JP2002156897A - Pronunciation learning system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a pronunciation learning system which is capable of more adequately evaluating the pronunciation state over the entire part of model sentences by learners. SOLUTION: When the speech of the time the learner reads aloud the model sentences for learning of pronunciation is inputted, the pronunciation by the learner is subjected to processing (S12) to evaluate the pronunciation in word units and processing (S13) to evaluate the pronunciation by the learner in model sentence units. When the levels of importance in the pronunciation of the words constituting the model sentences are previously set, processing (S15) to change the evaluation points of the pronunciation of the model sentence units is performed according to the levels of importance.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pronunciation learning device suitable for use in learning a foreign language conversation.

[0002]

2. Description of the Related Art In recent years, with the growing need for English conversation and the like, conversational learning of foreign languages such as English has been emphasized. While these conversational learning opportunities are offered on television and radio programs,
A third party cannot evaluate the learner's pronunciation.

Therefore, conventionally, there has been provided a pronunciation learning device having a pronunciation evaluation function for evaluating the pronunciation of a learner. Some pronunciation learning devices of this type evaluate the voice when a learner reads a sentence example in units of words and sentence examples (entire sentence examples), and notify the learner of the evaluation score of the entire sentence example. .

FIG. 8 is a flowchart showing the operation procedure of a conventional pronunciation learning device. First, it is determined whether or not a learner's voice has been input (step S1). After that, when a voice is input, an evaluation score of pronunciation in words is calculated (step S2), and then an evaluation score of pronunciation in sentences is calculated (step S3). Next, it is determined whether or not the evaluation score of the pronunciation of each sentence example is less than a predetermined value (step S4). At this time, if the pronunciation evaluation score of each sentence example is equal to or more than a predetermined value, a message meaning "good pronunciation" is displayed together with the pronunciation evaluation score of each sentence example (step S5). If the evaluation score for each sentence example is less than the predetermined value, a message meaning "repronounce again" is displayed together with the pronunciation evaluation score for each sentence example (step S6), and then the pronunciation evaluation score for each word is determined. Words less than the value are inverted and displayed (step S7). As a result, the learner can check the evaluation of his or her pronunciation and can recognize words that could not be pronounced well.

[0005]

By the way, in the conventional pronunciation learning device, the evaluation points of each word are averaged (simple averaging), and the evaluation points of the respective words constituting the sentence example are compared with the evaluation points of the respective words. Weighted by the number of characters (pronunciation length), averaged (weighted average), and sentence example units (entire sentence examples)
Is evaluated. Therefore, even if a word important in pronunciation (for example, a word whose meaning cannot be conveyed to the other party if it is not correctly pronounced) is not good, if the pronunciation evaluation of other words is good, the whole sentence example is "good pronunciation" In some cases, the evaluation was lacking.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a pronunciation learning device capable of more appropriately evaluating the pronunciation state of a whole sentence example by a learner. It is in.

[0007]

In a pronunciation learning apparatus according to the present invention, an input means for inputting a voice when a learner reads out a sentence example to be learned for pronunciation, and a pronunciation of a word constituting the sentence example. Setting means for setting the importance of the speech, and calculating the evaluation score of the pronunciation based on the result of comparison between the voice data input by the input means and the basic pronunciation data corresponding to the sentence example, and the importance set by the setting means And a pronunciation evaluation means for changing an evaluation point in accordance with the sound quality.

[0008] In the pronunciation learning apparatus having the above-described configuration, the importance of pronunciation is set in advance by the setting means for words constituting the sentence example for the sentence example for which the pronunciation is to be learned. When the pronunciation state when the learner reads out the sentence example is evaluated by the pronunciation evaluation means, the evaluation score of the pronunciation calculated based on the comparison result between the voice data input by the input means and the basic pronunciation data corresponding to the sentence example is obtained. Are changed according to the importance set in advance. Thereby, for example, when the pronunciation of a word set to high importance is not good, or when the pronunciation of a word set to low importance is not good, the importance set for each word is By changing (adjusting) the evaluation point of the pronunciation accordingly, it is possible to appropriately evaluate the pronunciation state of the entire sentence example.

[0009]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is an external view of a pronunciation learning device according to an embodiment of the present invention. In FIG. 1, a display 2 is provided on a front portion of the apparatus main body 1. The display 2 is configured by, for example, a dot matrix type liquid crystal display (LCD). The display 2 displays various information such as a conversation selection screen serving as a pronunciation learning menu, a conversation screen including a sentence example to be learned, a pronunciation evaluation result, and various selection screens.

Along with the display 2, various buttons (cursor button 3, enter button 4, return button 5, listen button 6, talk button 7, power button 7) Button 8) is provided. Further, a head-mounted earphone 10 and a microphone 11 are connected to the apparatus main body 1 via a connection cord 9. The cursor button 3 is used to change the position of a cursor mark (arrow display, bar display, inverted display, blinking display, etc.) on the display 2 and to scroll the display screen up, down, left and right. OK button 4
Is used to switch from the conversation selection screen displayed on the display 2 to the conversation screen indicated by the cursor.

The return button 5 is used to interrupt processing, return from the conversation screen to the conversation selection screen, and the like. The listen button 6 is the basic (example) of the sentence specified by the cursor
Use when listening to audio When the listen button 6 is pressed, the sound that is the basis of the above sentence is output from the earphone 10. The speak button 7 is used to pronounce a sentence example specified by the cursor. After the learner presses the talk button 7, the learner reads the sentence example specified by the cursor toward the microphone 11. Power button 8
Is used to turn the device on and off.
When the power button 8 is pressed when the apparatus power is not turned on, the apparatus power is turned on. Conversely, when the power button 8 is pressed while the apparatus power is turned on, the apparatus power is turned off.

FIG. 2 is a block diagram showing a hardware configuration of the pronunciation learning device according to the embodiment of the present invention. In FIG. 2, a CPU (Central Processing Unit) 12, various buttons 13, an analog / digital converter (hereinafter, A / D)
14, a ROM (read only memory) 15, a display driver 16, and a digital / analog converter (hereinafter, D / A) 17 are connected to a common data bus 18.

The CPU 12 controls the operation of each unit connected to the data bus 18 according to a control program stored in the ROM 15. In addition to the above-mentioned control program, the ROM 15 includes text data and font data of sentence examples used for learning pronunciation, voice data for outputting basic sounds of sentence examples, and basic pronunciation for evaluating the pronunciation of a learner. Data is stored. The various buttons 13 include the above-described cursor button 3, enter button 4, return button 5, listen button 6, talk button 7, power button 8, and the like.

The A / D 14 converts a learner's voice signal (analog signal) input from the microphone 11 into a digital signal that can be processed by the CPU 12. The display driver 16 controls driving of the display 2. The D / A 17 converts the audio data (digital signal) sent out by the CPU 12 into an analog audio signal when the CPU 12 reads out the audio data stored in the ROM 15 and outputs it from the earphone 10. As a result, the audio signal is sent from the D / A 17 to the earphone 10 and output as audio.

Next, the operation of the pronunciation learning device according to the embodiment of the present invention will be described. First, as a basic operation example, when the power button 8 is pressed when the power of the apparatus main body 1 is not turned on, the CPU 12 starts processing according to the control program stored in the ROM 15. As a result, a conversation selection screen for selecting a conversation to be learned is displayed on the display 2 as shown in FIG. On this conversation selection screen, the titles of the conversations to be learned are displayed side by side. The arrow in the screen can be moved up and down by the cursor button 3. Therefore, the learner can select a conversation title by operating the cursor button 3.

Next, the learner positions the arrow on the conversation title to be learned and presses the enter button 4. Then,
The screen of the display 2 switches to the conversation screen corresponding to the conversation title selected by the learner. For example,
When "Basic Conversation 1" is selected in FIG. 3A, the conversation screen is switched to a conversation screen as shown in FIG. Incidentally, when the return button 5 is pressed while the conversation screen shown in FIG. 3B is displayed, the screen returns to the conversation selection screen shown in FIG. 3A.

On the conversation screen shown in FIG. 3B, sentence examples for which pronunciation is to be learned are displayed side by side. The arrow in the screen can be moved up and down with the cursor button 3 as before. Therefore, the learner can select a sentence example displayed on the display 2 by operating the cursor button 3. When the listen button 6 is pressed with the arrow set to the desired sentence example, the earphone 10 outputs a voice that is the basis of the sentence example specified by the arrow. As a result, the learner can select a desired sentence example and listen to the pronunciation example.

On the other hand, when the talk button 7 is pressed while the conversation screen of FIG. 3B is displayed, the display screen of the display 2 switches as shown in FIG. 3C, and the learner waits for a speech. State. Here, the learner reads out the sentence example specified by the arrow. The voice when the learner reads the sentence example is input from the microphone 11. The input voice (analog signal) is converted into a digital signal by the A / D 14, and the sound generation is evaluated by the processing operation of the CPU 12 based on the control program stored in the ROM 15. Then, when the learner finishes reading out the sentence example, the evaluation result regarding the current pronunciation of the learner is automatically displayed on the display 2.

The evaluation of the pronunciation is performed, for example, using a point formula with the highest score being "100". In addition, the pass / fail judgment of the sounding state of the learner is performed based on a predetermined value (threshold) set in advance. Here, it is assumed that the predetermined value for pass / fail determination is set to “70”.

When the evaluation point of the sound generation in the processing operation of the CPU 12 is equal to or more than the predetermined value (70), a message meaning "good sound" is displayed on the screen of the display 2 as shown in FIG. “Your speech is goo
d! ”is displayed together with the evaluation score“ 86 pt. ”On the other hand, when the pronunciation evaluation score is less than the predetermined value, FIG.
As shown in (E), an evaluation score “67pt.” Is displayed together with the message “Excuse me?” Meaning “Please pronounce again”, and the word “Tanak” in the sentence example is displayed.
The word "a" is displayed in an inverted manner. This inverted word indicates a word whose evaluation score in word units is lower than a predetermined value, and which the learner has not pronounced well.

Next, the processing procedure from the input of the learner's voice to the display of the pronunciation evaluation in the pronunciation learning apparatus according to the embodiment of the present invention will be described with reference to the flowchart of FIG. . The following processing procedure is executed by the CPU 12 based on the control program stored in the ROM 15.

First, in step S11, it is determined whether or not a learner's voice has been input. Thereafter, when a voice is input, the evaluation score of the pronunciation in word units is calculated (step S12), and then the evaluation score of the pronunciation in sentence example units is calculated (step S13).

Here, the pronunciation evaluation in word units is performed for each word constituting the sentence example for which the pronunciation is to be learned. The evaluation points are based on the voice data input through the microphone 11 and the ROM 15. Is compared with the basic pronunciation data stored in the program, and based on the comparison result, the degree of pronunciation of the learner is calculated based on how close the pronunciation is to the basic pronunciation. For example, the sentence to be learned is “Nice to meet you, Mr. Ta.
naka. ”means“ Nice ”,“ to ”,“ meet ”,
Specifically, pronunciation is evaluated for each word such as “you”, “Mr.”, “Tanaka”, and the evaluation score is calculated for each word.

On the other hand, the pronunciation evaluation for each sentence example is the evaluation of the pronunciation of the whole sentence example (Nice to meet you, Mr. Tanaka.). It is calculated by averaging (weighted average) by weighting the evaluation points of each word constituting the sentence example with the number of characters (pronunciation length) of each word. Note that the pronunciation evaluation in word units is also used when the pronunciation evaluation in sentence example units is not good (in the case of rejection), particularly when extracting words whose pronunciation is not good.

Then, regarding the sentence example read out by the learner,
It is determined whether or not the pronunciation of the words constituting the sentence example is set (step S14). The setting information regarding the pronunciation importance is stored in the ROM 15 in advance corresponding to the text data of each sentence example. Various forms can be considered as a method of setting the importance in pronunciation.
For example, when the pronunciation importance is set in two stages of “high” and “low”, or in a plurality of stages (for example, 5
Stage).

Also, various forms are conceivable as a setting method for setting the degree of importance of pronunciation for a word included in a certain sentence example. For example, in a sentence example, as a word that is set to high importance, if the word is not pronounced correctly, the meaning of the entire sentence example will not be transmitted to the other party accurately or difficult to convey, or it will be misunderstood that it will not be pronounced correctly Words that can easily lead to In addition, in the sentence of a certain conversational screen, for example, when the difference between pronunciations of “R” and “L” is to be emphasized, words including pronunciations of “R” and “L” are not included. Set to high importance.

Contrary to this, as a word set to a low importance, a word in which the meaning of the entire sentence example is transmitted or transmitted to the other party without any problem even if the word is not correctly pronounced, or a word that is accurately pronounced Words that are not easily misunderstood even if they are not used are conceivable. In addition to this, words such as proper nouns included in the sentence example and words for which pronunciation is difficult to evaluate (recognize) on the device side are set to low importance.

Specifically, for example, in the sentence example displayed on the conversation screen of FIG. 3B, "I'veheard a lot about
In the case of the sentence example "you.", the word "heard" is assigned a high importance.
In the case of the sentence example "you, Mr. Tanaka.", the word (proper noun) "Tanaka" is set to a low importance.

In step S14, when it is determined that the pronunciation importance is set (when there is a word for which the pronunciation importance is set), the pronunciation is made according to the set importance. Then, a process of changing the evaluation point is performed (step S15). The evaluation point of the pronunciation changed here is an evaluation point when the pronunciation by the learner is evaluated for each sentence example in step S13. On the other hand, step S14
When it is determined that the pronunciation importance is not set (when there is no word for which the pronunciation importance is set), the process proceeds to step S16.

FIG. 5 is a flowchart showing an example of a processing procedure when changing the evaluation point of the pronunciation. First, it is determined whether or not a word set to low importance exists in the sentence example read out by the inventor (step S151).

If it is determined that there is a word set to a low importance, the pronunciation evaluation score of the word (word set to a low importance) in word units in step S12 is set to a predetermined value (70). ) Is determined (step S152). At this time, when it is determined that the pronunciation evaluation score of each word is less than the predetermined value, the pronunciation evaluation score of each sentence example is added (step S153). S1
The process goes to step S154 after passing 53. Thereafter, in step S154, it is determined whether or not a word set to high importance exists in the same sentence example, and if there is, the process proceeds to step S155, and if not, the process exits as it is.

On the other hand, if it is determined in step S151 that there is no word set to low importance in the sentence example read out by the inventor, only words set to high importance in the sentence example Therefore, it is determined whether or not the pronunciation evaluation score of the word (word set to high importance) in word units in step S12 is less than a predetermined value (step S155). At this time, when it is determined that the pronunciation evaluation score in word units is less than the predetermined value, the pronunciation evaluation score in sentence example units is deducted (step S1).
56) If it is determined that the pronunciation evaluation score of each word is equal to or greater than the predetermined value, the process skips step S156 and exits the process.

Here, the method of adding evaluation points in step S153 and the method of deducting evaluation points in step S156 will be described. First, as a method of adding evaluation points,
A method of adding a value obtained by dividing the difference between the pronunciation evaluation point of a word set to low importance and a predetermined value by the number of words of the sentence example to the pronunciation evaluation point of each sentence example calculated in step S13 (first method) Point addition method), or a method of adding a predetermined fixed number of points (second addition method) can be considered. On the other hand, as a method of deducting an evaluation point, a method of deducting a point from the pronunciation evaluation point of each sentence example calculated in step S13 under the condition that the pronunciation evaluation point becomes a value (69) that is less than a predetermined value is obtained. (First deduction method) or a method of deducting a predetermined fixed point (second deduction method) can be considered.

Thereafter, in step S16, it is determined whether or not the pronunciation evaluation score for each sentence example is less than a predetermined value. At this time, if the pronunciation evaluation score of each sentence example is equal to or more than a predetermined value,
A message meaning "good pronunciation" is displayed together with the evaluation points for each sentence example (step S17). If the pronunciation evaluation score for each sentence example is less than the predetermined value, a message meaning "reproduce again" is displayed together with the pronunciation evaluation score for each sentence example (step S18), and then the pronunciation evaluation score for each word is increased to the predetermined value. The words less than are inverted and displayed (step S19).

As described above, for a sentence example for which pronunciation is to be learned, the importance of pronunciation is set for the words constituting the sentence example, and the pronunciation evaluation score (sentence example) is set in accordance with the importance. By changing (adjusting) the pronunciation evaluation point of the unit, the pronunciation state when the learner reads out the sentence example can be more appropriately evaluated for the entire sentence example.

As a specific example, for example, "I've heard
When the learner reads out the sentence example "a lot about you.", the evaluation score when performing the word-by-word and sentence example units is as shown in FIG. 6, and the word "heard" is an important word in pronunciation. In such a case, conventionally, even if the evaluation score of the important word in pronunciation (evaluation score in word units) is less than the predetermined value of “50”, it is assumed that other Since the evaluation score of the word is high, the evaluation score for each sentence example is “86”, and the score is directly displayed as the evaluation result. On the other hand, in the present embodiment, when the evaluation score of an important word in pronunciation is less than a predetermined value, the evaluation score “86” in sentence example units is changed to “69” according to the above-described first deduction method, for example. Deduct points. For this reason, as the evaluation result of the pronunciation, the evaluation score for each sentence example is “6” together with a message meaning “reproduce again”.
9 "is displayed. Also, "heard" with low pronunciation evaluation
Is displayed in reverse video.

Further, for example, “Nice to meet you, Mr. Tana
When the learner reads aloud the sentence example "ka.", the evaluation score when performing the word unit and the sentence example unit is as shown in FIG. 7, and the word "Tanaka" is a non-important word (low In such a case, conventionally, if the evaluation score (word-based evaluation score) of a non-important word in pronunciation is “40”, which is significantly lower than a predetermined value, Even when all the evaluation points of the other words are equal to or more than the predetermined value, the evaluation point in the sentence example unit is “67”, and the score is displayed as the evaluation result as it is,
The word "Tanaka" is displayed inverted. On the other hand, in the present embodiment, when the evaluation score of a non-important word in pronunciation is less than a predetermined value, the evaluation score “67” in sentence example units is changed to “7” in accordance with, for example, the above-mentioned first addition method.
2 ”. For this reason, the pronunciation evaluation results
An evaluation score of "72" is displayed together with a message meaning "good pronunciation".

In the above-described embodiment, the processing procedure in the case where the pronunciation importance is set to two stages of “high” and “low” for the words constituting the sentence example has been described as an example. For example, when the pronunciation importance is set to five levels of “highest”, “middle and high”, “middle”, “middle and low”, and “lowest”, the evaluation of each sentence example according to each importance Points may be changed (adjusted) as appropriate.

Further, in the above embodiment, the setting information relating to the pronunciation importance is stored in the ROM 15 in advance, but the pronunciation importance and the words for which the importance is set are stored in the apparatus main body 1. Various buttons 13 provided
It is good also as a structure which can be set arbitrarily on the learner side using. By adopting such a configuration, for example, when there is a word that the learner himself wants to focus on pronunciation practice, the pronunciation of the word is improved by setting the pronunciation importance to the word high. You can practice pronunciation repeatedly up to. As a result, for example, it is possible to improve the efficiency of pronunciation practice, for example, the pronunciation of words that the learner is not good at can be accurately pronounced in a shorter period of time.

[0041]

As described above, according to the pronunciation learning apparatus of the present invention, for a sentence example for which pronunciation is to be learned, the significance of pronunciation is set by the setting means for the words constituting the sentence example. By changing the pronunciation evaluation point by the pronunciation evaluation means according to the importance, it is possible to more appropriately evaluate the pronunciation state of the entire sentence example by the learner.

[Brief description of the drawings]

FIG. 1 is an external view of a pronunciation learning device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a hardware configuration of a pronunciation learning device according to the embodiment of the present invention.

FIG. 3 is a diagram showing a display example related to pronunciation learning.

FIG. 4 is a flowchart showing a processing procedure of the pronunciation learning device according to the embodiment of the present invention.

FIG. 5 is a flowchart illustrating a processing procedure when changing a pronunciation evaluation point.

FIG. 6 is a diagram (part 1) for explaining a specific example of pronunciation evaluation;

FIG. 7 is a diagram (part 2) illustrating a specific example of pronunciation evaluation.

FIG. 8 is a flowchart showing a processing procedure of the conventional device.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Device main body, 2 ... Display, 10 ... Earphone, 1
DESCRIPTION OF SYMBOLS 1 ... Microphone, 12 ... CPU, 13 ... Various buttons, 15 ...
ROM

Claims (6)

[Claims] 1. An input unit for inputting a voice when a learner reads out a sentence example to be learned for pronunciation, a setting unit for setting pronunciation importance to words constituting the sentence example, and the input unit A pronunciation evaluation score is calculated based on a comparison result between the voice data input by the user and the basic pronunciation data corresponding to the sentence example, and the evaluation score is changed according to the importance set by the setting unit. A pronunciation learning device comprising an evaluation unit. 2. The method according to claim 1, wherein the pronunciation evaluation unit includes a first evaluation unit that evaluates pronunciation by the learner in word units, and a second evaluation unit that evaluates pronunciation by the learner in sentence example units. The pronunciation learning device according to claim 1, characterized in that: 3. The pronunciation learning device according to claim 1, wherein said setting means enables the learner to arbitrarily set the pronunciation importance. 4. The pronunciation learning device according to claim 1, wherein the setting means can set the pronunciation importance in a plurality of stages. 5. The pronunciation evaluation unit, wherein when the pronunciation of a word set to high importance by the setting unit is evaluated by the first evaluation unit and an evaluation score is lower than a predetermined value, the pronunciation evaluation unit 3. The pronunciation learning device according to claim 2, wherein the pronunciation evaluation point is reduced by the second evaluation means. 6. The pronunciation evaluation unit, wherein when the pronunciation of a word set to low importance by the setting unit is evaluated by the first evaluation unit and the evaluation score is lower than a predetermined value, the pronunciation evaluation unit 3. The pronunciation learning device according to claim 2, wherein an evaluation point of pronunciation by the second evaluation means is added.