KR102372070B1 - Apparatus and method for foreign language conversation learning - Google Patents

Abstract

A plot-based dialog model generator for generating a learning conversational conversation on a topic to learn foreign language conversation according to the present invention in the form of a plot-based conversational model, plot-based by automatically changing the learning conversational conversation to knowledge usable in a destination-oriented conversation system Conversation is conducted with the learner so as to correspond to the foreign language level information and progress information registered by the learner using the destination-oriented dialogue knowledge unit that creates the topic-specific dialogue knowledge DB and the information stored in the plot-based topic-specific dialogue knowledge DB, and the conversation It includes a plot-based conversation processing unit that provides accuracy for progress results.

Description

Foreign language conversation learning apparatus and method {APPARATUS AND METHOD FOR FOREIGN LANGUAGE CONVERSATION LEARNING}

The present invention relates to an apparatus and method for learning foreign language conversation.

Conversational learning apparatuses for foreign language education currently being released give the user a role in a conversation selected as learning content and repeatedly practice the conversational expression of the role. However, the method of repeatedly learning a given sentence may be suitable for beginners, but in the case of intermediate and advanced learners, it is possible to lower the learning efficiency.

In addition, in the case of a voice communication system in which a machine plays the role of a native speaker to practice conversation, there is a problem in that it requires a lot of money and time because conversation knowledge and rules that require expert processing are required.

In addition, in the case of the above-described prior art, there is a problem in that conversational learning is conducted without considering the learning level and progress of the learner.

Accordingly, there is a need for an apparatus and method for learning foreign language conversation that can significantly reduce the construction time and cost for learning foreign language conversation and at the same time improve the learning efficiency of the learner.

An embodiment of the present invention enables conversation in a foreign language on a computer to match the learning level and progress of the learner through the topic conversation implemented as a plot-based dialog model, and provides evaluation information on the expression uttered by the learner to communicate in a foreign language A foreign language conversation learning apparatus and method capable of improving learning efficiency are provided.

However, the technical task to be achieved by the present embodiment is not limited to the above-described technical task, and other technical tasks may exist.

As a technical means for achieving the above-described technical problem, the foreign language conversation learning apparatus according to the first aspect of the present invention creates a plot-based conversation model that generates a learning conversation conversation on a topic to learn foreign language conversation in the form of a plot-based conversation model By automatically changing the learning conversational dialogue to knowledge usable in the destination-oriented dialogue system, the conversational knowledge unit toward the destination creates a conversational knowledge DB for each topic based on the plot, and the learner using the information stored in the plot-based conversational knowledge DB for each topic and a plot-based conversation processing unit that conducts a conversation with the learner to correspond to the foreign language level information and progress information registered by , and provides accuracy for the conversation progress result.

The plot-based dialog model generating unit may generate a role conversation with a learner in units of plots, and may generate the plot-based dialog model through a connection between the plots.

The plot-based dialog model generating unit may generate the plot unit to be composed of one starting utterance and a plurality of corresponding utterances.

The starting utterance may include a plot title as plot information, and the corresponding utterance may include a plot correspondence method for the plot title and the foreign language level information and progress information of a learner capable of the corresponding utterance as plot information.

The destination-directed dialogue knowledge generating unit may include: a utterance intention generation unit configured to receive the plot-based dialogue model and generate utterance intention information to be recognized on a computer through representative utterances of the starting utterance of the plot and the corresponding utterance; and a dialogue rule generator generating dialogue rule information using the connection information of the plot based on the speech intention information, and learning the speech intention so that representative speeches and paraphrases of the plot are recognized according to the speech intention information The intention of utterance may also include a learning unit.

The generated dialogue rule information and the speech intention learning data generated as a result of the learning may be stored in the plot-based dialogue knowledge DB for each topic.

The plot-based conversation processing unit may include: a speech recognition unit that recognizes the learner's utterance and converts it into text; an intention recognition unit that recognizes the converted text as the speech intention information; A dialogue management unit that generates a system response based on the speech information and the dialogue rule information, and a speech evaluation unit that compares and evaluates the learner's speech with the learning sentence for the speech intention information recognized by the speech intention recognition unit can do.

The speech intention recognition unit may recognize the speech intention information using information learned according to the foreign language level information of the learner, or recognize the speech intention information as the speech intention information only when the converted text matches the learning sentence. .

The conversation management unit may receive the foreign language level information and progress information of the learner, and select and provide a rule to be answered in the conversation rule information to correspond to the learner's level.

The plot-based conversation processing unit may further include a general conversation processing unit configured to provide a general conversation response to the utterances not classified through the utterance intention information in the utterance intention recognition unit by using the dialogue rule information.

In addition, the learning method in the foreign language conversation learning apparatus according to the second aspect of the present invention comprises the steps of: generating a conversation about a topic to learn foreign language conversation in the form of a plot-based conversation model; generating a plot-based conversational knowledge DB by topic by automatically changing the generated plot-based dialog model into knowledge usable in a destination-oriented dialog system; and conducting a conversation with the learner so as to correspond to foreign language level information and progress information registered by the learner using information stored in the plot-based conversational knowledge DB for each subject, and providing accuracy for the conversation progress result.

According to any one of the above-described problem solving means of the present invention, by constructing a dialogue on a foreign language learning topic with a plot-based dialogue model, even a person without knowledge of the dialogue system can easily and quickly build the corresponding system. .

In addition, since the constructed plot can be reused and changed as a composition in other learning topics, it is possible to practice conversation in various situations and topics when learning foreign language conversation.

In addition, as conversational learning is implemented in consideration of the learner's foreign language level and learning progress, the learner can proceed with conversational learning appropriate to their level even in a conversation composed of the same topic.

In addition, it not only provides foreign language learners with learning effects through the system and dialogue in line with the flow of topic conversation, but also provides various evaluation services such as content suitability evaluation, foreign language expression suitability evaluation, and task performance evaluation, allowing learners to You can recognize the evaluation of your own speech in the process of your own conversation.

In addition, the learning effect can be further improved by minimizing the foreign language expression suitability information delivered to the foreign language learner during the conversation, allowing them to focus on the conversation content, and delivering detailed feedback on the simplified evaluation information during the conversation after completing the conversation. .

1 is a block diagram of a foreign language conversation learning apparatus according to an embodiment of the present invention.
2 is a view for explaining the function of the foreign language conversation learning apparatus according to an embodiment of the present invention.
3 is a diagram for explaining an example of a plot.
4 and 5 are diagrams for explaining another example of a plot.
6 is a block diagram for explaining the function of a destination-directed conversational knowledge unit.
7 is a block diagram illustrating a function of a plot-based conversation processing unit.
8 is a flowchart of a foreign language conversation learning method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted.

In the entire specification, when a part "includes" a certain element, it means that other elements may be further included, rather than excluding other elements, unless otherwise stated.

The present invention relates to a foreign language conversation learning apparatus (1) and a method.

Conversational learning devices for foreign language education that are currently released give learners a role in a conversation selected as learning content, and allow them to practice the conversational expression of that role repeatedly. These conversational learning devices for foreign language education recognize the utterance of the learner using a voice recognition device, determine whether it is vocally consistent with the correct answer or whether the recognized result matches the content, and deliver the evaluation to the learner. make learning

In such existing foreign language conversation learning devices, the learner can practice uttering exactly the same given sentence.

However, repeating learning by following a given sentence may be a good learning method for some learners such as beginners, but there is a problem that it is difficult to expect fun and time-related learning effects for foreign language learning for intermediate and advanced learners.

Recently, in order to obtain the expected learning effect from conversation with an actual native speaker, conversation learning devices for foreign language education in which a machine plays the role of a native speaker using a voice communication system to practice conversation have been developed and studied.

However, chatbot systems capable of simple chat-chat are not effective in learning because only conversations for fun rather than learning are possible.

In order to solve this problem, there is a method of enabling a conversation suitable for a given learning topic using goal-oriented dialogue systems.

However, the destination-oriented dialogue system requires dialogue knowledge and rules established by dialogue processing experts to enable dialogue on learning topics, and this requires a lot of money and time.

Due to this problem, there are almost no foreign language conversation learning devices using a commercialized voice conversation system, but research and development for this are actively in progress.

In most foreign language conversation learning, it is difficult to expect a high learning effect because the learner only communicates with the computer in the range of conversations that can be realized with a computer on a given topic.

Therefore, a computer-learning response to the learner's foreign language expression is required. That is, an appropriate expression should be responded to in the next conversation, and if it is grammatically or expressively incorrect, it needs to be expressed during the conversation or presented after the conversation so that the user can know it.

On the other hand, the expected effect of a foreign language conversation learning device in which a computer replaces a human role is bound to increase. However, since the conversation topic implemented by the destination-oriented dialogue system is implemented as dialogues to complete a specific purpose, there are cases in which conversations about the performance of the purpose of the topic occupy a higher weight than the foreign language expressions that should be the learning target. many.

For example, suppose you set the topic to order food to learn “I would like to have~.”, an expression used when ordering food in English. The purpose of the food ordering subject is simply to ask what menu the computer wants and to ask the user "I would like to have~." If only using expressions, it would be more effective to simply follow a dialogue consisting of two dialogues rather than a complex destination-oriented dialogue system.

In conversational learning, learners using the destination-oriented dialogue system want more diverse conversations to be expressed along with the learning expression on the topic of ordering food, which is the example above, and to be implemented in the same way as the actual situation.

Therefore, on the topic of ordering food, simply adding only a menu request, adding today's recommended cooking question more complicatedly, adding even more billing for food expenses, complicatedly adding a dish to order If you ask how the ingredients and each taste are different, if you complain or praise the food you eat, if you order different types of sauces and side menus depending on the food you order, it needs to be added.

And according to the learning level and learning progress of the learner, all of these situation examples need to be implemented differently.

In this way, effective conversational learning becomes possible only when the learning level and progress of the learner are considered for learning topics implemented as a destination-oriented dialogue system.

In order to solve such a problem, the foreign language conversation learning apparatus 1 and method according to an embodiment of the present invention are based on a plot that can variously define a foreign language conversation learning topic according to a learner's level and progress process. It provides a dialogue model, enables conversation in a foreign language on the computer to match the learning level and progress of the learner through topic dialogue implemented as a plot-based dialogue model, and provides positive or negative evaluation information for the expression uttered by the learner. By providing it, it is possible to increase the efficiency of learning foreign language conversation.

Hereinafter, a foreign language conversation learning apparatus 1 according to an embodiment of the present invention will be described with reference to FIGS. 1 to 6 .

1 is a block diagram of a foreign language conversation learning apparatus 1 according to an embodiment of the present invention.

The foreign language conversation learning apparatus 1 according to an embodiment of the present invention includes a microphone 10 , a memory 20 , a processor 30 , and an output unit 40 .

The microphone 10 receives the user's utterance. In addition, the utterance input to the microphone 10 or an evaluation result thereof may be output through the output unit 40 including at least one of a speaker outputting a voice and a display outputted to a screen.

A program for providing foreign language conversation learning is stored in the memory 20 . In this case, the memory 20 collectively refers to a non-volatile storage device and a volatile storage device that continuously maintain stored information even when power is not supplied.

For example, the memory 20 may include a compact flash (CF) card, a secure digital (SD) card, a memory stick, a solid-state drive (SSD), and a micro SD card. NAND flash memory such as cards, magnetic computer storage devices such as hard disk drives (HDD), etc., and optical disk drives such as CD-ROMs and DVD-ROMs. can

The processor 30 executes the program stored in the memory 20 . A detailed process performed by the processor 30 will be described with reference to FIG. 2 .

2 is a view for explaining the function of the foreign language conversation learning apparatus 1 according to an embodiment of the present invention.

The foreign language conversation learning apparatus 1 according to an embodiment of the present invention includes a plot-based conversation model generating unit 100 , a destination-directed conversation knowledge unit 200 , and a plot-based conversation processing unit 400 .

The plot-based dialog model generating unit 100 generates a learning conversational conversation on a topic to learn foreign language conversation in the form of a plot-based conversational model so that it can be used in various conversation processing methods.

Referring to FIG. 2 , the plot-based dialog model format basically generates a role conversation between two people between the system and the learner in plot units, and the plot-based dialog model generating unit 100 connects these plot units to each other and Create a dialog model.

Each plot is an integral part of the dialogue plot, and includes plot titles that indicate some topic.

In addition, each plot may consist of one starting utterance and a plurality of corresponding utterances. In this case, the starting utterance is an utterance that starts the corresponding plot, and the corresponding utterance means a response to the starting utterance.

In addition, the starting utterance and the corresponding utterance may have representative utterance expressions and plot information according to the situation for each role.

The starting utterance may include a plot title as plot information, and the learner's foreign language level and progress information for the starting utterance may be automatically set to the minimum level and progress of the corresponding utterances.

Corresponding utterances may include a plot correspondence method for a plot title and foreign language level information and progress information of learners capable of corresponding utterances as plot information. The plot response method is provided to the learner at the time of learning conversation, and it is possible to determine whether the conversation task is performed by performing the task of which utterance should be made in a specific plot.

For each corresponding utterance, a different plot may be defined that develops into the next subsequent dialogue situation. At this time, if there is no other plot connection following the corresponding utterance, it means that the dialogue ends with the current corresponding utterance. Also, if there is another plot connection following the corresponding utterance but marked as “END”, only the starting utterance of the other plot is performed. and the conversation ends.

Through the connection of plots in this way, the conversation to learn conversation is composed. And one plot can be re-applied to multiple conversational learning conversations. In addition, the plot-based dialog model generating unit 100 according to an embodiment of the present invention may have a different plot configuration according to the level of the learner.

In addition, the plot-based dialog model generated by the plot-based dialog model generating unit 100 is implemented with paraphrases, which are other expressions representing the same meaning in addition to the representative utterances expressed in the starting utterances and corresponding utterances of each plot. can be Accordingly, according to an embodiment of the present invention, even when the learner makes various utterances, it is possible to understand and present various expressions to the learner, and also to enable the learner to utter various expressions having the same meaning.

The plot-based dialog model generated in this way will be described with reference to the examples of FIGS. 3 to 5 .

3 is a diagram for explaining an example of a plot. 4 and 5 are diagrams for explaining another example of a plot.

3 is an example of a plot with the plot title “Why I broke my promise”. The starting utterance is “S<1>”, and the corresponding utterance consists of 4 response utterances (U<1>, U<2>, U<3>, U<4>).

For each response utterance, the corresponding information in the given plot is described, the corresponding plot content of U<1> is “Plot: Reason for breaking the promise = something else happened”, the learner word level is '4 level', and the progress is at least '3 It indicates that it should have proceeded beyond -2'.

After the corresponding utterance U<1>, it is connected to another plot “Plot: Inquiry of Current Status”

Corresponding utterance U<2> is linked to another plot “Plot: Forgetting attention”, and “END”, which requests to end the conversation with only the beginning utterance, is described.

After the corresponding utterance U<3>, various plot connections are possible. When various plot connections are possible as described above, in actual conversational learning, a plot in which a corresponding utterance of a task given to the learner or a corresponding utterance suitable for the current learning process is followed may be selected.

“Unknown” in the corresponding utterance U<4> is an exception handling method for the case where the learner does not give the corresponding response defined in the corresponding plot. The processing method for “unknown” may proceed to the next plot with an appropriate response as in S<7>, or through an auto-response, that is, a response from a chatbot prepared for general conversation, as in S<9>.

4 and 5 are an example of a plot with a plot title of “Today’s Schedule”.

As in U<1> of FIG. 4, various examples can be used instead through the variable '@action'.

5 shows a notation method for a case in which the following plot changes according to the example as the values of the examples change despite the same corresponding utterances U<1> and U<2>.

As described above, the plot-based dialog model generating unit 100 generates a dialog model including dialog topics composed of the set of plots described above. And each dialogue topic is composed by connecting other plots to the starting plot and the corresponding utterances of each plot. Accordingly, according to an embodiment of the present invention, it is possible to construct as many conversational learning topics as possible with a defined set of plots.

Referring back to FIG. 2 , the destination-oriented dialogue knowledge conversion unit 200 automatically changes the learning conversational dialogue generated by the plot-based dialogue model generation unit 100 into knowledge usable in the destination-oriented dialogue system, and plot-based conversational knowledge by topic DB 300 is created.

6 is a block diagram for explaining the function of the destination-directed conversation knowledge unit 200 .

Referring to FIG. 6 , the destination-directed conversation knowledge generating unit 200 may include an utterance intention generating unit 210 , a conversation rule generating unit 220 , and an utterance intention learning unit 230 .

The utterance intention generating unit 210 receives the plot-based dialogue model and generates utterance intention information that can be recognized on a computer through the starting utterance of the plot and the representative utterance of the corresponding utterance.

In detail, the utterance intention generating unit 210 is configured to say “Hello. Why didn't you come to the meeting yesterday?" or U<1> of “I’m sorry. Yesterday, something else suddenly happened, so I couldn't go out.” Any specific intentional name can be automatically assigned.

In this case, the intention name may consist of a series of numbers or plot names or a corresponding order, and is given so as to be distinguishable from the intention names of other plots. However, the same intention should be given to the utterances that are the same as the representative utterances.

In addition, utterances that include variables must express both the intention and the variable. For example, it should be expressed in the form of intention to speak (variable = variable value). Accordingly, the intentions of U<1> and U<2> in FIG. 5 can be equally expressed as Plot22-U1 (@action=value), and “let’s go to exercise” is expressed as Plot22-U1 (@action=exercise). “Let’s go hiking” can be expressed as Plot22-U1 (@action=climbing), and “let’s go for a walk” can be expressed as Plot22-U1 (@action=walking).

Next, the dialog rule generating unit 220 converts the representative utterances of the plots included in the plot-based dialog model to the dialogue rule as connection information of the plots based on the utterance intention information generated by the utterance intention generating unit 210 . create information

The dialog rule information generated by the dialog rule generator 220 may be composed of an utterance condition and an action utterance.

The utterance condition simply means the intention to speak in the previous part, and the behavioral utterance can be expressed as the response intention shown in the next part.

For example, in case S<1> of FIG. 5 is a condition in which the utterance “what shall we do today?” appears, the action utterance of U<1> or U<2> “let’s go to @action” is the rule get angry And if the utterance condition is “Let’s go to @action” and the value of @action is “exercise”, a rule is created in which the starting utterance for a Plot: movement type is an action utterance.

In addition, the dialog rule generating unit 220 uses information such as the learner level and progress defined in the plot to be used in the dialogue rule so that the plot-based conversation topic can be communicated based on information such as the learner's level and educational progress. Conversation rule information can be created so that the defined response and dialogue flow proceed only when the information of the learner must be satisfied with the condition by being recorded together with the condition.

Next, the utterance intention learning unit 230 is configured to recognize the representative utterances and paraphrases of the plots included in the plot-based dialogue model as the utterance intention information generated by the utterance intention generating unit 210 . to learn

In this case, the speech intention learning unit 230 may learn the intention classification model in advance in order to be recognized as speech intention information. Here, as the intention classification model, various models such as various CRF, SVM, and DNN used in language processing can be applied.

For example, the utterance intention learning unit 230 displays “Let’s go to exercise” in FIG. 5 and its paraphrases “Let’s exercise” and “How about exercising” as Plot22-U1 (@action=exercise). You can learn to be recognized.

As such, the conversation rule information on the learning topic generated by the dialogue rule generating unit 220 and the speech intention learning data generated as a result of learning by the speech intention learning unit 230 are plot-based dialogue knowledge DB 300 for each topic. ) is stored in

Referring back to FIG. 2, the plot-based conversation processing unit 400 conducts a conversation with the learner to correspond to the foreign language level information and progress information registered by the learner using the information stored in the plot-based topic-specific dialog knowledge DB 300, It provides the accuracy of the content uttered by the learner during or after the conversation has ended.

That is, when a foreign language learner selects a specific conversation topic using the plot-based conversational knowledge DB 300 for each topic, the plot-based conversation processing unit 400 may practice and educate conversation within the topic.

7 is a block diagram for explaining the function of the plot-based conversation processing unit 400 .

The plot-based conversation processing unit 400 includes a voice recognition unit 410 , a speech intention recognition unit 420 , a conversation management unit 430 , and a speech evaluation unit 440 .

The voice recognition unit 410 recognizes the input utterance uttered by the learner by voice and converts it into text.

The utterance intention recognition unit 420 recognizes the utterance recognized as text by the voice recognition unit 410 as the utterance intention information defined through the destination-directed dialogue knowledge unit 200 .

The speech intention recognition unit 420 recognizes the speech intention information by using the learned information according to the learner's foreign language level information, or recognizes the speech intention information only when the converted text matches the learning sentence. can do.

At this time, recognizing the speech intention information only when it matches the learning sentence is useful in the process of learning a specific pattern repeatedly, and recognizing the speech intention information using the learned information means that the learner can use various expressions Because it is useful on a level where it can be tried.

The dialog management unit 430 generates a system response based on the conversation history information and current utterance information that the learner and the system have had conversations so far, and the dialog rule information defined by the destination-directed conversation knowledge unit. In this case, the conversation management unit 430 may receive foreign language level information and progress information of the foreign language learner, and through this, a rule to be answered in the conversation rule information may be selected and provided to correspond to the learner's level.

The speech evaluation unit 440 compares and evaluates the learning sentence for the speech intention information recognized by the speech intention recognition unit 420 with the current learner's speech. The speech evaluation unit 440 performs a content suitability assessment to evaluate whether the learner's current speech intention is an expression suitable for the flow of a conversation topic, a foreign language expression assessment to evaluate whether it is appropriate as a foreign language expression, and whether the learner's topic task has been well performed It is possible to conduct a mission performance evaluation to evaluate the

The content suitability evaluation determines whether the learner is an accurate learner in the conversation flow by figuring out whether it is consistent with the conversation topic flow composed of the plot-based conversation model, or whether it is the expected intention after the utterance intention in the current system.

The foreign language expression evaluation determines the degree of similarity between the expressions spoken by the learner and the learning sentences corresponding to the recognized intentions of speech. That is, the speech evaluation unit 440 may evaluate the foreign language expression by determining the degree of matching according to the case of the same as the learning sentence and the importance of a verb, a noun, and the like, calculating and grading the degree of similarity.

Task performance assessment determines and evaluates whether a learner is asked to perform one of the responses appropriate to the plot title of a specific plot defined in the topic dialog, whether that request has been made for each plot.

On the other hand, the plot-based conversation processing unit 400 uses the conversation rule information used by the conversation management unit 430 for utterances that are not classified through the utterance intention information defined in the subject in the utterance intention recognition unit 420 . Thus, it may further include a general conversation processing unit 450 that provides a response as a general conversation.

The general dialog processing unit 450 activates a rule composed of “auto-response” in the dialogue rule information defined by the plot in the dialogue management unit 430 to automatically respond to the learner’s utterance in a large-scale dialogue pattern. It is possible to automatically generate and provide a response using an example set composed of

In addition, the plot-based conversation processing unit 400 may further include a comprehensive evaluation processing unit 460 that stores the evaluation results in the speech evaluation unit 440 and provides statistical information and a total score for each evaluation information at the end of the conversation. there is.

Also, the plot-based conversation processing unit 400 may further include a conversation output unit 470 and a multimedia output unit 480 .

The dialogue output unit 470 synthesizes the evaluation contents of the speech evaluation unit 440 and the system response from the dialogue management unit 430 to evaluate and feedback speeches on the topic flow for the learner's speech and foreign language expressions. print out

The dialog output unit 470 utters the educational feedback utterance first and then performs the system response utterance, so that the learner can first determine whether there is an error in the self-utterance, and then proceed with the topic conversation. can do.

For educational feedback utterances, target learners who have passed the content relevance evaluation in the utterance evaluation unit 440 but have received not bad grades above a preset grade in the foreign language expression evaluation, by detecting the learning sentence most similar to the learner's utterance, grammatical, It provides learners with expressively perfect sentences so that minor errors can be corrected.

In this case, the dialogue output unit 470 may change and output the learning sentence that is most similar to the learner's utterance in the form of an utterance that confirms the other's utterance, rather than outputting the learning sentence as it is. In addition, a method of automatically learning patterns may be applied to change the learning sentence into an utterance form that confirms the utterance of the other party.

For example, when uttering U<1> in FIG. 3, the learner says “I’m sorry. When you say, "I'm not going because something else happened yesterday," the speech intention recognition unit 420 accurately recognizes the speech intention information and has the most similar intention in the learning sentence, "I'm sorry. Yesterday, suddenly something else happened, so I didn’t go.”

If this is changed to an utterance that confirms the above learning sentence as the utterance of the other party, “I couldn’t come out because something else suddenly happened yesterday” is generated, and through this, the learner can check the wrong part in the content he uttered.

The multimedia output unit 480 outputs the system output and sentence-type evaluation of the dialogue output unit 470 through voice or screen, and displays the evaluation from the viewpoint of foreign language expression for the utterance and the evaluation of content suitability in the flow of the topic conversation. It can be output as a character's expression or as a display in a specific part of the screen.

That is, the dialogue output unit 470 may provide the results of the content suitability evaluation, foreign language expression evaluation, and task performance evaluation evaluated by the speech evaluation unit 440 to the multimedia output unit 480 , and the multimedia output unit 480 . ) can enable learners to receive the results of each evaluation through a variety of methods depending on the type of service during the conversation.

For reference, the components shown in FIGS. 1 to 7 according to an embodiment of the present invention may be implemented in the form of software or hardware such as a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC). can perform roles.

However, 'components' are not limited to software or hardware, and each component may be configured to reside in an addressable storage medium or to reproduce one or more processors.

Thus, as an example, a component includes components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, procedures, sub It includes routines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables.

Components and functions provided within the components may be combined into a smaller number of components or further divided into additional components.

Hereinafter, a learning method in the foreign language conversation learning apparatus 1 according to an embodiment of the present invention will be described with reference to FIG. 8 .

8 is a flowchart of a foreign language conversation learning method according to an embodiment of the present invention.

In the foreign language conversation learning method according to an embodiment of the present invention, first, a learning conversation conversation on a subject to learn foreign language conversation is generated in the form of a plot-based conversation model ( S110 ).

Next, the generated plot-based dialog model is automatically changed to knowledge usable in the destination-oriented dialog system to generate the plot-based dialog knowledge DB 300 for each topic (S120).

Next, a conversation is conducted with the learner to correspond to the foreign language level information and progress information registered by the learner by using the information stored in the plot-based topic-specific dialog knowledge DB 300 (S130), and the accuracy of the conversation progress result to provide (S140)

Meanwhile, in the above description, steps S110 to S140 may be further divided into additional steps or combined into fewer steps according to an embodiment of the present invention. In addition, some steps may be omitted if necessary, and the order between steps may be changed. In addition, even if other omitted content, the content already described in FIGS. 1 to 7 may be applied to the foreign language conversation learning method of FIG. 8 .

Meanwhile, an embodiment of the present invention may be implemented in the form of a computer program stored in a medium executed by a computer or a recording medium including instructions executable by the computer. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer-readable media may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transport mechanism, and includes any information delivery media.

Although the methods and systems of the present invention have been described with reference to specific embodiments, some or all of their components or operations may be implemented using a computer system having a general purpose hardware architecture.

The description of the present invention described above is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

1: Foreign language conversation learning device
100: plot-based dialog model generation unit
200: Conversation towards destination knowledge building
300: Plot-based conversational knowledge DB by topic
400: plot-based conversation processing unit

Claims (9)

A plot-based dialogue model generation unit that generates a learning conversational conversation on a topic to learn foreign language conversation in the form of a plot-based dialogue model;
a destination-oriented dialogue knowledge unit that automatically changes the learning conversational dialogue to knowledge usable in the destination-oriented dialogue system to create a plot-based topic-specific dialogue knowledge DB; and
It includes a plot-based conversation processing unit that conducts a conversation with the learner to correspond to the foreign language level information and progress information registered by the learner using the information stored in the plot-based conversational knowledge DB for each topic, and provides accuracy for the conversation progress result and,
The plot-based dialog model generation unit generates a role conversation with a learner in a plot unit, generates the plot-based dialog model through a connection between the plots, and generates the plot unit to consist of a starting utterance and a corresponding utterance,
The plot-based conversation processing unit induces a conversation flow by performing exception handling when the learner does not respond to the corresponding response defined in the corresponding plot.
A foreign language conversation learning device.
delete According to claim 1,
The starting utterance includes a plot title as plot information, and the corresponding utterance includes a plot correspondence method for the plot title and the foreign language level information and progress information of a learner capable of the corresponding utterance as plot information
A foreign language conversation learning device.
According to claim 1,
The destination-oriented conversational knowledge unit
a utterance intention generating unit that receives the plot-based dialogue model as an input and generates utterance intention information that can be recognized on a computer through a representative utterance of a plot starting utterance and a corresponding utterance;
a dialogue rule generation unit generating dialogue rule information using the connection information of the plot based on the representative speech and the speech intention information; and
including a speech intention learning unit that learns the speech intention so that the representative speech and paraphrase of the plot are recognized according to the speech intention information
A foreign language conversation learning device.
5. The method of claim 4,
The generated dialogue rule information and the speech intention learning data generated as a result of the learning are stored in the plot-based topic-specific dialogue knowledge DB
A foreign language conversation learning device.
5. The method of claim 4,
The plot-based conversation processing unit
a speech recognition unit for recognizing the learner's speech and converting it into text;
a speech intention recognition unit for recognizing the converted text as the speech intention information;
a conversation management unit for generating a system response based on the conversation history information and current utterance information with the learner, and the conversation rule information; and
Comprising a speech evaluation unit that compares and evaluates the learning sentence for the speech intention information recognized by the speech intention recognition unit and the speech of the learner
A foreign language conversation learning device.
7. The method of claim 6,
The speech intention recognition unit recognizes the speech intention information by using the information learned according to the foreign language level information of the learner, or recognizes the speech intention information only when the converted text matches the learning sentence
A foreign language conversation learning device.
7. The method of claim 6,
The dialog management unit receives the foreign language level information and progress information of the learner, and selects and provides a rule to respond to in the dialog rule information to correspond to the learner's level
A foreign language conversation learning device.
7. The method of claim 6,
The plot-based conversation processing unit further comprises a general conversation processing unit that provides a response as a general conversation by using the dialogue rule information for utterances that are not classified through the utterance intention information in the utterance intention recognition unit
A foreign language conversation learning device.

Images