33...Display/output unit, 34...Memory for user registration, 35...Hiragana dictionary, 36...Dictionary for compound clause conversion, 3a...Hiragana selection means , 3b... Dictionary search means, 3c... Reading registration means, 3d... Registered reading confirmation means. Patent applicant: Brother Industries, Ltd.

Claims (2)

[Claims] (1) In a Japanese pronunciation practice machine that has the function of displaying the pronunciation of Japanese sentences with kanji and kana on the display screen, the pronunciation display in hiragana and the pronunciation display in other languages such as Roman characters, a kana character dictionary having information such as at least headings for hiragana, phonetic characteristics, and the display screen position; a kana character selection means for selecting the designated kana character from the kana character dictionary; and the kana character. A reading registration means for registering a kana character selected by the character selection means with a “yomi” in the other language; and a registered reading for confirming the kana character to which a “yomi” has been added by the reading registration means. A Japanese pronunciation training device comprising: a confirmation means; and an output means for outputting the "pronunciation" confirmed by the registered pronunciation confirmation means and displaying it on the display screen. (2) The Japanese pronunciation training device according to claim 1, wherein the output means has a function of outputting the sound of the "yomi" in conjunction with the display of the "yomi" of the confirmed kana character.