KR100287228B1 - Language study method by interactive conversation - Google Patents

Abstract

The present invention relates to an interactive language learning method for enabling user-led language learning using a computer.

In particular, the present invention comprises a first step of constructing and outputting an initial screen for learning guidance at the same time as program start, and waiting for a learning request; A second step of waiting for the user's answer data to be recognized by reading and outputting the question data stored in the database, processing the question data stored in the database, and executing a subroutine for speech recognition when the learning request is received by the user in the standby state; A third step of determining whether the answer data is correct or incorrect when comparing the answer data of the user in the standby state; A fourth step of, if the answer is correct, reading the next question data having contents corresponding to the answer data of the user from the database and outputting the result; And a fifth step of displaying a message informing that the answer is incorrect, as a result of the determination.

Therefore, the present invention allows the user to proceed with a foreign language language with the computer as the subject of the conversation, so that the user can select a variety of materials in the direction intended by the user, and the language can be expressed in a mutually interactive three-dimensional conversation method between one-to-one or one-to-one algebra By learning, the user's concentration and learning effect can be enhanced.

Description

Language study method by interactive conversation}

The present invention relates to a language learning method that can perform language learning in a one-to-one or one-to-many user interaction using a computer. In particular, the computer recognizes and judges a user's question or response through voice recognition. This allows the user-led interactive dialogue to be developed between the user and the computer, and also provides the data of the conversations in various situations and the background image or sound data suitable for the conversations. The present invention relates to an interactive language learning method that can actively store a user's intention by selecting a next conversation in response to a user's question or response and outputting a background screen and sound according to the user's question or response. .

Conventional education system is composed of a one-way computer-driven conversation process of setting the story to be developed in advance, and requires an answer for a given sentence. Therefore, the user is asked to memorize and answer the sentence, so that the story development of the learning conversation is monotonous, there is no variety of choices in the answer, and there is a lack of flexibility, so that the user is easily tired and the learning efficiency is reduced. There was this. That is, the conversation in daily life develops various situations according to the contents of the question, and since the conversation proceeds with various forms of answers, the conversation form fixed in a specific form is unrealistic. This results in loss of learning motivation and disgust of the user, resulting in poor learning effect.

In addition, since it does not include a speech recognition function, there is no ability to determine the validity of the user's answer contents. After giving the user time to answer for a certain time, the conversation proceeds to the next sentence regardless of the contents of the answer, or to the next conversation only by the manual sentence input by the keyboard or the selection of the answer sentence presented by the mouse. It was so constructed that no real interaction was possible.

As described above, the conventional English learning systems are insufficient in the mutual cooperation system using the multimedia and the speech recognition method is not used, so that it is impossible for the user to communicate with the computer and it is difficult to learn English autonomously. In addition, the development of real-time interactive three-dimensional English language education software is not at home and abroad.

Therefore, the user can present a variety of answers appropriate to the situation according to the dialogue question, and the three-dimensional language that allows the conversation to proceed to a new situation according to the answer sentence presented so that the real-life English language such as the actual daily conversation situation can proceed. Development of an interactive English learning system is required.

Accordingly, the present invention has been made to solve the above problems, the present invention is to enable the computer to recognize and determine the user's questions or responses through voice recognition, various situations that can be deployed in real-life conversation By storing the dialogue contents and background image data or sound data suitable for the dialogue contents in the database in advance, the user can select the next dialogue in response to the user's question or response and simultaneously output the background image and sound accordingly. It is an object of the present invention to provide an interactive language learning method in which a user-directed two-way conversation can be developed between a user and a computer by actively coping with a user's intention.

In order to achieve the above object, an embodiment of the present invention provides a first step of constructing and outputting an initial screen for learning guide at the same time as program start, and waiting for a learning request through the initial screen; When the learning request is made by the user in the waiting state, the second step of reading and outputting the question data stored in the database and processing the same, and simultaneously executing a subroutine for voice recognition to wait for the user's answer data to be recognized. ; A third step of determining whether the answer data is correct / incorrect when the answer data of the user is recognized in the standby state of the second step; As a result of the determination of the third step, if the correct answer is a fourth step of reading the next question data corresponding to the user's answer data from the database and outputting the data, and returning to the user's answer data recognition standby state of the second step. ; As a result of the determination of the third step, it provides a interactive language learning method comprising a fifth step of displaying a message indicating that the incorrect answer is incorrect.

1 is a block diagram of an interactive language learning system according to an embodiment of the present invention;

FIG. 2 is a block diagram illustrating in detail the subprocessor and the data storage of FIG. 1;

3 is an operation flowchart of an interactive language learning method according to the present invention;

4A is an operation flowchart for the voice output subroutine of FIG. 3;

4b is an operation flowchart of the speech recognition subroutine of FIG.

4C is a flowchart illustrating an operation of a background image output subroutine of FIG. 3;

4d is an operation flowchart of the background sound output subroutine of FIG. 3;

5 is a block diagram illustrating another embodiment of an interactive language learning system according to the present invention.

<Explanation of symbols for main parts of the drawing>

10: key input unit 20: voice recognition unit

21: microphone 30: audio output unit

31: speaker 40: video output unit

41: monitor 50: main processing unit

60: subprocessing unit 61: background image processing unit

62: background sound processor 63: voice output processor

64: speech recognition processing unit 70: data storage unit

71: background image database 72: background sound database

73: voice output database 74: voice recognition database

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically showing the configuration of an interactive language learning system according to the present invention. The voice input means 21 converts a voice signal input from the outside into an electrical signal and an external signal through the voice input means. The voice recognition means 20 analyzes the voice signal coming from the voice signal, recognizes the meaning of the voice signal, and converts the text signal into text data, and processes the video signal and the audio signal to be output to the monitor screen and the speaker. Output means 40 and 41 and voice output means 30 and 31, key input means 10 for inputting the learning data to be stored in the database and for selecting the type of learning requested by the user; Question data composed of various situations, and answer data of the question and background image data or sound data according to the conversation contents learning data The data storage means 70 for storing the data and the learning data stored in the data storage means, and read out the desired type of learning data to the image and audio output means, and recognizes the user's voice through the speech recognition means And control means (50, 60) for comparing with the corresponding learning data read from the data storage means, and reading out the next question material corresponding to the contents answered by the user from the database according to the comparison result.

The image output means 40 may be configured as an image output device that receives the signal output from the control means and converts it into an image signal, and further includes a monitor 41 for outputting the converted signal to an external image screen. can do. The voice output means 30 may be configured as an audio device that receives a signal output from the control means and outputs a driving signal, and receives a driving signal output from the audio device and outputs sound or voice. It can be comprised including the 31 further.

The voice recognition means 20 may include a microphone 21 for receiving a human voice and converting the voice into an electrical signal, and receiving the signal output from the microphone 21 based on the pronunciation and accent of the user. It can be configured as a speech recognition device 20 that is converted into a matching word to output to the control means, it can also be configured as a sub-routine called and driven by the control means.

The control means may include a sub-processing unit 60 having respective processing routines separately, and a main processing unit 50 for outputting a plurality of process control signals for controlling the sub-processing unit 60. Data storage unit 70 for storing the may be included or may be configured separately. In the control means, when there is a learning request by the user, the main processing unit 50 controls the sub processing unit 60 using a plurality of tasks to read the learning data of the data storage unit 70 or read the read data. Export to audio and video output. In addition, the signal received from the outside through the voice recognition means 20 is compared with the answer data of the data storage unit 70, and compares the output of the background image and the background sound according to the comparison result, and at the same time You can read the training material for the next conversation or iterate the current training data along the previous flow of motion for repetitive learning.

FIG. 2A is a detailed block diagram illustrating the subprocessing unit and the data storage unit of FIG. 1. As shown in the figure, the data storage unit 70 is divided into a language learning question material for voice output, a plurality of answer data for the question material, and a background image data and background sound data according to the conversation contents. A database 71-74, wherein the questionnaire and a plurality of answer materials for the questionnaire, and a new questionnaire for each answer, and a plurality of new answer materials for the new questionnaire, Formed into a pile, it is configured to have the form of a tree structure as shown in FIG. 2B.

In this case, the background image database 71 stores image data for outputting a background screen, the background sound database 72 stores sound data for outputting a background sound, and the voice output database 73 outputs audio. For user data (such as question data and a number of answer data corresponding to them, each memory address for storing background image data and sound data for creating an atmosphere according to the conversation contents), and a voice recognition database ( 74 may store word data corresponding to the recognized voice.

The sub-processing unit 60 loads the data corresponding to the process control signal output from the main processing unit 50 from the background image database 71 and outputs the data to the monitor 41 and outputs data on the current state. A background image processor 61 which returns to the main processor 50; The background sound data corresponding to the process control signal output from the main processor 50 is loaded from the background sound data data 72 and output to the speaker 31, and the data on the current state is output to the main processor 50. A background sound processor (62) to return to; The voice data corresponding to the process control signal output from the main processor 50 is loaded from the voice output database 73 and output to the speaker 31, and the data on the current state is sent to the main processor 50. A voice output processor 63 for returning the voice signal; Controlled according to the process control signal of the main processor, the signal recognized by the voice recognition unit 20 and the data loaded from the voice recognition data data 74 are compared, and the data of the comparison result is output to the main processor 50. It is composed of a speech recognition processing unit 64 to return to.

3 is an operation flowchart of the interactive language learning method according to the present invention, which shows a signal processing flow executed by the main processing unit 50 of the control means.

In the drawings, the present invention comprises a first step (S311-S313) of constructing and outputting an initial screen for learning guidance at the same time as program start, and waiting for a learning request through the initial screen; When there is a learning request by the user, a subroutine for voice output (FIG. 4A) is executed to read and process the question data of the training data stored in the database, and the voice is output, while the subroutine for voice recognition (FIG. A second step (S318, S319-1, S320) of determining whether the user's response is correct from the result of speech recognition and the stored answer data in the database by executing 4b); According to the determination result, the question material and the answer data answered by the user are combined to form a file, the file is read from the database, and the data for the next question are selected and outputted by voice, thereby providing a response to the user's response. It can be configured as a third step (S320-1, S321) to continue the next learning of the matching content.

Here, in the first step, for example, a learning type (a) that can be divided by 'sentence practice', 'following a sentence' or 'pronounced practice', and a learning level (b) that can classify difficulty according to each learning type. Outputting a guide screen to select a user's intention and the like (S314); Waiting to select a learning type (a) or a learning grade (b) through the guide screen (S315); Setting the selected learning type and grade as a learning level when the selection is made in the waiting state (S316); If the selection is not made in the standby state, the method may further include setting a predetermined basic learning type and grade as a learning level (S317).

In the second step, the background data of the learning data stored in the database is read and monitored by executing a subroutine (FIG. 4C) for outputting the background screen simultaneously with the voice data processing subroutine (FIG. 4A) of the training data. The process may further include a step of outputting the screen to the screen, and executing the subroutine for outputting the background sound (FIG. 4D) to read the background sound data of the learning data stored in the database and output the process to the speaker (S319).

The third step further includes the step (S322) of outputting a warning message for notifying a correct answer when it is determined that the user's answer and the answer stored in the database are inconsistent from the subroutine execution result. In addition, the method may further include a step of executing a voice output subroutine to read out the stored answer contents and the new question data in the database and output the result in voice (S323). At this time, the warning message, the stored answer contents, and the new question data can be output by text, and voice or text can be used simultaneously.

In addition, the third step may further include a step (S324) of checking whether to repeat the learning data for the incorrect answer data, it can be configured to perform the repetitive learning according to the user's intention.

4A to 4D are flowcharts illustrating operations of the voice output subroutine, the voice recognition subroutine, the background image output subroutine, and the background sound output subroutine of FIG. 3, respectively. In the figure, each subroutine is called and driven by the main routine (FIG. 3), and the data for searching the database is provided from the main routine, and the result of searching the database and the speech recognition result are again. It is configured to return to the main routine.

FIG. 5 is a block diagram illustrating another embodiment of an interactive language learning system according to the present invention. The management fish 100 having the components shown in FIG. And a plurality of learning fish (110-130), which are operated by the management fish by sending and receiving an address and receiving a control signal, are connected in parallel to the management fish and have a large number of students (one professor and several students). It is organized in such a way that language learning can be done.

Referring to the operation of the present invention configured as described above and the effect thereof is as follows.

Referring to each of the drawings, at the same time as the start of the program operation (S311) outputs the initial screen (S312), and determines whether there is a learning request (S313). As a result of the determination, if there is a learning request, a selection screen for selecting a learning type or a learning grade (difficulty level) is output again (S314). At this time, the selection screen allows the user to select a desired type of learning and presents and selects menus of various levels so that the user can select the education level. The learning type and difficulty selected in the state are set to a level to be learned later (S316). If there is no selection, the learning level is set using a predetermined value (S317).

Then, the voice output subroutine is called (S318) to execute. In this case, in the voice output subroutine, as shown in FIG. 4A, the initial value of the database is checked using data provided from the main routine, and the training data corresponding to the set level is loaded from the database (S402). The question data of the data is output as voice (or text) (S403), and the value is returned to the main routine (S404).

Next, the background image output subroutine and the background sound output subroutine are simultaneously called (S319) and executed by using the data returned from the voice output subroutine, and the voice recognition subroutine is executed (S319-1). . Accordingly, in the speech recognition subroutine, as shown in FIG. 4B, the database initial value is checked using the data provided in the main routine, and then the voice recognition result is waited for (S412). In step S413, the answer data is compared with the speech recognition result and the answer data in the database (S414), and the comparison results are returned to the main routine, respectively (S415 and S416). In the background image output subroutine and the background acoustic subroutine, as shown in FIGS. 4C and 4D, the initial value of the database is checked using the data provided by the main routine, and the background image data of the training data corresponding to the set level is checked. And load background sound data from respective databases (S422 and S432), and output background image data and background sound data of the loaded learning data (S423 and S433).

Next, the result returned from the speech recognition subroutine is checked (S320), and if the answer is correct, the question data and the response data are combined into a file to read the next user fee from the database (S320-1). By searching the file in the database and executing the voice output subroutine (S321), new question material corresponding to the user's response is output as voice, so that new learning can be continued.

If the result returned by the speech recognition subroutine is checked (S320), and if it is not a correct response, a message for notifying the correct response is output as voice or text (S322), and the correct answer stored immediately in the database By reading the data and outputting in voice or text (S323), the user can check the correct answer. If the question is asked again to review the wrong problem (S324), and if the answer is affirmatively, the screen returns to the step (S319) of performing the background screen and sound output and speech recognition in order to learn the wrong problem again. If you do not want to review the wrong problem, in order to read the next user fee from the database, the question material and the answer data are combined into a file (S320-1), and the file is found in the database and the voice output sub By executing the routine (S321), a new question material corresponding to the user's response is output as a voice so that new learning can be continued.

The above processes are described as specific examples.

First, assuming a conversational situation where a Korean who greets a foreign visitor at Gimpo Airport asks about the visitor's experience in Korea, the mutual three-dimensional conversational English learning method proceeds as follows.

Welcome to Korea! Have you been in Korea before？

All answer sentences capable of answering the above questions are stored in the data storage unit 70. At this time, the wallpaper becomes Gimpo Airport, and the dynamic background and sound effects of the plane take off and land simultaneously. Assume that the answer stored in the data storage unit 70 is one of the following.

A) Yes, I have.

B) Yes, I have been there many times.

C) No, I have not.

When the user gives a matching answer among the above three answers, the contents of the next conversation according to the selected answer are stored in the data storage 70. For example, it is assumed that the new questions stored in the data storage unit 70 for the sentence A are as follows.

How long have you stayed？

As a possible answer here, it is assumed that the data storage unit 70 is stored as follows.

A) a week

2) two weeks

A) a month

D) More than two months

When the user answers a), it is assumed that the question of the next progress question and answer stored in the data storage unit 70 is as follows.

What was your impression of Korea？

The background screen for the question is a content for properly displaying the content of the question and is stored in the data storage unit 70. In the case of the question above, various three-dimensional background images such as a heart symbol representing a good feeling, an autumn Seorak mountain landscape representing beauty, and an interior of a complicated subway with a disgusting commute time are used as the wallpaper for emphasizing the impression. It is stored at 70.

As described above, the image output unit 40 visually stimulates the user's understanding while outputting the image data on the screen to increase learning efficiency.

Therefore, according to the learning method of the present invention, by enabling the language development of various contents according to the user's learning content, there is an advantage that the user can proceed the language as a subject, the three-dimensional image and the background sound according to the progress of the conversation By outputting to the monitor screen and the speaker, it is possible to educate in a three-dimensional dialogue like the real language, there is an advantage that can increase the user's concentration and learning effect.

Claims (4)

A first step of constructing and outputting an initial screen for learning guidance at the same time as program start, and waiting for a learning request through the initial screen; When the learning request is made by the user in the waiting state, the second step of reading and outputting the question data stored in the database and processing the same, and simultaneously executing a subroutine for voice recognition to wait for the user's answer data to be recognized. ; A third step of determining whether the answer data is correct / incorrect when the answer data of the user is recognized in the standby state of the second step; As a result of the determination of the third step, if the correct answer is a fourth step of reading the next question data corresponding to the user's answer data from the database and outputting the data, and returning to the user's answer data recognition standby state of the second step. ; And a fifth step of displaying a message indicating that an incorrect answer is a wrong answer as a result of the third step. The method of claim 1, wherein the fifth step Interactive language learning method characterized by outputting a warning message in voice and / or text to indicate that the correct answer data. The method of claim 1, wherein the fifth step If it is determined that the user's answer data and the answer data of the database do not match, executing the subroutine for voice output and outputting the correct answer data read from the database. Way. The method of claim 1, wherein the fifth step When it is determined that the answer data of the user and the answer data of the database do not match, the method may further include checking whether to repeat the learning again. And returning to the second step to output again.