JP2001051587A - Foreign language learning device, foreign language learning method, and computer readable recording medium recording foreign language learning program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a learner to diagnose by him-/herself the causes of an inability to catch speech and to effectively learn according to the causes, by converting a speaking speed of the speech synthesized from ordinary speech data and clear speech data without changing the speech characteristics. SOLUTION: In the 1st main matching, sentences are aurally outputted according to a speaking speed and narration set on a setting screen. Namely, the set narration is accessed by a storage part, and ordinary speech data or clear speech data corresponding to this narration is accessed by a speech data storage part 2 and outputted to a synthesized speech creating part 4. And, after a speech is synthesized by the synthesized speech creating part 4, a set speaking speed is read from a storage part, and when conversion of the speaking speed is necessary for outputting the speech at this speaking speed, the speaking speed is converted by a speaking speed conversion part 5, and then, the speech is outputted from a loudspeaker 7, and when the speaking speed conversion is not required, the speech is outputted from the loudspeaker 7 as it is.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a foreign language learning apparatus, a foreign language learning method, and a computer recording a foreign language learning program for learning a foreign language by outputting a sentence in a foreign language by voice. The present invention relates to a readable recording medium.

[0002]

2. Description of the Related Art Conventionally, a foreign language sentence read out by a native speaker or the like is recorded on a tape or the like,
Foreign language learning devices that reproduce this and allow learners to listen to it are widely used. Such a conventional foreign language learning apparatus includes a recording medium such as a tape for recording voice data, a synthesized voice generating unit for synthesizing voice from voice data recorded on the recording medium, and a synthesized voice generating unit. And a speaker for outputting the synthesized voice.

In the listening learning using such a foreign language learning apparatus, it is natural that when the learner's language level is low, the sentence may not be able to be heard even if the sentence is repeatedly heard. . The causes of the inability to hear are: (1) the utterance speed of the sentence is too fast to follow (this cause is referred to as "insufficient speech speed tracking power"); (2)
Like a liaison, linking, or reduced form, each word in a sentence is pronounced with a pronunciation change and cannot be heard. (3) There are unknown words in the sentence (this cause is called "Insufficient word power"), (4) There are other causes, (5) Insufficient speech speed following ability, Insufficient ability to follow changes or insufficient ability of words occurs in a complex manner.

[0004]

However, in a conventional foreign language learning apparatus, only sentences read out at a normal speed by a normal pronunciation method are recorded, and these are simply repeatedly output at a constant speed. Therefore, when the learner cannot hear the sentence, there is a problem that it is extremely difficult to determine which of the above causes is the cause. For example, when a sentence cannot be heard despite repeated listening, the learner can only confirm whether there is an unknown word in the sentence by referring to the attached text. And even if it is confirmed that there is no unknown word, at this point, since the sentence has been seen once, even if the same sentence is heard again, the meaning of listening learning is lost. . Therefore, it has been particularly difficult to confirm whether or not the cause is insufficient speech speed following ability or insufficient change following ability.

In addition, since the cause cannot be specified, efficient learning is performed, for example, learning for improving the insufficient change following ability is performed despite the lack of the speaking speed following ability. There was a problem that it was difficult. In addition, there has been proposed a foreign language learning device that can adjust the listening speed by reducing the rotation speed of the tape in order to compensate for the lack of speaking speed following ability. However, if the number of rotations of the tape is simply reduced as described above, the timbre and pitch of the original voice (these are referred to as “voice characteristics”) are different from each other. Meaning is lost.

[0006] Even if the learner himself has recognized the cause of the inability to hear, it has been difficult to perform effective learning corresponding to the cause. For example, even if it is recognized that the lack of change tracking power is the cause, there is no other way than to listen to the same sentence repeatedly or to reduce the tape rotation speed to listen. Could not be resolved.

The present invention has been made in view of such a conventional problem, and allows a learner to make a self-diagnosis of the cause of the inability to hear a voice, and to perform effective learning according to the cause. It is an object of the present invention to provide a computer-readable recording medium in which a foreign language learning device, a foreign language learning method, and a foreign language learning program that can be recorded are recorded.

[0008]

According to the present invention, there is provided a foreign language learning apparatus for learning a foreign language by outputting a predetermined content in a predetermined foreign language by voice. And stores normal voice data obtained by uttering the predetermined content with a normal pronunciation, and clear voice data obtained by uttering the predetermined content with a clear pronunciation without inflection. A voice data storage unit, a synthesized voice generation unit that synthesizes respective voices from the normal voice data and the clear voice data stored in the voice data storage unit, and a speech of the voice synthesized by the synthesized voice generation unit A speech speed conversion unit that converts the speech speed without changing its speech characteristics, a speech synthesized by the synthesized speech creation unit, and a speech output unit that outputs the speech converted by the speech speed conversion unit. , The voice Over data storage unit, synthesized speech creation unit is configured speech speed converting section, and a control unit that controls the sound output unit, in that it comprises as a feature.

According to a second aspect of the present invention, in the first aspect of the present invention, the normal voice data is obtained at a normal voice speed, and the clear voice data is obtained at a voice speed lower than a normal voice speed. Stored in the voice data storage unit, the speech speed conversion unit converts the voice speed of the voice synthesized from the normal voice data in the synthesized voice creation unit to a lower voice speed than at the time of data acquisition, and And converting the speech speed of the speech synthesized from the clear speech data in the synthesized speech creation unit into a speech speed higher than that at the time of data acquisition.

According to a third aspect of the present invention, in the first or the second aspect of the present invention, the data for outputting an audio out of the normal audio data and the clear audio data stored in the audio data storage section is provided. It is characterized by comprising an input unit for selecting the speed and for instructing the conversion speed in the speech speed conversion unit.

According to a fourth aspect of the present invention, there is provided a foreign language learning method for learning a foreign language by outputting a predetermined content in a predetermined foreign language by voice. A storage procedure for storing, in a voice data storage unit, normal voice data obtained by uttering, and clear voice data obtained by uttering the predetermined content with a clear pronunciation without inflection; A synthesizing procedure for synthesizing respective voices from the normal voice data and the clear voice data stored in the voice data storage unit; and a speech speed converting unit for talking on the voice synthesized by the synthesized voice generating unit. A conversion procedure for converting the voice speed without changing its voice characteristics, and outputting a voice synthesized by the synthesized voice generation unit and a voice converted by the speech speed conversion unit in the voice output unit. It is configured as characterized by comprising an output procedure for, the.

According to a fifth aspect of the present invention, in the fourth aspect of the present invention, in the storing step, the normal voice data is obtained at a normal voice speed, and the clear voice data is obtained at a lower speed than normal. Acquiring and storing at a speech speed, and in the conversion procedure, converting the speech speed of the speech synthesized from the normal speech data in the synthesized speech creation unit to a speech speed lower than that at the time of data acquisition; The production unit converts the speech speed of the speech synthesized from the clear speech data into a speech speed higher than that at the time of data acquisition.

According to a sixth aspect of the present invention, in the input device according to the fourth or fifth aspect, an audio output of the normal audio data and the clear audio data stored in the audio data storage is provided at the input unit. Select the data to perform
Further, an input procedure for inputting a conversion speed in the speech speed conversion unit is provided.

According to a seventh aspect of the present invention, in the fourth aspect of the present invention, when the learner cannot hear a voice synthesized from the normal voice data and output at a normal voice speed. In the above, the speech synthesized from the normal speech data is output at a lower speech speed than usual, and when the speech is heard by the learner, it is determined that the learner lacks the speaking speed following ability. A first cause determining step, and when the learner cannot hear the sound output in the first cause determining step, a sound synthesized from the clear sound data is output at a normal speech speed. If the learner can hear the voice, a second cause determination procedure for determining that the learner has a shortage of change following power, and the voice output in the second cause determination procedure is learned. If the person cannot hear the sound, The synthesized speech is output at a lower speech speed than normal, and if this sound is heard by the learner, it is determined that the learner lacks the speaking speed following ability and the change following ability. A third cause determination procedure for performing
When the learner cannot hear the voice output in the cause determination procedure, the content of the voice is displayed to the learner, and the display is used to determine whether the learner has insufficient word power. And a fourth cause determination procedure for determination.

According to the present invention as set forth in claim 8, in the present invention as set forth in claim 7, when it is determined in the first cause determination procedure that the learner has insufficient speaking speed following ability, In a first learning procedure of outputting a voice synthesized from the normal voice data at a lower speech speed than usual and further outputting the same voice at a gradually higher voice speed, and in the second cause determination procedure, When it is determined that the learner has a shortage of change following power, the voice synthesized from the normal voice data and the voice synthesized from the pre-clear voice data,
If it is determined in the second learning procedure that outputs alternately at the normal speech speed and the third cause determination procedure that the learner has a lack of a speech speed following ability and a change following ability,
The voice synthesized from the normal voice data and the voice synthesized from the clear voice data are alternately output at a lower voice speed than normal, and both voices are gradually alternated at a higher voice speed. And a third learning procedure for outputting to the third party.

According to a ninth aspect of the present invention, there is provided a computer-readable recording medium recording a foreign language learning program for outputting predetermined content in a predetermined foreign language by voice. In the case where normal voice data obtained by uttering predetermined content with normal pronunciation, and clear voice data obtained by uttering the predetermined content with clear pronunciation without inflection are stored, A synthesis procedure for synthesizing respective voices from the normal voice data and the clear voice data stored in the voice data storage section in the synthesized voice generation section; A conversion procedure for converting the speech speed of the synthesized voice without changing the voice characteristics thereof; and a voice output section for voice synthesized by the synthesized voice generation section. Is configured as a recording medium for executing an output procedure for outputting the converted audio, foreign language learning program characterized by comprising the computer in the speech speed conversion unit.

The present invention according to claim 10 provides the present invention according to claim 9.
In the present invention, in the audio data storage unit, the normal audio data is acquired at a normal speech speed, and when the clear audio data is acquired and stored at a speech speed lower than normal, In the conversion procedure, the speech synthesis unit converts the speech speed of the speech synthesized from the normal speech data into a speech speed lower than that at the time of data acquisition, and is synthesized from the clear speech data in the synthesized speech creation unit. It is configured as a recording medium for causing a computer to execute a foreign language learning program, which is characterized by converting the speech speed of the speech into a speech speed higher than that at the time of data acquisition.

The present invention according to claim 11 provides the present invention according to claim 9
In the present invention described in or 10, in the input unit, data for performing voice output is selected from the normal voice data and the clear voice data stored in the voice data storage unit, and conversion is performed in the speech speed conversion unit. An input procedure for inputting a speed is provided as a recording medium for causing a computer to execute a foreign language learning program.

[0019]

An embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of a foreign language learning device according to the present embodiment, FIG. 2 is a general flowchart of a foreign language learning program according to the present embodiment, FIG. 3 is a flowchart of essential parts in FIG. 2, and FIG. FIG. 5 shows an example of a display screen on a display.

First, the outline of the present invention will be described. The foreign language learning apparatus of the present invention is configured as, for example, a personal computer and a program installed therein. This foreign language learning device is intended to roughly output the contents of a foreign language and make the learner listen to it.
In particular, when the learner cannot hear the voice uttered at the normal speed and normal pronunciation, the voice is converted to a low speed without changing the voice characteristics and output, or a speed lower than the normal speed. In addition, a voice uttered with a clear pronunciation with no inflection can be converted and output at high speed without changing voice characteristics.

The learner listens in these various patterns, and according to the result of the listen, the learner cannot hear the voice uttered at the normal speed and the normal pronunciation. Can be analyzed.
Further, the learner can perform effective learning according to the cause. Note that the present device can perform learning of an arbitrary foreign language, but in the present embodiment, the case of learning English will be described. The content of the foreign language output as voice in the apparatus may be in any language unit such as a segment, a sentence (short sentence, long sentence), etc. In the present embodiment, it is assumed that one sentence is output as one unit. explain.

Next, the configuration of the foreign language learning apparatus will be described. As shown in FIG. 1, a computer body 1 of a foreign language learning apparatus includes a voice data storage unit 2, a text storage unit 3,
It contains a synthesized speech creation unit 4, a speech speed conversion unit 5, and a control unit 6. Further, a speaker 7, a display 8, a keyboard 9, and a mouse 10 are connected to the computer main body 1.

As shown in FIG. 1, the voice data storage unit 2 stores English sentences 1 to n (n is an integer of 1 or more) in normal speech speed and normal pronunciation (inflection in inflection) in English. A plurality of voice data obtained by uttering with the accompanying pronunciation (this voice data is referred to as “normal voice data.”
, "Sentence 1-normal" to "sentence n-normal"), and sentences 1 to n are uttered with pronunciation lower in speed than normal and with high intelligibility of individual words without inflection. A plurality of acquired voice data (this voice data is referred to as “clear voice data”. In FIG. 1, “sentence 1-clear”
To “sentence n-clear”).

The text storage unit 3 stores the above sentence 1
To n are stored as text data. The text storage unit 3 stores questions 1 to n (for example, questions of numbers and idioms appearing in the text) regarding the contents of the sentences 1 to n,
Correct answers 1 to n for each of the questions n are stored as text data. Note that the voice data storage unit 2 and the text storage unit 3 can be configured by a semiconductor storage element or a magnetic or optical storage medium.

The synthesized voice generation unit 4 includes a voice data storage unit 2
The speech is synthesized from the normal speech data or the clear speech data stored in. The voice synthesized in this way is output to the voice speed converter 5 when the voice speed conversion is required, and is directly output to the speaker 7 when the voice speed conversion is not required.

The voice speed converter 5 converts the voice speed of the voice output from the synthesized voice generator 4 and outputs the voice to the speaker 7. Specifically, the voice synthesized from the normal voice data in the synthesized voice generation unit 4 is converted into a voice lower in speed than the data acquisition speed, and the voice synthesized from the clear voice data in the synthesized voice generation unit 4 Is converted into a voice faster than the speed at the time of data acquisition. This speech speed conversion can be performed within an arbitrary range, but within a practical range (from the slowest speed that even a beginner of English conversation can sufficiently listen to, the fastest speed that can be uttered in normal English conversation) It is preferable to carry out the conversion in the range of up to).

The speech speed conversion by the speech speed conversion unit 5 can be performed in a plurality of stages. However, it is meaningless if the learner cannot grasp the change in the speech speed even if the speech speed is converted too finely. Therefore, it is preferable that the range in which the learner can grasp the change in the speech speed is set as the minimum unit of the conversion. In the present embodiment, a numerical value in increments of 0.1 is used as a scale indicating the minimum unit of the speech speed conversion (the larger the numerical value, the slower the speech speed). The normal voice data can be converted into seven steps of speech speed 1.0 to 1.6, and the clear voice data can be converted into six steps of speech speed 1.0 to 0.5.

In each case, the speech speed of 1.0 is used as the data acquisition speed. Here, as described above, since the normal voice data is acquired at the normal speech speed, the normal speech data is converted to the lower speed side with the normal speech speed being the maximum. In addition, since clear speech data is acquired at a speech speed lower than usual, the speech speed is converted to a higher speed with this speech speed being minimized. The maximum speech speed 0.5 of the clear voice data is set to be, for example, a normal speech speed in English. Such speech speed conversion by the speech speed conversion unit 5 is performed without changing the voice characteristics. That is, the voice speed is converted without changing the waveform of the voice data by digitally thinning out the voice data or adding the data.

In addition, the speaker 7 outputs the voice output from the synthesized voice generator 4 or the speech speed converter 5 to the learner. The display 8 displays each screen to the learner. The keyboard 9 and the mouse 10 are input units for the learner to input arbitrary contents in cooperation with the display 8 as necessary. By using the display 8, the keyboard 9, and the mouse 10, the learner can perform interactive input and output. In the following description, the input performed by the mouse 10 can be performed by the keyboard 9 as a matter of course. The control unit 6 is a control unit that controls the voice data storage unit 2, the text storage unit 3, the synthesized voice creation unit 4, the speech speed conversion unit 5, the speaker 7, the display 8, the keyboard 9, and the mouse 10. The control unit 6 specifically includes
It can be composed of a CPU and a program stored in storage means (not shown). In addition, the computer main body 1 is provided with a storage unit such as a main storage device and an auxiliary storage device (not shown), and stores a foreign language learning program.

A learning procedure performed by the foreign language learning apparatus thus configured will be described with reference to FIG. The learning procedure is roughly divided into a basic stage and a listening stage. First, the basic stage and the listening stage will be schematically described. In the basic stage, it is possible to set an initial value and give an instruction to end learning. In the listening stage, three stages of a main match, a revenge match, and a short training can be executed.

In the main match and the revenge match, learning is performed in a game-tailored story development in which the learner competes with a virtual questioner in order to enhance the learner's motivation and learning effect. If the learner wins the first main match, he can move up one level to the next level of the main match. on the other hand,
If the learner loses, proceed to the revenge match. In this revenge match, the questions are lowered one level.
The meaning of the level will be described later.

After the revenge match, the learner
Regardless of the result of this revenge match, proceed to short training. In this short training, the learner can output the speech of any speech method of the normal speech and the clear speech at an arbitrary speech speed and listen. The cause of the loss in the main match can be self-diagnosed, and learning according to the cause can be performed. Of course, it is also possible to use an arbitrary story other than that shown in the present embodiment, or to omit such a story.

Further, each stage will be described in detail with reference to FIG. First, the basic stage will be described. After the program start (S1), the basic screen of the basic stage is displayed (S2). In this basic screen, the user can proceed to the setting screen. In this setting screen, the standard speaking speed of the sentence output at each stage of the listening stage, the speech style of normal voice or clear voice, the number of levels to increase when winning in the main match, lower when losing in the main match Setting items such as the number of levels and the number of executions of the revenge match can be set (S3). This set value is stored in a storage unit (not shown). By instructing the start of learning on the basic screen, the user can proceed to the listening stage.

Next, the listening stage will be described in detail. In the first main match (S4),
Sentences are output as speech according to the speech speed and speech style set on the setting screen. That is, the set speech is called from a storage unit (not shown), and normal speech data or clear speech data corresponding to the speech is called from the speech data storage unit 2 and output to the synthesized speech creation unit 4. After the synthesized voice is synthesized by the synthesized voice generating unit 4, the set voice speed is called from a storage unit (not shown). If voice speed conversion is required to output at this voice speed, the voice speed converting unit 5 After the speech speed conversion is performed, the signal is output from the speaker 7, and when the speech speed conversion is not required, the signal is output from the speaker 7 as it is. For example, if “speech style = normal speech, speech speed = 1.6” is set, the speech rate is converted by the speech rate conversion unit 5 and then output from the speaker 7. Speed = 1.0 "
Is set to be output from the speaker 7 as it is.

Thereafter, as a question from a virtual questioner, a question related to the content of the sentence output earlier is called from the text storage unit 3 and displayed on the display 8. Also, a plurality of candidates for the answer to the question are displayed on the display 8, and the learner selects the correct answer from the plurality of candidates via the mouse 10. After an answer is input from the learner, the correct answer to the question is called from the text storage unit 3, and the correct answer and the answer input by the learner are cross-checked to determine whether the answer is correct. This result is displayed on the display 8 and stored in a storage unit (not shown). Such questions and answers are repeated a predetermined number of times, and the main match ends. At this end point, the number determined to be incorrect in the previous correctness judgment is compared with a predetermined number stored in the storage unit in advance, and if the number is within the predetermined number, the learner wins in the main match, and the predetermined number is determined. If it exceeds, it is determined that the game has been lost (S
5). This determination result is displayed on the display 8.

When the learner wins, the level is increased by one level, and the process proceeds to the next main match. Here, raising the level means raising the speech speed or changing the clear voice to normal voice and outputting the voice. In the present embodiment, the voice speed is increased by the number of levels set on the setting screen. I do. For example, if the first main match is performed with "speech style = normal voice, speech speed = 1.6" and the number of levels set on the setting screen is "0.2", the next main match is " Speech method = normal voice, speech speed = 1.4 ". In such a main match of different levels, a virtual questioner matching the image of each level is displayed on the screen, and the question is performed. When the learner proceeds in the main match in this way and further wins the main match of the highest level, that is, the level of “speech style = normal speech, speech speed = 1.0”, the listening battle ends ( S6) Return to the basic screen. Also, when the total number of losses in the main match exceeds a predetermined number, the listening battle ends and the screen returns to the basic screen.

On the other hand, if the learner loses in each main match, the level is lowered by one step, and the process proceeds to a revenge match (S7). Here, lowering the level means lowering the voice speed or changing the normal voice to clear voice and outputting the voice. In this embodiment, the voice speed is lowered by the number of levels set on the setting screen. I do. This revenge match is performed in the same manner as the main match. That is, the sentence is output as speech, and after the question is set, the learner inputs the answer. Then, the answer and the correct answer are collated, and the correctness of the answer is determined. This result is stored in a storage unit (not shown). Such questions and answers are repeated a predetermined number of times, and the revenge match ends.
At the end of the revenge match, the process shifts to short training without judging the win or loss (S8).

In this short training, the learner can switch the speaking style and the speaking speed arbitrarily and listen repeatedly as many times as possible. It can be carried out. First, a device configuration for performing the self-diagnosis and learning will be described. Upon completion of the revenge match, the short training screen shown in FIG. On this screen, a speech style selection column 11 for selecting which speech style, "normal speech" or "clear speech", is to be output is provided. By clicking one of the speech styles with the mouse 10, the speech style can be arbitrarily selected. Further, a speech speed designation field 12 for designating a speech speed of voice output is provided. By clicking the UP button 12a or Down button 12b provided on the side of the speech speed designation field 12 with the mouse 10, the speech speed designation field 12 is provided. Can be set to an arbitrary speed by increasing / decreasing the speech speed displayed. Further, a sentence display field 13 for calling up the sentence output from the text storage unit 3 and displaying the sentence as a text, and an output instruction button 14 for instructing the sound output by clicking the mouse 10 are provided. In addition, an end instruction button 15 for instructing the end of the short training by clicking the mouse 10 is displayed. Of course, other screen configurations and input formats can be used.

Next, the concept of the self-diagnosis will be described with reference to FIG.
This will be described with reference to FIG. In FIG. 4, the vertical column shows the speech (normal voice or clear voice), and the horizontal column shows the voice speed (high speed or low speed). Each column in the table indicates that when a voice is output in various combinations 1 to 4 of the corresponding speech speed and speech style, the voice can be heard (listening possible). In each of the cases where it was not possible to listen (cannot listen), the ability (the cause of the inability to listen) that the learner could consider to be insufficient at the time of listening to the previous main match is shown.

In FIG. 4, if it is possible to listen to the voice output in combination 1 "speech style = normal voice, voice speed = high speed" (when the main match is won), it is natural. However, it can be diagnosed that the cause of inaudibility did not exist. On the other hand, if it is not possible to listen to the voice output in combination 1, the learner experiences one or all of "sufficient speaking speed following ability, insufficient change following ability, insufficient word power, and other causes". It is thought that it was. In this case, it is possible to identify a cause that may be considered to be insufficient for the learner by sequentially performing the listening with the combinations 2 to 4.

That is, when the combination 1 fails to hear, the listening is further performed by the combination 2 “speech method = normal voice, speech speed = low speed”. Here, in the case where the listening is possible, since the listening can be performed by lowering the speaking speed, it can be specified that the learner has only “sufficient speaking speed following ability”. On the other hand, when listening is impossible, it is considered that one or all of “insufficient ability to follow changes, insufficient vocabulary, and other causes” have occurred in the learner.

If the combination 2 fails to listen,
Further, listening is performed by the combination 3 “speech method = clear speech, speech speed = high speed”. Here, in the case where the listening is possible, it is possible to specify that the learner has only “inadequate change follow-up power” since the listening can be performed by eliminating the inflection. On the other hand, when it is impossible to listen, it is considered that one or all of "sufficient speaking speed following ability, lack of word ability, and other causes" have occurred in the learner.

If the combination 3 cannot be heard,
Finally, listening is performed by the combination 4 "speech method = clear speech, speech speed = low speed". Here, when listening is possible, it can be specified that the learner has both "sufficient speaking speed following ability and insufficient changing following ability". It is considered that any or all of the "other causes" have occurred. If the word cannot be heard in combination 4, the learner looks at the text display of the sentence that could not be heard and checks whether there is a word that he does not know in the sentence. Can be diagnosed. In other words, if there is an unknown word in the sentence, it is considered that the learner has "word deficiency", and if there is no unknown word, it is considered that the learner has "other causes". Can be As described above, by performing listening in each combination in the short training and by visually checking the text display, the learner can self-diagnose the cause of the inability to hear.

In this diagnostic procedure, the speech speed is expressed as "high speed" or "slow speed", but these are relative to each other, and the speech speed of combination 1 is higher than the speech speed of combination 2. It is sufficient that the speech speed of the combination 3 is higher than that of the combination 4. Conversely, the speech speed of the combination 2 may be lower than the speech speed of the combination 1, and the speech speed of the combination 4 may be lower than the speech speed of the combination 3. For example, if the combination 1 is changed to “speech method = normal speech,
1.2 ", the combination 2 is" speech = normal speech,
Speaking speed = low speed = 1.4 "or" speech method = normal voice, talking speed = low speed = 1.6 ". The speech speed may be switched to a plurality of stages. For example, if the speech cannot be heard in combination 3, the speech speed is gradually reduced in steps of 0.1 and the combination 4
, It is possible to make a detailed diagnosis of the degree of lack of the following force.

Next, a more specific procedure of self-diagnosis and learning in the short training will be described with reference to FIG. In the flowchart of FIG. 3, the short training is conceptually broadly divided into a diagnosis stage for diagnosing the cause of the inability to hear and a learning stage for performing learning for eliminating the cause diagnosed in the diagnosis stage.

In the diagnosis stage, the learner makes various inputs using the speech selection column, speech speed designation column, sentence display column, keyboard 9 and mouse 10 shown in FIG. 4 (Combination 1 is excluded because it has already been heard in the main match)
Listening by.

That is, first, the speech style and the speech speed are set to the combination 2 and the listening is performed (S8-1). Then, it is judged by himself / herself whether or not to listen (S8-2). If the user can listen to this, self-diagnosis can be made that "slow speed following ability" has occurred (S8-3). ). Therefore,
To solve this, move to the learning stage, first listen to normal voice at low speed, gradually increase the speech speed to high speed,
Desirably, the listening is performed until the normal speaking speed is reached (S8).
-4). Then, by learning, the user becomes accustomed to a high speech speed, so that the shortage of the speech speed following ability can be resolved.

On the other hand, in the case where the combination 2 cannot be heard, the speech is set with the speech style and the speech speed set to the combination 3 (S8-5). Then, as in the previous case, it is judged by himself / herself whether or not to listen (S8-6), and if this can be heard, it can be diagnosed that "insufficient change following power" has occurred (S8-7). . Then, the process proceeds to the learning stage, in which the normal voice and the clear voice are alternately switched to perform the listening (S8-8). By learning the state of the pronunciation change of the liaison or the like by this learning, it is possible to solve the shortage of the change following ability.

[0049] If the combination 3 cannot be heard, the speech style and the speech speed are set to combination 4 and the listening is performed (S8).
-9), it is determined whether or not this is possible (S8-10). If this can be heard, the cause can be specified that "insufficient speaking speed follow-up ability and change following ability" has occurred (S8-).
11). Therefore, in order to solve this, we switch between normal voice and clear voice at low speed first and listen.
This is gradually increased to a high speed, and listening is desirably performed until a normal speech speed is reached (S8-12). By this learning, both the lack of speaking speed following ability and the lack of change following ability can be solved.

If the combination 4 cannot be finally heard, the text displayed in the text display column of FIG. 5 is checked (S8-13) to check whether there is any unknown word (S8-13). S8-14). If there is a word that is unknown here, the self-diagnosis can be made that "insufficient word power" has occurred (S8-15), and the word power can be increased by memorizing this word (S8-16). . On the other hand, if there is no unknown word, it can be self-diagnosed that "other cause" has occurred (S8-17). Step S8
It is desirable that the text display in -13 is performed after the learner's instruction in order to prevent the learner from looking at the text carelessly. As described above, according to the present foreign language learning apparatus, the learner can identify the cause of lack of himself / herself and perform learning according to the cause. Can be.

After the completion of such learning, the learner returns to FIG.
Is clicked with the mouse 10 to terminate the short training. Then, as shown in FIG. 2, at the end of the short training,
The number determined to be incorrect in the revenge match is compared with a predetermined number.If the number is within the predetermined number, the learner wins the revenge match, and if the number exceeds the predetermined number, the learner is determined to lose. Is displayed on the display 8 (S
9). Here, if the learner wins, the level is raised by one level, and the player can move on to the main match. On the other hand, if the learner loses, the listening battle ends and the screen returns to the basic screen. By instructing termination of the program on this basic screen, the program is terminated (S10).

In the above description, the cause diagnosis and learning in the short training are performed by the learner's self-determination and manual operation. However, the cause diagnosis can be automatically performed by a computer. It is also possible to automatically start learning according to the result of the diagnosis. For example, each output step (S8-1, S8-5, S8-5,
An output control unit that automatically performs S8-9 and S8-13) may be provided, and the output control may be performed according to the flow of FIG. 3 according to the determination result of the learner.

Further, the control section 6 performs each cause diagnosis step (S8-3, S8-7, S8-) in the diagnosis stage of FIG.
11, S8-15, S8-17) may be provided, and a cause diagnosis may be performed according to the flow of FIG. In this case, for example, the diagnosis result may be automatically displayed on the display 8. Further, the control unit 6 is provided with a learning control unit for automatically performing each learning step (S8-4, S8-8, S8-12, S8-16) in the learning stage of FIG. 3, and according to the flow of FIG. The learning content may be automatically presented. According to such a program, the cause diagnosis and learning of the inability to hear can be automatically specified, and the cause diagnosis and learning can be performed more easily and reliably. When the cause diagnosis and the learning are automatically performed in this manner, some of the input units such as the speech style selection column 11 and the speech speed designation column 12 may be omitted.

Although one embodiment of the present invention has been described, the present invention is not limited to the above-described embodiment, and may be embodied in various forms within the scope of the technical idea. Hereinafter, these different embodiments will be described. In the above embodiment, the present apparatus has been described as being constituted by a personal computer and a program installed in the personal computer. However, the apparatus may be constituted as a video game machine or a device exclusively for learning foreign languages.

With respect to the data stored in the voice data storage section 2, normal voice data does not necessarily need to be uttered at a normal voice speed and acquired, and clear voice data is not necessarily uttered at a voice speed lower than normal. It does not need to be obtained. The normal voice data and the clear voice data may be obtained by uttering at a voice speed other than those described above, and converted into a predetermined voice speed by the voice speed conversion unit. In addition, even when the main match and the revenge match are omitted in the flow of FIG. 2 and only the short training is executed, it is possible to perform the diagnosis of the cause of the inability to hear and the learning according to the cause.

[0056]

As described above, the foreign language learning device according to claim 1, the foreign language learning method according to claim 4, or the recording for causing a computer to execute the foreign language learning program according to claim 9 According to the medium, it is possible to listen to a normal pronunciation voice at a low speed, or to listen to a clear pronunciation voice without inflection at a high speed. The missing cause can be easily identified.

According to the apparatus described in claim 2, the method described in claim 5, or the recording medium described in claim 10, it is possible to convert a normal voice acquired at a normal speech speed to a low speed,
Also, it is possible to convert a clear voice acquired at a lower speech speed than usual to a high speed. Here, it is easy for a speaker to pronounce normal voice at a normal voice speed and pronounce clear voice at a low voice speed, so by enabling voice output based on such data, In addition, data can be more easily obtained.

According to the apparatus described in claim 3, the method described in claim 6, or the recording medium described in claim 11, it is possible to listen by changing the speech style and the speech speed via the input unit. Self-diagnosis can be easily performed manually.

According to the method of the seventh aspect, by sequentially listening in accordance with each procedure, the cause of the lack of the learner can be easily specified.

According to the method of the eighth aspect, by learning in accordance with each procedure, the learner can perform learning according to the cause of his / her inability to listen, and a more efficient foreign language learning It can be performed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an overall configuration of a foreign language learning device according to an embodiment of the present invention.

FIG. 2 is an overall flowchart of a foreign language learning program according to an embodiment of the present invention.

FIG. 3 is a flowchart of a main part of FIG. 2;

FIG. 4 is a diagram categorizing causes of inaudibility.

FIG. 5 is an example of a display screen on a display.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Computer main body 2 Voice data storage part 3 Text storage part 4 Synthetic voice creation part 5 Speech rate conversion part 6 Control part 7 Speaker 8 Display 9 Keyboard 10 Mouse 11 Speech style selection field 12 Speech rate specification field 12a UP button 12b Down button 13 Text display field 14 Output instruction button 15 End instruction button

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Eiichi Miyasaka 1-10-11 Kinuta, Setagaya-ku, Tokyo Japan Broadcasting Corporation Research Institute (72) Inventor Motohiro Takano 2-6-Nishishinjuku, Shinjuku-ku, Tokyo No. 1 Stock Company CSK F-term (reference) 2C028 AA03 BB07 BC02 BD03 CA12 5D045 AB12 BA02

Claims (11)

[Claims] 1. A foreign language learning apparatus for learning a foreign language by outputting a predetermined content in a predetermined foreign language by voice, comprising: a normal voice acquired by uttering the predetermined content with a normal pronunciation. Data, and a voice data storage unit for storing clear voice data obtained by uttering the predetermined content with a clear pronunciation without inflection, and the normal voice data stored in the voice data storage unit, A synthesized voice generating unit that synthesizes each voice from the clear voice data; a voice speed converting unit that converts the voice speed of the voice synthesized by the synthesized voice generating unit without changing its voice characteristics; A voice synthesized by the voice generating unit, and a voice output unit that outputs the voice converted by the voice speed converting unit; the voice data storage unit, the synthesized voice generating unit, the voice speed converting unit;
And a control unit that controls a voice output unit. 2. The method according to claim 2, wherein said normal voice data is obtained at a normal voice speed, said clear voice data is obtained at a voice speed lower than normal, and stored in said voice data storage unit. Converting the speech speed of the speech synthesized from the normal speech data in the synthesized speech creation unit to a speech speed lower than that at the time of data acquisition, and converting the speech synthesized from the clear speech data in the synthesized speech creation unit. 2. The foreign language learning device according to claim 1, wherein the speech speed is converted into a speech speed higher than that at the time of data acquisition. 3. The data processing apparatus according to claim 1, further comprising: a selector for selecting data for outputting a voice from the normal voice data and the clear voice data stored in the voice data storage unit. 3. The foreign language learning device according to claim 1, further comprising an input unit for instructing. 4. A foreign language learning method for learning a foreign language by outputting predetermined content in a predetermined foreign language by voice, comprising: a normal voice acquired by uttering the predetermined content with a normal pronunciation. A data and a storage procedure for storing clear voice data obtained by uttering the predetermined content with a clear pronunciation without inflection in a voice data storage unit; A synthesis procedure for synthesizing respective voices from the stored normal voice data and the clear voice data, and a voice speed conversion unit for changing a voice speed of the voice synthesized by the synthesized voice generation unit. A conversion procedure for converting the speech without converting the speech, and an output procedure for outputting the speech synthesized by the synthesized speech creation section and the speech converted by the speech speed conversion section in the speech output section. A foreign language learning method characterized by the following. 5. In the storing step, the normal voice data is obtained at a normal voice speed, and the clear voice data is obtained and stored at a voice speed lower than a normal voice speed. The voice synthesizer converts the voice speed of the voice synthesized from the normal voice data into a voice speed lower than that at the time of data acquisition, and the voice speed of the voice synthesized from the clear voice data in the synthesized voice generator is data. 5. The foreign language learning method according to claim 4, wherein the speech speed is converted to a higher speech speed than at the time of acquisition. 6. An input unit for selecting data for outputting a voice from the normal voice data and the clear voice data stored in the voice data storage unit, and inputting a conversion speed in the speech speed conversion unit. 6. The foreign language learning method according to claim 4, further comprising an inputting step. 7. When a learner cannot hear a voice synthesized from the normal voice data and output at a normal voice speed, the voice synthesized from the normal voice data is reduced to a voice speed lower than normal. A first cause determination procedure for determining that the learner lacks speaking speed following power when the learner can hear the voice, and a first cause determination procedure. If the learner does not hear the sound, the synthesizer outputs a speech synthesized from the clear speech data at a normal speech speed, and if the learner can hear the sound, the learner can follow the change. A second cause determination procedure for determining that a shortage has occurred, and a speech synthesized from the clear speech data when a learner cannot hear the speech output in the second cause decision procedure. Output at a lower speaking speed than normal, this sound Is output by the third cause determination procedure for determining that the learner lacks the speaking speed following ability and the change following ability when the learner can hear the third cause determination procedure. A fourth cause determining step of displaying the content of the voice to the learner when the learner cannot hear the voice and determining whether or not the learner has insufficient word power by looking at the display; 7. The foreign language learning method according to claim 4, comprising: 8. When the first cause determination step determines that the learner has insufficient speech speed following ability, the speech synthesized from the normal speech data is reduced to a speech speed lower than normal. A first learning procedure of outputting the same voice at a gradually higher speech speed, and a second learning procedure in which it is determined that the learner has a shortage of change following power in the second cause determination procedure. A second learning procedure for alternately outputting a voice synthesized from the normal voice data and a voice synthesized from the pre-clear voice data at a normal speech speed; and the third cause determination procedure. When it is determined that the learner lacks the speaking speed following ability and the change following ability, the voice synthesized from the normal voice data and the voice synthesized from the clear voice data are compared with the normal voice data. Output alternately at a low talk speed, and Third learning procedure and, foreign language learning method according to claim 7, characterized in that it comprises an outputting both of these audio gradually alternately at a high speed of speech speed. 9. A computer-readable recording medium on which a foreign language learning program for outputting predetermined content in a predetermined foreign language by voice is recorded, wherein said predetermined content is recorded in a voice data storage unit by a normal pronunciation. In a case where normal voice data obtained by uttering and clear voice data obtained by uttering the predetermined content with a clear pronunciation without inflection are stored, A synthesis procedure for synthesizing respective voices from the normal voice data and the clear voice data stored in a data storage unit; and a speech speed conversion unit for determining a voice speed of the voice synthesized by the synthesized voice creation unit. A conversion procedure for converting without changing the characteristic feature, a voice output section, a voice synthesized by the synthesized voice creation section, and a conversion performed by the speech speed conversion section. And a computer-readable recording medium storing a foreign language learning program for causing a computer to execute the foreign language learning program. 10. The method according to claim 1, wherein said normal voice data is obtained at a normal voice speed and said clear voice data is obtained and stored at a voice speed lower than normal. In the procedure, the speech speed of the speech synthesized from the normal speech data in the synthesized speech creation unit is converted to a speech speed lower than that at the time of data acquisition, and the speech is synthesized from the clear speech data in the synthesized speech creation unit. A computer readable recording of a foreign language learning program according to claim 9 for causing a computer to execute the foreign language learning program, wherein the computer converts the voice speed of voice into a higher voice speed than at the time of data acquisition. Recording medium. 11. An input unit for selecting data to output a voice from the normal voice data and the clear voice data stored in the voice data storage unit, and inputting a conversion speed in the speech speed conversion unit. 11. A computer-readable recording medium storing a foreign language learning program according to claim 9 or 10 for causing a computer to execute the foreign language learning program, comprising: