KR101743708B1 - System for studying Korean language using accuracy of final consonant, and operating method thereof - Google Patents

Abstract

The present invention relates to a system and a method for providing a Korean language learning service by determining the accuracy of a base, and is based on determining the accuracy of the pronunciation (pronunciation) , And Korean language learning can be efficiently conducted.

Description

[0001] The present invention relates to a system and method for providing a Korean language learning service,

The present invention relates to a system for providing a Korean language learning service and a method of operating the same, and more specifically, to a technique for acquiring pronunciation information of a learner and determining the accuracy of a learner (longevity) as an element of an evaluation criterion.

With the spread of the Korean Wave culture centering on Korean dramas and K-pops, interest in Korea has been rapidly increasing, and more and more foreigners are trying to learn Korean. The number of foreigners who want to learn Korean is increasing, but the education infrastructure is still lacking.

Also, even in the case of Koreans, elementary school students who are unable to express their opinions correctly due to their lack of ability to pronounce correct Korean are increasing, and measures are needed to improve their communication skills.

In other words, all the learning is the same, but the language learning also needs to grasp the current level of the learner and then the teaching method should be done. Especially in case of language learning, pronunciation accuracy is the highest priority.

Currently, various learning services are provided to improve language skills. However, most of them are adapted to foreign language learning, and there is no learning service that reflects the characteristics of Korean language.

On the other hand, Korean Patent Registration No. 10-1584254 is known as a related art related to language learning.

SUMMARY OF THE INVENTION The present invention has been devised to solve the problems of the prior art as described above, and it is an object of the present invention to provide a learning method, And to provide a technique that enables efficient learning.

According to another aspect of the present invention, there is provided a system for providing a Korean language learning service, comprising: a communication processing unit for transmitting and receiving information to and from a learner terminal; A waveform analyzer for analyzing a waveform and extracting a consonant after receiving learner pronunciation information from the learner terminal through the communication processor; An achievement level determining unit for comparing the extracted pedestrian with the standard pronunciation information to determine accuracy of pronunciation of the pedestrian (consonant) sound and scoring it to calculate achievement information; A member information management unit for storing the achievement degree information in the member information D / B; And a learning information operating unit for providing customized learning information to the learner terminal based on the achievement information.

Here, the waveform analyzing unit extracts only the waveform information about the utterance after separating each syllable of the learner's pronunciation information into the preliminary, the neutral, and the utterance, and the achievement degree judging unit preliminarily displays the standard pronunciation information through the learning information operating unit The coincidence rate can be scored by comparing the waveform information about the utterance of one word with the waveform information about the utterance of the learner pronunciation information extracted by the waveform analyzer.

In addition, the learning information operating unit may check whether the performance information is the standard or more, and if it is confirmed that the performance information does not meet the criteria, the learning information operation unit may provide a practice word having the same baseline and different initials or neutrality to the learner terminal.

When the user requests the conversation information from the learner terminal through the communication processing unit, the learning information operation unit extracts the conversation information corresponding to the conversation information request and provides the conversation information to the learner terminal, After confirming the achievement degree information of the learner who has requested the conversation information, the degree of providing the conversation information may be differentiated according to the achievement degree information.

Further, the present invention may further include a member recommendation unit for recommending face-to-face learning by providing high-score learner information adjacent to the location of the low-score learner through the achievement degree information.

According to another aspect of the present invention, there is provided a method for providing a Korean language learning service, the method comprising: providing learning information including standard pronunciation information to a learner terminal in a Korean learning service providing system (a); (B) receiving learner pronunciation information generated by the learner terminal in response to the learning information; (C) analyzing a waveform of the learner's pronunciation information and extracting a consonant; (D) comparing the extracted pedestal with standard pronunciation information to determine the accuracy of pedestal pronunciation and scoring it to calculate achievement information; (E) storing the achievement degree information in the member information D / B; And (f) providing customized learning information to the learner terminal based on the achievement information.

In the step (c), only the waveform information for the utterance is extracted after phonemes of the syllables of the learner's pronunciation information are divided into the initials, the neutral and the utterances, and the step (d) The waveform information of the utterance of the word presented in advance and the waveform information of the utterance of the learner pronunciation information extracted in the step (c) can be compared to score the coincidence rate.

In the step (f), if it is confirmed that the achievement degree information is not lower than the reference level, it is possible to provide a practice word having the same affinity and different initial or neutral to the learner terminal.

(G) receiving a conversation information request from the learner terminal; And (h) extracting the conversation information corresponding to the conversation information request and providing the extracted conversation information to the learner terminal, wherein the step (h) includes the steps of: extracting the conversation information requested by the learner After confirming the achievement degree information, the form of providing the conversation information may be differentiated according to the achievement degree information.

The method may further include recommending face-to-face learning by providing high-score learner information adjacent to the location of the low-score learner through the achievement degree information.

According to the Korean learning service providing system and the method of operating the Korean learning service according to the present invention, it is possible to accurately determine the pronunciation ability of the learner by reflecting the characteristics inherent in Korean language, and to provide subsequent services accordingly.

That is, it is possible to provide Korean language learning by providing learning information according to the progress to the learner terminal, but it is possible to make correction by confirming whether the word or sentence suggested by the system is uttered properly.

In this case, since the learner pronunciation information only determines the accuracy of the pronunciation of the pedestrian when judging the achievement level, the accuracy can be determined by a simple algorithm. In other words, due to the nature of Korean, it is possible to communicate if the pronunciations are correct even if the pronunciations are not accurate. However, the accuracy of the rest of speech The analysis method is very simple, but the pronunciation accuracy can be correctly judged.

Furthermore, it is possible to repeat learning by providing exercise words on the basis of the performance information scored through pronunciation accuracy determination, so that it is possible to perform correction learning on weak words or sentences. At this time, the exercise word is made such that the whole or a part of the support of the weak word is the same, and the remaining phonemes are variously changed, thereby mastering the support word. In particular, after dividing groups of exercises by their characteristics, it is possible to maximize the learning efficiency by concentrating only on the groups that the learners can pronounce well.

Also, after the achievement has been determined, the learner can be provided with conversational information that can be used in a specific place or situation. In this case, by providing conversational information such as voice, text, and pronunciation symbol in accordance with the learner's achievement information, it is possible to reduce the amount of data usage by providing all necessary information to the low-score learner and providing unnecessary information to the high-score learner.

In addition, it is possible to improve the learning efficiency by making face-to-face learning possible by recommending a high-score learner to the low-score learner. In this case, the recommended high-score learner is not only close to the location but also has the same nationality or uses the same mother tongue, so that the face-to-face learning can be performed more intimately and the learning efficiency can be further improved. In addition, high-scoring learner who has instructed low-scoring learner can benefit from the discount of tuition fee, thereby inducing active promotion of face-to-face learning.

1 is a block diagram for explaining a Korean language learning service providing system according to an embodiment of the present invention;
2 is a flowchart for explaining an example of a process of analyzing learner pronunciation information among operating methods of a Korean learning service providing system according to an embodiment of the present invention.
3 is a diagram for explaining an example of waveform analysis;
4 is a flowchart for explaining an example of a process of providing practice words among operating methods of a Korean language learning service providing system according to an embodiment of the present invention.
5 is a diagram for explaining an example of providing an exercise word;
FIG. 6 is a flowchart illustrating an example of a process of providing conversation information based on a location and an achievement level of a learner among operating methods of a Korean learning service providing system according to an embodiment of the present invention.
FIG. 7 is a flowchart for explaining an example of a procedure for recommending another learner among a method of operating a Korean language learning service providing system according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. However, some configurations which are not related to the gist of the present invention may be omitted or compressed, but the configurations omitted are not necessarily those not necessary in the present invention, and they may be combined by a person having ordinary skill in the art to which the present invention belongs. .

In addition, each component, server, and system described below does not necessarily have to be an independent component or server that performs each function, but may be implemented as one or more programs or one or more servers or a collection of one or more systems, May be shared.

FIG. 1 is a block diagram for explaining a Korean language learning service providing system according to an embodiment of the present invention. 1, a Korean language learning service providing system 100 according to an embodiment of the present invention is connected to a plurality of learner terminals 210, 220, and 230 and a teaching terminal 310 through a wire / wireless network to transmit / receive information required for Korean language learning . The Korean language learning service providing system 100 may be a system 100 operated by an educational institution such as a school, a school, or an educational broadcasting station, or a system 100 for providing a learning service commissioned from the educational institution.

The learner or the instructor can access the Korean language learning service providing system 100 through a web page access method through a terminal such as a desktop computer or a notebook computer to transmit and receive the learning information or to install the learning information in a terminal such as a smart phone or a tablet PC And can transmit and receive learning information by driving a dedicated application. That is, the Korean language learning service providing system 100 provides various interfaces through which learning information can be transmitted and received through a web page or a dedicated application. Hereinafter, the Korean learning service providing system 100 shown in FIG. 1 will be referred to as 'system 100' for the sake of convenience, and a learner will be assumed to receive learning information through a dedicated application installed on a smartphone or a tablet PC I will explain it later.

1, the system 100 includes a communication processing unit 105, a member information operation unit 110, a member recommendation unit 115, a learning information operation unit 120, a waveform analysis unit 125, (130), a membership information D / B (135), and a learning information D / B (140).

The communication processing unit 105 is provided for connecting a communication channel with the learner terminals 210, 220, and 230 or the instructor terminal 310 through a wire / wireless communication network, providing various user interfaces, and processing information transmission / reception.

The member information operation unit 110 processes member subscription by receiving the member information from the learner terminals 210, 220 and 230 and storing the member information in the member information D / B 135, or the learning progress of the registered learner, And so on.

Membership recommendation section 115 is designed to provide learner with low academic achievement with high academic achievement to support Korean learning in offline meeting. When the member recommendation section 115 recommends another member, it is performed by referring to the achievement information, the residence (current position), and the nationality or mother tongue information among the learners.

The learning information operating unit 120 stores learning information or standard pronunciation information input through an operator terminal (not shown) in the learning information D / B 140, and when requested by the learner, To the learner terminals 210, 220, and 230, and to receive and store learner pronunciation information acquired by the learner terminals 210, 220, and 230. Also, the learning information operation unit 120 is provided for providing the learner terminals 210, 220, and 230 with exercise words (or exercise phrases) stored in the learning information D / B 140 according to the learner's achievement information.

The waveform analyzer 125 is provided for analyzing the waveform of the learner's pronunciation information received from the learner terminals 210, 220 and 230. The achievement level determining unit 130 compares the waveform of the learner's pronunciation information received from the learner terminals 210, And stores it in the member information D / B 135 in a matching manner. The analysis algorithm of the waveform analyzer 125 and the performance determination of the achievement level determining unit 130 will be described in detail below.

The membership information D / B 135 is a storage medium for storing the member information received from the learner terminals 210, 220 and 230. The member information stored in the membership information D / B 135 may include personal information of each learner, such as ID, name, residence (current position), nationality, mother tongue, The pronunciation information of learners uttered by the learner, and the like can be stored in a matching manner. Also, the achievement degree information determined by the achievement degree determination unit 130 may be stored in a matching manner.

The learning information D / B 140 is a storage medium that stores basic learning information, standard pronunciation information, exercise words, and painting information to be provided to a learner.

The basic learning information is the content for the Korean language learning to be provided to the learner terminals 210, 220, and 230 and output. The basic learning information may be provided in various forms such as a moving picture, an image, a voice, and a text. The basic learning information may be based on information previously stored in the learning information D / B 140, Lt; / RTI > Standard pronunciation information means words or sentences to be provided to a learner are stored in a sound source form.

The standard pronunciation information is used as basic data for determining the accuracy of the learner pronunciation information in the future, and may further include information such as a text (sound), a waveform, a mouth shape (image or moving picture) Also, the standard pronunciation information is preferably a sound source recorded by a Korean person, but it may be a sound source information recorded by a foreigner having various nationalities in Korean.

Practice words are practice word information to be provided to learners with low pronunciation accuracy. Practice words can be words that have the same but different initial and neutrality to enhance the accuracy of certain supporting words. These exercises can be stored in packages of different words with the same longitudinality and with the same initiality and neutrality, and can be stored in various forms such as video, image, voice, and text as well as basic learning information.

Conversational information is information about sentences that can be used in a specific place or situation. The conversation information may be stored in a state where the sound source, the text (pronunciation), the pronunciation symbol, etc. are matched, and may further include moving image or image information if necessary.

The basic learning information, the standard pronunciation information, the practice word, and the pictorial information described above are classified according to the necessity and the providing form, and the respective information is shared and can be extracted and used as needed.

The learner terminals 210, 220 and 230 are terminals used by a learner who wishes to receive a Korean language learning service. Various types of terminals such as a desktop computer, a notebook computer, a smart phone, and a tablet PC can be applied. (210) as a reference.

The first learner terminal 210 includes a communication unit 211, a learning information processing unit 212, a screen output unit 213, a key input unit 214, a voice output unit 215, a voice input unit 216, (217).

The communication unit 211 is provided for connecting a communication channel for information transmission / reception with the system 100.

The learning information processing unit 212 is provided for outputting the learning information received through the communication unit 211 through the screen output unit 213 and the audio output unit 215. [ The learning information processing unit 212 is provided for transferring the member information and learner's pronunciation information input through the key input unit 214 or the sound output unit 215 to the system 100. [

The screen output unit 213 is provided to display a user interface for inputting member information and the learning information, and the key input unit 214 is provided for inputting member information, a learning information request command, and the like. Here, the screen output unit 213 and the key input unit 214 may be integrally manufactured by a touch screen method.

The voice output unit 215 may be a speaker or the like for outputting a reproduced sound source and the voice input unit 216 may be a microphone or the like for inputting pronunciation information of a learner.

Meanwhile, the position information acquisition unit 217 acquires the current position of the first learner terminal 210 in real time, and a GPS module or the like can be applied.

In addition, the first learner terminal 210 may further include a camera (not shown) for photographing the learner.

The learning process through the first learner terminal 210 can be performed through the dedicated application installed in the system 100 or through a separate app store. That is, the learning information processing unit 212 of the first learner terminal 210 may be a software-like configuration that is processed through a dedicated application installed in the first learner terminal 210, rather than a hardware configuration.

The descriptions of the components of the Korean language learning service providing system 100 and the learner terminals 210, 220 and 230 through the determination of the pedestrian accuracy according to the embodiment of the present invention described above will be described with reference to the Korean language learning From the description of how the service providing system 100 is operated.

FIG. 2 is a flowchart for explaining an example of a process of analyzing learner pronunciation information among operating methods of a Korean language learning service providing system according to an embodiment of the present invention. That is, the basic learning information is outputted from the learner terminals 210, 220 and 230 connected to the system 100 shown in FIG. 1, and the pronunciation information of the learner is obtained and the accuracy is determined. It is assumed that the membership application procedure has already been completed while the dedicated application is installed in the first learner terminal 210 prior to the description.

First, the first learner accesses the system 100 through the first learner terminal 210, receives the member authentication through a predetermined login procedure, and then requests basic learning information (S205). The learning information operating unit 120 extracts the basic learning information from the learning information D / B 140, and transmits the basic learning information to the first learning information processing unit 105 via the communication processing unit 105. [ To the learner terminal 210 (S210).

The basic learning information can be provided in various forms such as a moving image, an image, a sound source, and text. At this time, the learning information operating unit 120 preferably extracts and provides the learning information corresponding to the learning progress of the first learner. The membership information managing unit 110 confirms the learning progress of the first learner through the membership information D / B 135, and the learning information management Unit 120 extracts learning information according to the learning progress confirmed by the member information operation unit 110. [

The learning information processing unit 212 outputs the basic learning information through the screen output unit 213 and the audio output unit 215. The basic learning information is output from the communication unit 211 of the first learner terminal 210. [ Thus, the first learner can proceed the Korean language learning according to the progress. At this time, there may be a speaking process during the learning process. In other words, follow the suggested word or sentence and practice pronunciation.

In the case of off-line teaching or one-on-one remote lecture, the instructor can instantly correct the learner's pronunciation by pointing out the wrong part, but it is difficult to immediately grasp the learner's speaking ability in the online learning service. To this end, in the present embodiment, the pronunciation information obtained by the learner is acquired, and the achievement degree can be determined by comprehensively digitizing the pronunciation accuracy, the height, and the pronunciation speed.

That is, when the first learner terminal 210 outputs a spoken word (or sentence), the first learner speaks as it is. The learning information processing unit 212 transmits the learned learner's pronunciation information to the system 100 through the communication unit 211 <S220> when the first learner's pronunciation is acquired through the speech input unit 216 (S215).

When the learner's pronunciation information is received in the communication processing unit 105 of the system 100, the learning information operating unit 120 stores the learner's pronunciation information in association with the member information of the first learner. The waveform analyzing unit 125 analyzes the waveform of the learner's pronunciation information And the achievement level determining unit 130 determines the accuracy of the waveform based on the longitudinal waveform extracted by the waveform analyzing unit 125 (S230).

3 is a diagram for explaining an example of waveform analysis. That is, when the speaking learning information provided to the first learner terminal 210 is 'Hello', the waveform analyzer 125 analyzes the learner's pronunciation information (sound source) obtained by the first learner as it is.

The waveform analyzer 125 separates the wavelength of sound into frequency bands using a frequency conversion function. The frequency conversion function used in this case may be a Fast Fourier Transform (FFT) function. By using the fast Fourier transform function, the wavelength of sound is separated into the high frequency domain and the low frequency domain, and the values in the high frequency domain and the low frequency domain are known. In addition, since the fast Fourier transform results in a complex number, the values of the real part and the imaginary part are separated. Analysis of the combination of these two values can lead to accurate phoneme analysis. Also, it is possible to remove the noise of the acquired sound by using the fast Fourier transform function. In the waveform analyzing unit 125, various methods other than the above-described examples can be applied to the algorithm for separating the preliminary, neutral, and longitudinal waveforms.

The waveform analyzer 125 extracts only the middle portion of the waveform from the waveform analyzer 125. The performance analyzer 130 extracts only the middle portion from the waveform analyzer 125, And compares the end portion extracted in the learning information D / B 140 with the standard pronunciation information stored in the learning information D / B 140 to determine the accuracy of the support pronunciation S230. The determination of the accuracy of the pedestal pronunciation can be made through waveform comparison. For example, in the learner pronunciation information, the coincidence rate is quantified (scored) by comparing the 'b' of 'inside' and the 'o' of 'ni' with the waveform information of the corresponding support of the standard pronunciation information. The comparison of the waveform information for the pedestal pronunciation can be made by comparing feature points such as amplitude, length, angle, maximum value, and energy amount. Also, the achievement level determining unit 130 can score the overall syllable height and the pronunciation speed together with the determination of the accuracy of the support pronunciation. Of course, the accuracy determination is not performed for syllables without a pedestal.

The process of comparing the learner's pronunciation information with the standard pronunciation information and determining the accuracy by the waveform analyzing unit 125 and the performance level determining unit 130 may be similar to the first embodiment. However, the algorithm for determining the accuracy in the performance determining unit 130 is fundamentally different.

Conventional speech recognition technology compares a waveform of a speech signal with a codebook stored in a memory to extract the most similar words to find out what words the speaker has spoken. In other words, the speech signal must be compared to the entire word (or syllable or sentence) of the memory in the absence of any criterion, which makes the algorithm very complex and takes a long time to analyze.

However, in the present invention, it is not an object to find out what the learner is talking about, but it is only necessary to grasp whether or not the pronunciation of the presented word (hello) provided in the system 100 is good (accuracy determination). The algorithm is very simple and the analysis time is short because it does not judge the whole word of the presentation, but it only needs to judge the accuracy of the pronunciation of the pedestal (except for the initial and neutral).

This is based on the unique characteristics of Korean. In other words, due to the nature of Korean, conversation is possible if the accuracy of the pronunciation is high even if the pronunciations or neutral pronunciations are insufficient. For example, 'Accuracy' is not 'pronounced', but if you pronounce it correctly, you can see what you want to say. On the contrary, the prefix pronounced 'preposition' is pronounced correctly, but it is very difficult to find its meaning if the foot is wrong.

Therefore, it can be seen that the pronunciative ability is high when the accuracy of pronunciation of the support (consonant) is judged in consideration of the characteristics inherent in Korean. Accordingly, it is unnecessary to use a complicated algorithm that analyzes all the sentences, words, and syllables as in the existing speech recognition technology and finds out which one of the words stored in the memory is stored. The algorithm is very simple and the analysis time is not long.

In addition, when comparing the initial, neutral, and ending points using the conventional speech recognition technology, the comparison of the longitudinality is only a part of the judgment, and the accuracy is high even if the pronunciations are not good. However, as described above, even if the pronunciations of the chorus and neutrality are accurate, if the pronunciations of the utterances are indefinite, the pronunciation ability is not high. Therefore, the entire syllable judgment method of the conventional speech recognition technology does not meet the object of the present invention.

Of course, the achievement level determining unit 130 may determine the achievement level by determining only the accuracy of the pronunciation of the base (pronation), and then calculating the accuracy by comparing the standard pronunciation information with the standard pronunciation information, The final achievement can be determined in a manner that the weight is increased.

(Score) is determined by the achievement degree determining unit 130 of the system 100, the achievement degree information is stored in the member information managing unit 110 in association with the first learner's membership information (S235) do.

Further, the determined performance information may be provided to the instructor terminal 310, and the professor may perform the pronunciation correction learning based on the achievement degree information to the first learner terminal 210 side. Of course, the performance information is also provided to the first learner terminal 210 so that the learner himself can determine the level of the pronunciation accuracy of the user.

In the above embodiment, the achievement degree is determined based on the pronunciation of the support in the system 100. However, the achievement degree is first determined based on the support pronunciation in the system 100, and secondarily, After evaluating the pronunciation ability, the instructor may be provided with the evaluation result, and then the correction learning may be performed.

The learning information is provided to the learner through the process of FIG. 2, and various personalized learning services can be provided based on the determination of the achievement level. Hereinafter, an example in which the learning service is provided using the performance information evaluated based on the pronunciation of the backing through Figs. 4 to 7 will be described.

4 is a flowchart for explaining an example of a process of providing practice words among operating methods of a Korean language learning service providing system according to an embodiment of the present invention.

That is, the learner pronunciation information of the first learner is obtained through the process of FIG. 2, the achievement degree information is analyzed by analyzing the learned learner pronunciation information, and the result is stored in the member information D / B 135 (S235) The operation unit 120, in cooperation with the member information operation unit 110, checks whether the achievement degree information of the learner is equal to or greater than the standard (S405). If the achievement information is above the standard, there is no need to learn additional words. On the other hand, if the performance information does not meet the criteria (S410), the learning information operating unit 120 extracts a practice word from the learning information D / B 140 and provides the extracted exercise word to the first learner terminal 210. [

At this time, the practice words are not provided at random, but are provided as words that can be used to familiarize the subject. For example, it is assumed that the presentation word provided to the first learner terminal 210 in the process of FIG. Practice words related to 'imperial room' are stored in groups according to a specific rule in the learning information D / B 140.

An example of a state where practice words are stored for each group with characteristics is shown in FIG.

Group 1 is the same as the progeny [ㅇ,]] of the proposed word 'empire', neutrality is [ㅘ, ㅣ], and the beginning is the changed words. For example, it can be a word such as royal family,

Group 2 is the same as the progeny [ㅇ,]] of the propositional word 'empire', neutrality is [a, ㅗ], and prefixes are changed words. For example, it can be a word such as tungdol, gangil.

Group 3 is the same as the progeny [ㅇ,]] of the proposed word 'empire', neutrality is [ㅗ, ㅜ], and initials are changed words. For example, words such as caves, animals, and poems.

Group 4 is the same as the prologue [ㅇ,]] of the proposal word 'empire', the neutrality is changed, and the prefix is [heh, g]. For example, words such as conduct, red room, and the like.

Group 5 is the same as the progeny [ㅇ,]] of the proposal word 'empire', neutrality is changed, and prefix is [ㅁ, ㅁ]. For example, it can be a word such as a cold or tide.

Group 6 is the same as the prologue [ㅇ,]] of the proposal word 'empire', the neutrality is changed, and the prefix is also changed. Such as jaundice, droplet, birthday, alignment, and the like.

Group 7 is only the first syllable [ㅇ] of the proposed word 'empire' [ㅇ], and the rest are all changed words. For example, study, video, happiness, mullet, and the like.

Group 8 is the same as the second syllable []] of the proposed word 'imperial', and all the rest are the changed words. For example, it can be a word such as a gil, a transfer.

When we look at the characteristics of each group, at least one of the consonants of the 'empire', which is a common word, consists of the same words. This is to enable mastery pronunciation of the word (imperial room) with insufficient pronounciation.

Also in groups 1 to 6, the entire progeny [ㅇ,]] of the proposed word 'empire' is the same, and the rest are words that have been changed with certain rules. In other words, groups 1 to 6 are the words to make [○,]], which is a poor base, to practice at the same time.

On the other hand, Group 7 and Group 8 are the same as the first syllable [ㅇ] or the same as the second syllable []]. In other words, groups 7 and 8 do not want to practice two bases at once, but want to master only one bases. Therefore, the whole syllable of a word can also be changed.

In this way, in a state where words of the learning information D / B 140 are grouped and stored, the learning information operating unit 120 of the system 100 generates a plurality of groups of words for pronunciation correction of 'imperial room' , One or more words representing each group are extracted. For example, you can extract 'royal, tungdol, cave, deceit, fire, jaundice, study, gil' for each group. The reason for extracting representative words one by one for each group is to understand which pronunciation is better for words having certain characteristics. Of course, it is desirable to extract and provide at least three words for each group in order to obtain accurate characteristics, and then find out groups with a higher average of the accuracy score.

When the representative word is extracted from each group, the learning information operating unit 120 provides the first word to the first learner terminal 210 in connection with the communication processing unit 105 (S420).

When the representative words are received in the first learner terminal 210, the learning information processing unit 212 processes the representative words so that the first learner can perform the speech exercises through the representative words.

However, pedestrian practice of representative words in each group may not be effective. That is, even if the first learner is pronouncing the same pedestal, the pronunciation may be inaccurate for a word having a specific rule, and the pronunciation may be correct for a word having another rule. However, if the pronunciation is incorrect, I can not expect much.

Moreover, there are separate characteristic words that can be easily digested by each person. Therefore, practicing the same words to all learners is not helpful at all to improve pronunciation. To this end, the present invention grasps how a learner can easily digest a word having a characteristic, and enables intensive speech exercises with words of the group having the characteristic.

That is, when the first learner 210 outputs the representative words for each group and the first learner inputs the pronunciation information for the representative words through the speech input unit 216 (S425), the learning information processing unit 212 The learner's pronunciation information is transmitted to the system 100 (S430).

Then, the waveform analyzer 125 and the achievement level determining unit 130 of the system 100 determine the accuracy of the trailing words for each representative word in the above-described manner (S435), and determine the representative word having the highest accuracy (S440). If three representative words are extracted for each group, check which group has the highest average accuracy score.

For example, 20 points for group 1, 30 points for tungsten for group 2, 50 points for cave 3 for group 3, 30 points for conduct of group 4, 20 points for group 5, , The jaundice accuracy of the group 6 'jaundice' is 30 points, the group 7 'study' is 80 points, and the group 8 'gilil' is 40 points. At this time, even though the words such as groups 1, 2, 4, 5, 6 are the same pedestal, the learner has difficulty pronouncing because of the relationship between other beginners and neutral. In other words, it is difficult to correct the pronunciation of words in groups that share the characteristics, no matter how much practice is given.

Accordingly, the learning information operating unit 120 of the system 100 confirms representative words having the highest accuracy (or a certain score or more) among the representative words, and recognizes the words of the group to which the representative word belongs as a practice word (S445) and provides it to the first learner terminal 210 (S450). In other words, the words of group 7 such as study, video, happiness, and mullet, which were judged to have the highest accuracy for the representative word, are provided, or the words of group 3 and group 7 which have obtained a certain score (for example, 50 points) Words can be learned by repeating the words (group 3 and group 7 words) that are able to perform well, such as 'ㅇ and 인', which are poor pronunciation of the first learner.

Of course, while providing the practice word, the pronunciation information of the learner may be acquired again to determine whether the achievement is improved in real time and to store it.

According to the process of FIG. 5, it is possible to provide learning words on the basis of the performance information scored through the pronunciation accuracy determination, so that it is possible to repeatedly learn, thereby enabling correction learning on weak words or sentences. At this time, the exercise word is made such that the whole or a part of the support of the weak word is the same, and the remaining phonemes are variously changed, thereby mastering the support word.

Especially, even if the same pronunciation is pronounced, the learners may have different learning ability depending on how the remaining chorus and neutrality are combined. However, difficult words can not help to improve pronunciation ability.

In other words, in the present invention, not all the learners are provided with the same practice word, but the test is first performed with words having unique characteristics, and then the learners can concentrate on the words of the group that can be easily digested. , It is possible to maximize learning efficiency.

6 is a flowchart illustrating an example of a process of providing conversation information based on a location and an achievement level of a learner among operating methods of a Korean learning service providing system according to an embodiment of the present invention. That is, the performance information of the learner is determined in the process of FIG. 2 and stored in the member information D / B 135, and then the service is provided.

After the learner has the learner terminals 210, 220 and 230 and visits a specific place, the learning information processing unit 212 inputs a command for requesting the conversation information through the key input unit 214 by driving the dedicated application, Acquires position information of the learner terminals 210, 220 and 230 in association with the acquiring unit 217 (S610). Then, the learner information processing unit transmits the conversation information request information including the position information acquired in association with the communication unit 211 to the system 100 (S615).

Upon receipt of the conversation information request from the communication processing unit 105 of the system 100, the member information operation unit 110 confirms the achievement degree information of the learner requesting the conversation information through the member information D / B 135 <S620> .

Thereafter, the learning information operating unit 120 extracts (S625) <S625> the conversation information based on the achievement degree information and the position information of the learner through the learning information D / B 140, and transmits the extracted conversation information through the communication processing unit 105 To the learner terminals 210, 220, and 230 (S630).

The conversation information based on the location information may be the conversation information used by the bank when the learner visits the bank or the conversation information which can be used in the cafe when the learner visits the cafe. Here, the location information may be coordinate information obtained through the GPS module in the location information acquisition unit 217 of the learner terminals 210, 220, and 230, but may be a method of acquiring a beacon signal installed at a site such as a bank or a cafe.

In addition, when the learner information operating unit of the system 100 provides the conversation information, it is preferable to provide differential information based on the achievement information. For example, even if the same content information is provided, the first learner terminal 210 having low achievement information (score) is provided with the conversation information including all the voice, text, and pronunciation symbols, Only the voice and text information is provided to the third learner terminal 220 and only the text information is provided to the third learner terminal 230 having high achievement information.

According to the achievement information, the way of providing the information of the conversation differs in order to reduce the data usage by not providing the unnecessary information as much as possible. In other words, for learners with low achievement level information, conversation information including voice, text, and phonetic symbols are provided so that they can be voiced after the preliminary exercise. The learners who have high achievement level information can use speech information It is enough to provide conversation information in the form of text because it is possible to communicate sufficiently.

It is preferable that the audio information included in the picture information is information recorded by a local person who uses the mother tongue used by the learner. In other words, if Korean is recorded by a person who uses the same accent as himself, he / she will be more familiar with it. Therefore, the conversation information stored in the learning information D / B 140 includes information recorded by people using various native languages. When the learner terminals 210, 220, and 230 request the conversation information, the member information operating unit 110 The linguistic information of the learner is checked together with the mother tongue information, and when the learning information operating unit 120 extracts the painting information, the recording information recorded by the person who uses the mother tongue can be extracted. Of course, when extracting pronunciation symbols, it is preferable to extract pronunciation symbols using the native language of the learner.

Meanwhile, although FIG. 6 exemplifies the provision of the conversation information through the acquisition of the location information, the user may select or input the conversation information that can be used in a specific place or situation, rather than acquiring the location information. That is, the learner terminals 210, 220 and 230 may be a non-movable desktop computer, or may want to confirm the conversation information in advance while the learner is located at a place other than the place. In this case, when the learner terminals 210, 220, and 230 request the conversation information, information on a specific place or situation can be selected or input, and the system 100 provides conversation information corresponding to the selected place or situation will be. Of course, in this case, it is the same to provide differential information on the basis of the achievement information.

7 is a flowchart for explaining an example of a process of recommending another learner among the method of operating the Korean language learning service providing system according to the embodiment of the present invention.

It is most efficient for learners to learn one by one, face to face with learners. However, the number of teachers is very small compared to the number of learners, and face distance is physically far away, so face-to-face learning is practically impossible. For this purpose, in this embodiment, a high score learner is recommended to a low score learner, and face-to-face learning is made possible through meeting.

That is, when the achievement degree information of the learners is determined through the process of FIG. 2 and stored in the membership information D / B 135, the member recommendation unit 115 of the system 100 transmits the membership information to the member information management unit 110 And confirms whether the achievement level information of the learner is equal to or higher than the standard level (S705). If it is above the standard, <S710> will not require additional learning.

On the other hand, if the achievement information does not meet the criteria (S710), the member recommendation unit 115 confirms the location information of the low-score learner in conjunction with the member information management unit 110 (S715). It is preferable that the location information is the residence of the learner stored in the member information D / B 135. However, the location information may be requested by the learner terminals 210, 220 and 230 to obtain the current location.

The member recommendation unit 115 of the system 100 extracts the high score learner information within the standard distance from the position of the low score learner in step S720 in conjunction with the member information operation unit 110, And provides the information of the high score learner to the learner terminals 210, 220, and 230 (S725).

Here, in order to provide high-score learner information, it should be possible only when the member allows the information disclosure in advance. Also, it is preferable that the member recommendation section 115 recommends a learner who uses the same nationality or the same mother tongue as the low-score learner, rather than simply recommending the high-score learner in the vicinity.

In other words, if high-scoring learner information is provided to low-scoring learner, face-to-face learning is possible through learner-to-learner interaction. If the country of origin uses the same or the same native language, familiarity can be felt, .

In addition, high-scoring learners' investment of time to support low-scoring learner's learning can be seen as a kind of time lecturer concept, so the benefits should be provided. That is, if the high score learner information is provided to the low score learner &lt; S725 &gt; to perform the face learning, and if the two learner are authenticated through the transfer of the photograph information or the like, ) Can give some discounts on tuition fees for high-scoring learners, or provide coupons for popular culture tickets (concert tickets, etc.).

Meanwhile, after the process of FIG. 7, the system 100 can continuously manage the low-score learner and the high-score learner to enable additional services. For example, if face-to-face learning is achieved through recommendation of a high-score learner, the member information management unit 110 continuously monitors whether or not the achievement level information of the low-score learner is increased, and if the achievement degree information is confirmed to be high, And to increase the benefits (such as discounts on discounts and coupons) for high-scoring and high-scoring learners, and to recommend them to other low-scoring learners in the future.

As described above in detail, according to the Korean learning service providing system 100 and the method of operating the Korean learning service through the determination of the recognition accuracy according to the present invention, the pronunciation ability of the learner is accurately determined by reflecting the characteristic inherent to Korean, Service is possible.

That is, the learner terminals 210, 220, and 230 can provide learning information corresponding to the degree of progress so that the Korean language learning can be performed, but the system 100 can confirm that the word or sentence uttered by the system 100 is properly uttered.

In this case, since the learner pronunciation information only determines the accuracy of the pronunciation of the pedestrian when judging the achievement level, the accuracy can be determined by a simple algorithm. In other words, due to the nature of Korean, it is possible to communicate if the pronunciations are correct even if the pronunciations are not accurate. However, the accuracy of the rest of speech The analysis method is very simple, but the pronunciation accuracy can be correctly judged.

Furthermore, it is possible to repeat learning by providing exercise words on the basis of the performance information scored through pronunciation accuracy determination, so that it is possible to perform correction learning on weak words or sentences. At this time, the exercise word is made such that the whole or a part of the support of the weak word is the same, and the remaining phonemes are variously changed, thereby mastering the support word. In particular, after dividing groups of exercises by their characteristics, it is possible to maximize the learning efficiency by concentrating only on the groups that the learners can pronounce well.

Also, after the achievement has been determined, the learner can be provided with conversational information that can be used in a specific place or situation. In this case, by providing conversational information such as voice, text, and pronunciation symbol in accordance with the learner's achievement information, it is possible to reduce the amount of data usage by providing all necessary information to the low-score learner and providing unnecessary information to the high-score learner.

In addition, it is possible to improve the learning efficiency by making face-to-face learning possible by recommending a high-score learner to the low-score learner. In this case, the recommended high-score learner is not only close to the location but also has the same nationality or uses the same mother tongue, so that the face-to-face learning can be performed more intimately and the learning efficiency can be further improved. In addition, high-scoring learner who has instructed low-scoring learner can benefit from the discount of tuition fee, thereby inducing active promotion of face-to-face learning.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, And additions should be considered as falling within the scope of the claims of the present invention.

100: Korean learning service provision system
105:
110: Member Information Operations Department
115: Member Recommendation Department
120: Learning Information Operations Department
125: waveform analysis unit
130: Achievement level judge
135: Member Information D / B
140: Learning Information D / B
210: first learner terminal
211:
212: Learning information processor
213:
214:
215: Audio output unit
216:
217: Position information obtaining unit
220: second learner terminal
230: Third learner terminal
310: a teaching terminal

Claims (10)

Learning information D / B storing standard pronunciation information and practice word;
A communication processing unit for transmitting and receiving information to and from the learner terminal;
A waveform analyzer for analyzing a waveform and extracting a consonant after receiving learner pronunciation information from the learner terminal through the communication processor;
An achievement level determining unit that compares the extracted pedestrian with the standard pronunciation information to determine accuracy of pronunciation of the pedestrian (consonant) sound and scoring it to calculate achievement information;
A member information management unit for storing the achievement degree information in the member information D / B; And
And a learning information operation unit for providing customized learning information to the learner terminal based on the achievement degree information,
The exercise words stored in the learning information D / B are groups in which the words having the same consonant,
Wherein the waveform analyzing unit extracts only the waveform information of the utterance after separating each syllable of the learner's pronunciation information by the phoneme,
The achievement level judging unit compares the waveform information about the utterance of the previously presented word among the standard pronunciation information and the waveform information about the utterance of the learner pronunciation information extracted by the waveform analyzing unit through the learning information operating unit,
The learning information operating unit confirms whether the achievement degree information is equal to or more than the reference value and if it is confirmed that the achievement degree information is not equal to the reference value, Extracting at least one word representing each group of the words of the group and providing the selected word to the learner terminal,
When the learner inputs learner's pronunciation information for each representative word in the learner terminal, the learner's pronunciation information received from the waveform analyzing unit and the achievement degree determining unit through the communication processing unit Determining the longitudinal accuracy for each representative word included,
The learning information operating unit extracts words of a group to which a representative word having the highest accuracy or a predetermined score or more among the representative words is extracted and provides the learned word to the learner terminal so that the learner can repeatedly utter words Wherein the language learning service providing system comprises:
delete delete The method according to claim 1,
When receiving the conversation information request from the learner terminal through the communication processing unit,
Wherein the learning information operating unit extracts the conversation information corresponding to the conversation information request and provides the extracted conversation information to the learner terminal and confirms the achievement degree information of the learner requesting the conversation information through the membership information D / B, And provides a different form of presentation of the conversation information.
The method according to claim 1,
And recommending face-to-face learning by providing high-score learner information adjacent to the location of the low-score learner through the achievement degree information.
(A) providing learning information including standard pronunciation information to a learner terminal in a Korean learning service providing system (hereinafter referred to as "system");
(B) receiving learner pronunciation information generated by the learner terminal in response to the learning information;
(C) analyzing a waveform of the learner's pronunciation information and extracting a consonant;
(D) comparing the extracted pedestal with standard pronunciation information to determine the accuracy of pedestal pronunciation and scoring it to calculate achievement information;
(E) storing the achievement degree information in the member information D / B; And
And (f) providing customized learning information to the learner terminal based on the achievement information,
In the step (c), phonemic splitting is performed on each syllable of the learner's pronunciation information as a preliminary, a neutral, and a longitudinal,
The step (d) scales the coincidence rate by comparing the waveform information about the utterance of the word previously presented through the step (a) and the waveform information about the longevity of the learner pronunciation information extracted in the step (c)
If it is determined that the achievement level information is not lower than the standard level, the system determines that the learning information D / B is the same as the support level, Extracting at least one word representative of each group among the words of the group having the same sentence as the presented word but having a change in the sentence and the sentence, and providing the extracted word to the learner terminal,
When the representative word for each group is output from the learner terminal and the learner inputs learner's pronunciation information for each representative word, the system determines the longitudinal accuracy for each representative word included in the learner's pronunciation information f-1); And
(F-2) of extracting words of a group to which representative words having the highest accuracy or having a certain score or more are included in the system and providing the words to the learner terminal,
Wherein the learner is allowed to repeatedly utter words that the learner can perform well.
delete delete The method according to claim 6,
(G) receiving a conversation information request from the learner terminal; And
And (h) extracting the conversation information corresponding to the conversation information request and providing the extracted conversation information to the learner terminal,
Wherein the step (h) comprises the step of providing the Korean language learning service after confirming the achievement degree information of the learner requesting the conversation information through the membership information D / B, and providing the degree of providing the conversation information according to the achievement degree information. How the system works.
The method according to claim 6,
And (i) recommending face-to-face learning by providing high-score learner information adjacent to the location of the low-score learner through the achievement degree information.

Images