KR101004913B1 - An apparatus and method for evaluating spoken ability by speech recognition through computer-lead interaction and thereof - Google Patents

Abstract

The object of the present invention is to first speak the language in a general conversational manner using a computer, and in particular, to limit the number of conversations that are predicted, thereby using the limited resources of the computer to induce a natural conversational speech evaluation. Achievement can be assessed, and the second examinee's speaking input can be compared with the evaluation criteria (references) recorded in the database of the data section. Speaking attempts are stored in the database as a comprehensive scoring standard, taking into account the feelings, naturalities, and impressions of the examinee's speaking attempts. Clarity, fluency, etc.) To provide to the evaluation criteria utilized by the computer-driven Speaking speech recognition that can be quantified as part of the total score by summing the scores entered into the database language learning ability evaluation system and its evaluation method based on the ignition data.

Speech assessment system, assessment criteria building, speech, computer, number of conversations, speech recognition, helpers,

Description

Apparatus and method for evaluating speech in computer-driven interaction using speech recognition and evaluation method {AN APPARATUS AND METHOD FOR EVALUATING SPOKEN ABILITY BY SPEECH RECOGNITION THROUGH COMPUTER-LEAD INTERACTION AND THEREOF}

The present invention relates to a computer-driven speech proficiency evaluation system and method using speech recognition, and in particular, by constructing the speech abilities of a specific language interactively, the ability to speak through a process of exchanging natural conversations with a computer as in real life. By using the computer, you can speak (try to speak) the language in a general conversational manner and limit the number of conversations that are predicted. The present invention relates to a computer-driven speech proficiency evaluation system and method using speech recognition capable of evaluating speech abilities objectively and with high reliability.

Conventional speaking evaluation, which has been widely used in the past, has been mainly conducted by humans and has been evaluated by humans. However, these evaluations have a problem that it takes a lot of time. There was no problem.

In order to improve such a problem, a method of recording an examinee's speaking attempt through a computer and delivering the score to a grader has been widely utilized.

However, this evaluation also has a problem that the score of the examinee is still determined by the scorer's supervision and the scorer selection and the scoring period takes a long time, and the examinee must only answer the question for the speech evaluation manually. There was a problem of closedness because the examinee expressed his opinion and lacked negotiation with the other party for dialogue.

In addition, there is a problem that can not solve the unreasonable point that the score of the examinee is determined in the selection of the scorer because it is composed of a grading system as a grading standard.

In order to solve this problem, a recent computer-aided speech assessment is performed by the examinee by simply reading or repeating the contents presented by the computer or repeating the sentences, and storing the speaking attempts by the short answers. Speaking evaluation, which scores are determined according to fluency, is also introduced, but this speaking evaluation also evaluates speaking ability by indirect method by dividing examinee's pronunciation or speaking fluency by grading system, not assessment of living speaking ability through natural conversation. There was a problem.

In order to solve these problems, the applicants of the present invention were registered in 2000 and 4 cases (No. 0287227,0287228,0296271,0296272), and the patent application (patent application No. 10-2007-0063336) computer-driven dialog device In the method and method, it is designed to lead a 1: 1 or multiple language conversation by a computer and to limit the number of conversations predicted in the conversation so that the interactive conversation is performed.

Conventional speaking evaluation is based on the overall grading criteria for the examinee's feelings of speaking attempts and partial grading criteria, such as pronunciation, fluency, grammatical errors, and vocabulary selection. It is true that objective scoring is difficult, because it takes a long time and the score varies depending on the physical condition of the grader and mood. In the case of scoring by humans, it was very much dependent on the overall scoring standard that gives a score as an overall impression, and it was true that it was too difficult to score based on the partial scoring criteria. When humans scored, there was a problem that it was almost impossible to grade the scores by listening to the examinees and comparing them with the partial grading criteria.

Therefore, in the case of human grading, the grading depends on feelings, impressions, moods, and feelings of the examinee, so that there is a problem that objectivity cannot be maintained. There was a problem that can not maintain objectivity.

For example, the examinee's pronunciation and vocabulary selection are classified mechanically, and short answers close to short answers are classified into correct and incorrect answers, and grammatically perfect sentences are recognized as correct answers, and sentences containing some errors are recognized as incorrect answers. Even if the communication can be made smooth enough, there was a problem that is determined by a low score.

Accordingly, the present invention has been made to solve the above-described problems, and an object of the present invention is to limit the number of conversations that can be spoken (attempted to speak) by using a computer in a general conversational manner, in particular by limiting the number of conversations expected. The present invention provides a computer-driven speech language learning ability evaluation apparatus using speech recognition and a method of evaluating the speech using the resources to induce a speech evaluation of a natural conversation method.

Another object of the present invention is to use the speech recognition to obtain the evaluation results on the fly with an objective and high reliability by comparing the examinee's speaking input with the evaluation criteria previously recorded in the database of the data section. To provide a computer-led speech language learning ability evaluation apparatus and its evaluation method.

delete

It is another object of the present invention to consider the feelings, naturalness, and impression of the examinee's speaking attempt as an overall scoring criterion stored in the database of the examinee's speaking attempts. Appropriateness of choice, relevance of content, clarity of expression, fluency, etc.) is computer-driven using speech recognition that can be input into the database and summed the partial scores to quantify the total scores. The present invention provides an apparatus for evaluating speaking language learning ability and an evaluation method thereof.

The present invention has been made to achieve the above object, the computer-driven speaking ability evaluation apparatus using the speech recognition of the present invention includes a program storage unit for storing a program provided to the examinee who wants to receive a language assessment led by a computer; A control unit for inducing an answer by a helper stored in a helper unit of the processor if the answer is an unpredicted answer, and controlling the answer to be compared with the searched answer; A processing unit for recognizing a voice signal, converting an answer voice signal, a background sound signal, and a background image signal from corresponding data, determining whether the answer is a predictable answer, and outputting a designated helper signal; A data unit for recording the recognized voice signal, the voice signal to be output, the background sound signal, the background video signal, the predicted answer signal, and the helper information in a database; A microphone for inputting the examinee's voice response signal or voice question signal; A display unit for outputting a background image and text for evaluation in color; An input unit for inputting an evaluation request command and data information for an evaluation request; A speaker for outputting a speech signal and a background sound signal for evaluating speech; Characterized in that it comprises a communication unit for communicating with other computers by wire or wireless.

In addition, the computer-driven speaking ability evaluation method using the speech recognition of the present invention includes a background image, text / background sound output process of outputting the background image, text and background sound for language learning by the computer; An evaluating dialogue voice signal output process for outputting a dialogue speech signal for speech evaluation after the background image / text / background acoustic output process; After the voice signal is recognized in the evaluation dialogue voice signal output process, an answering process for answering the evaluation dialogue voice signal by a microphone, and a voice for determining whether the examinee has recognized the voice signal answered by the examinee. Recognition discrimination process; An answer discrimination process for determining whether the recognized speech signal answered by the examinee in the speech recognition determination process is a predicted answer when the speech signal recognized by the examinee in the speech recognition determination process is recognized; An answer data search process of searching for answer data stored in the predictive answer database of the data part under the control of the control part of the examinee's answer in the case of a predicted answer as a result of the determination of the answer discrimination process; An answer similarity determination step of determining whether the answer data searched in the answer data search process and the examinee's answer are similar; An appropriate answer evaluation step of evaluating an appropriate answer if the examinee's answer is similar to the answer data retrieved from the predicted answer database of the data unit in the process of determining whether the answer is similar or not; An evaluation result storing step of storing the evaluation result evaluated in the appropriate answer evaluation step in the evaluation result database part of the data part; A result of determining whether the answer is similar or not, when the answer data searched in the answer data search process and the answer of the examinee are not similar, an incorrect answer evaluation process of evaluating a wrong answer; And an evaluation result storing step of storing the evaluation result in the evaluation result database of the data unit.

According to an apparatus and method for evaluating computer-driven speaking ability using speech recognition according to the present invention, first, a language is spoken (attempted to speak) by a general conversation method using a computer, and in particular, the number of conversations predicted is limited. Using limited resources, natural speaking methods can be used to assess speech achievement, and second, the examinee's speech input is compared with the evaluation criteria pre-recorded in the database of the data section. In addition, the third examinee's speaking attempt (speech) is a comprehensive scoring criterion that is stored in the database of the data section, taking into account the feelings, naturalness, and impressions of the speaking examinee's attempts. Accuracy or appropriateness , Grammatical accuracy, appropriateness of vocabulary selection, relevance of content, clarity of expression, fluency, logicality, etc.) are input into the database based on the native speaker's speech utterance data, and the partial scores are summed to quantify immediately. There are many outstanding effects such as being able to.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT Hereinafter, a computer-driven speaking ability evaluation apparatus using speech recognition according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

In describing an embodiment of the present invention, the gist of the present invention will be clearly described by omitting drawings and descriptions for technical contents that are not directly related to the present invention and are well known.

1 is a block diagram schematically showing an apparatus for evaluating computer-driven speaking ability using speech recognition according to an embodiment of the present invention, FIG. 2 is a block diagram showing a detailed structure of a processing unit and a data unit in FIG. FIG. 3 is a diagram schematically illustrating a tree structure based on questions and responses of a conversation in the apparatus for evaluating computer-driven speaking ability using voice recognition according to an embodiment of the present invention.

1 and 2, a computer-driven speaking ability evaluation apparatus using voice recognition according to an embodiment of the present invention is a program storage unit for storing a program provided to examinees who want to receive an interactive language evaluation led by a computer. 10; A control unit 20 for inducing an answer by a helper stored in the helper unit 36 of the processing unit 30 and comparing the searched answer if the examinee's answer is an unexpected response; A processing unit 30 for recognizing a voice signal, converting an answer voice signal, a background sound signal, and a background image signal from corresponding data, determining whether the answer is a predictable answer, and outputting a designated helper signal; A data unit 40 for recording the recognized voice signal, the voice signal to be output, the background sound signal, the background image signal, the predicted answer signal, and the helper information in a database; A microphone 50 for inputting the examinee's voice response signal or voice question signal; A display unit 60 for outputting a background image and text for evaluation in color; An input unit 70 for inputting an evaluation request command and data information for an evaluation request; A speaker 80 for outputting a speech signal and a background sound signal for evaluating speech; The communication unit 90 communicates with other computers by wire or wirelessly.

In addition, as shown in FIG. 2, the processor 30 includes a voice recognition unit 31 for recognizing a voice signal and converting the voice signal into data; A voice output unit 32 which converts the retrieved and provided voice data into a voice signal and outputs the voice signal; A background sound unit 33 for converting the searched and provided background sound data into an audio signal; A background image unit 34 for converting the searched and provided background image data or text data into an image signal; A predictable dialogue unit 35 for confirming whether or not the answer signal provided is a predictable answer signal retrieved from the data unit 40; The helper unit 36 converts the designated and retrieved data into corresponding helper information.

In addition, the data unit 40 includes a voice recognition database 41 for storing voice signal data as data; A voice output database 42 for storing voice output data and outputting the retrieved voice data to the controller 20 for evaluation of speech; A background sound database 43 for storing background sound data for speech evaluation, retrieving the stored background sound data, and outputting the stored background sound data to the controller 20; A background image database 44 for storing background image data for speech evaluation, retrieving the stored background image data, and outputting the stored background image data to the controller 20; Predictive answer storing the predicted answer signal of the examinee for language evaluation and receiving the answer signal recognized by the voice recognition unit 31 of the processing unit 30 from the control unit 20 to search whether there is a similar answer signal A database 45; A helper database 46 for storing characters, characters, sounds, and colors as helper data, and searching and providing the designated helper data to the controller 20; While storing the evaluation criteria (evaluation criteria) for evaluating the examinee's speech utterance ability, receiving the voice output data stored in the voice output database 42 and comparing the answer signal to output the comparison result to the control unit 20 An evaluation criteria database 47; It consists of a dialogue subject / topic database 48 which stores the dialogue subject data or the subject data, and an evaluation result database 49 which stores the evaluation result data of the examinee.

Next, the operation and effects of the computer-driven speaking ability evaluation apparatus utilizing the speech recognition of the present invention configured as described above will be described.

First, the computer-driven speaking ability evaluation apparatus using voice recognition according to an embodiment of the present invention, if the examinee inputs an evaluation request signal for requesting to start language evaluation from the input unit 70 is input to the control unit 20, the control unit In operation 20, the corresponding application program for language evaluation is loaded from the program storage unit 10, and the standby state is provided for providing the language evaluation service to the examinee.

At this time, the control unit 20 monitors the communication unit 90 to determine whether it is connected to another learning computer or management computer and the communication unit 90, and as a result of the determination, the control unit 20 is connected to another evaluation and learning computer or management computer. In this case, the current operation status information and management information required for management are exchanged with each other, and the necessary information is periodically controlled to exchange information with each other as the speech evaluation progresses.

The controller 20 provides a background image of the speech evaluation service from the voice output database 42, the background acoustic database 43, and the background image database 44 of the data unit 40 to provide the speech evaluation service to the examinee. The voice signal including the signal, the background sound signal, and the text TEXT data is searched and provided to the processor 30.

The processing unit 30 transmits the background image signal, the background sound signal, and the text data provided from the control unit 20 to the background image unit 34, the background sound unit 33, and the audio output unit 32 of the processing unit 30, respectively. The background image unit 34 of the processing unit 30 converts the background image data into a background image signal (including a text signal), and the background sound data of the background unit 33 of the processing unit 30 performs the background. The audio signal is converted into an audio signal, and the text data, which is the audio data of the processor 30, is converted into an audio signal by TTS (TEXT TO SPEECH) technology.

The video signal and the text signal respectively converted by the processing unit 30 are output from the control unit 20 to the display unit 60 to display an image screen on the display unit 60, and at the same time, The background sound signal converted by the background sound unit 33 and the background sound signal converted by the voice output unit 32 of the processor 30 are provided to the speaker 80 to output sound and voice.

The control unit 20 monitors whether the examinee's voice signal is input from the microphone 50 and checks whether the examinee's voice signal is provided to the voice recognition unit 31 of the processing unit 30 and recognized.

The speech recognition unit 31 of the processing unit 30 provides the recognized answer speech signal to the predictable conversation unit 35 to check whether it is included in the predicted answer for speech evaluation. The predictable conversation unit 35 of the processing unit 30 searches the prediction answer database 45 of the data unit 40 and analyzes whether there is an answer similar to the recognized answer. That is, an answer that is similar to the recognized answer by a predetermined ratio or more is searched from the prediction answer database 45, and prepared by digitally processing the predicted answer and the recognized answer in the prediction answer database 45 of the data part 40. It may be determined whether the similarity is equal to or greater than a predetermined ratio preset in the prediction answer database 45 of the data unit 40.

In the above description, a predetermined ratio set for measuring similarity is a state in which the response voice signal input and recognized through the microphone 50 includes the noise of the surrounding environment, a change in the examinee's intonation, the use of similar words and phrases, and the examinee. The appropriate value is selected and evaluated in a repeated experiment in consideration of differences in voice signals due to age and gender.

If it is confirmed that the answer input by the examinee is included in the range of the predicted answer, the controller 20 searches for the specified answer data from the voice output database 42 of the data part 40 and processes the data to obtain similarity. Check it.

If the recognized answer and the designated answer found are more than a certain percentage set, provide the examinee with an expression such as "It is an appropriate answer" and return to the next step to proceed with language learning. The same expression is provided to the examinee through the speaker 80, and the process returns to the next step. The speaking evaluation result is stored in the evaluation result database 49 of the data unit 40 and used for future learning progress. Provide the examinee if requested by the examinee.

In addition, the controller 20 controls the helper 36 when the answer of the examinee inputted to the monitored voice recognition unit 31 is not the answer predicted by the predictable conversation unit 35, and controls the designated helper 36. Is outputted and a limited range of answers from the examinee is entered and recognized.

Designated helpers include characters, pictures, videos, text messages, voice messages, color messages, and the like. For example, "Are you wondering ~~? Ask me." Or "How about asking ~~?" The computer leads the conversation for evaluating speech and limits the scope of the conversation. The helper information is recorded and managed in the helper database 46 of the data unit 40, and records and manages the helper in advance for each evaluation step.

Conventionally, by simply recognizing the examinee's voice signal and comparing whether or not a designated voice signal has been input, a method of judging right and wrong, that is, judging only the success of voice recognition, leads the conversation. It was a way of answering a question or a conversation by reading or memorizing, but in the present invention, the guide is used to guide the conversation by using a helper to induce or limit the direction of the conversation so that a certain direction of the conversation is achieved. Evaluate speaking skills with the examinee freely or interactively.

In general conversations, there are an infinite number of conversations or questions in very complex forms, such as what conversations and questions with examinees will include, for example, positive, negative, questionable questions, and additional questions. The capacity of the database is infinitely needed to store the contents for the respondents. In addition, there is a problem that the processing method is very complicated and time-consuming, but in the present invention leads to the desired direction by limiting the conversation, that is, the language learning using a helper by a predetermined limited sentence, words, etc. The computer-driven direction of the conversation can help you assess your speaking skills in a natural way with a small computer database.

There are various interactive interactive deployment methods using the helper, but one embodiment of the present invention will be described as follows.

end. Drawing, video assistant

Since a certain picture or video is guided in a certain direction to generate a question or an answer, an unlimited number of questions and answers may be limited to a desired direction.

These conversations form a sequence of conversations for each stage of language learning, and the examinee responds to the induction questions of the computer by allowing the examinee to recognize the main character of the given situation as a picture or video for 1: 1 conversation with the computer. Evaluate your abilities.

I. Predictive Data Entry

Predictive answers of the examinee are recorded in advance, and recorded as data, and inputted as data to confirm the set development direction to control the progress of speaking evaluation and limit the number of conversations. In addition, the grammatical structure, communication sentences, and sentences with errors can be input as data and compared with the voice data recognized by the examinee's input to identify the type of error and assess speaking ability.

All. Utilization of helper by character and third signal

A situation where an examinee asks a computer by using a third character based on pictures and videos as an assistant, for example, "Aren't you curious about something? Ask him, What ....) ", Etc. and input the examinee's response as data for analysis. If the examinee's response is not similar to the predicted content, the computer asks the question again to guide the established conversation and evaluates the speaking ability.

In the case of not utilizing such a character, the question form by When, Where, What, Why, Who, How, and Yes-No is defined in advance and developed into a third signal such as a promised color, shape, picture, sound, or character. Evaluate speaking skills by introducing usage to the examinee, asking questions or answers, or encouraging subtitles (e.g., it's your turn.) To ask questions or answer questions through natural conversation.

In the intermediate situation where the language learning is smoothly performed, the examinee can recognize the order of questions or answers only with a predetermined signal instead of the helper, so that speaking can be easily and quickly evaluated.

la. Limit the development of dialogues to Tree-Structure

Since the dialogue with the computer develops the situation like the branch structure of the tree, the speech can be evaluated by evaluating the ability of speaking with various numbers of conversations and by restricting the situation from developing indefinitely by inducing the dialogue to proceed in a certain direction.

In the accompanying FIG. 3, a dialogue situation of such a tree structure is illustrated.

In other words, the examinee predicts the answer to the question of the conversation in advance, enters it into the computer, restricts the number of conversations so that they do not extend to infinity, and repeats the process of redeploying the limited number of conversations. Numbers can also be used to assess speech by selecting questions in a variety of situations.

4 is a flowchart schematically illustrating a method for evaluating computer-driven speaking ability using speech recognition according to an embodiment of the present invention. In the figure, S denotes a step.

As shown in FIG. 4, in the method of evaluating computer-driven speaking ability using speech recognition according to an embodiment of the present invention, the display unit 60 displays the background image, the background character, and the background sound for speech evaluation by the computer in step S1. Output (background image, background character / background sound output process), and proceed to step S2 and output the dialogue voice signal for speech evaluation after the background image, background character / background sound output process (dialogue speech signal output process for evaluation) After the voice signal is recognized in the evaluation dialogue voice signal output process, the process proceeds to step S3 and the examinee answers the answer to the evaluation dialogue voice signal in the microphone 50 (answering process).

Next, the process proceeds to step S4 to determine whether or not the voice signal answered by the examinee in the answering process is recognized (voice recognition discrimination process), and in the case of recognizing the voice signal answered by the examinee in the speech recognition process (YES). Go to step S5 to determine whether the recognized speech signal answered by the examinee in the speech recognition process is a predicted answer (answer discrimination process), and if the answer is a predicted answer (YES day) Case) to search for the answer data stored in the predicted answer database 45 of the data part 40 under the control of the control unit 20 (step S7), and proceed to step S7. Determine whether the answer data retrieved from the answer data search process and the examinee's answer is similar (determining whether or not there is a response).

Next, when the examinee's answer is similar to the answer data retrieved from the predictive answer database 45 of the data part 40 (YES) in the process of determining whether or not the answer is similar, the process proceeds to step S8 and is evaluated as an appropriate answer. (Evaluation process), the process proceeds to step S9 and the evaluation result evaluated in the evaluation process is stored in the evaluation result database part 49 of the data part 40 (evaluation result storage process), and whether the answer of step S7 is similar or not. As a result of the discrimination process, when the answer data searched in the answer data search process and the examinee's answer are not similar (NO), the process proceeds to step S10 to evaluate the wrong answer (wrong answer evaluation process), and the evaluation of the step S9. Proceeding to the result storing process, the evaluation result is stored in the evaluation result database 49 of the data part 40.

On the other hand, if the result of the voice recognition determination process in step S4 does not recognize the voice signal answered by the examinee (NO), it returns to the background image, background text / background sound output process of step S1 and the background image. • Background text / background sound output process The following process is repeated.

If the recognized speech signal answered by the examinee as a result of the answer discrimination process of step S5 is not a predicted answer (NO), the process proceeds to step S20 and is stored in the helper database 46 of the data unit 40. Induce a question limited to the designated helper (helper help process), and after the helper help process, the examinee proceeds to step S21 and the examinee asks the microphone 50 according to the helper help (question process), and proceeds to step S22 to ask the question. In step S23, it is determined whether the examinee has recognized the voice signal asked by the examinee (question speech recognition discrimination process), and if the examinee recognizes the voice signal asked by the examinee in the question speech recognition discrimination process (YES), the process proceeds to step S23. In the question voice recognition determination process, whether the recognized question voice signal asked by the examinee is a predicted question (question discrimination process) is determined. If the result of the judgment of the sentence discrimination process is a predicted question (YES), the process proceeds to step S24 and the question data stored in the helper database 46 of the data unit 40 is controlled by the examinee's question under the control of the control unit 20. Search (question data search process), and proceeds to step S25 to determine whether or not the question material and the examinee's questions searched in the question data search process is similar (determination of similar questions or not).

Next, when the examinee's question is similar to the questionnaire retrieved from the helper database 46 of the data part 40 (YES), the process proceeds to step S26 and is evaluated as an appropriate question in the process of determining whether or not the similar question is determined ( Similar question evaluation process), and proceeds to the similar evaluation result storage process and stores the evaluation result evaluated in the question evaluation process in the evaluation result database part 49 of the data part 40 (evaluation result storage process), and the step If the result of the determination process of the similar question determination process of S25 is not the question material and the examinee's question found in the question material search process (NO), go to step S27 to evaluate the wrong question (incorrect question evaluation process) The process of storing the evaluation result of step S9 is performed to store the evaluation result in the evaluation result database 49 of the data part 40.

In the present invention, the evaluation criteria for evaluating the speaking utterance ability of the examinee is composed of the evaluation criteria database 47 is composed of a comprehensive-analytical scoring grade, which is constructed by integrating the overall scoring and the analytical scoring criteria, and constructing a database. Scores can be broken down or simplified depending on the purpose of the test. Evaluation criteria of each area are as follows.

1) task achievement ability: evaluates the ability to respond to and realize a given conversational task and to determine whether the task is finally realized.

2) Communication skills: to judge the ability to convey meaning, to understand and to continue conversations.

3) Logical Logic: To determine the ability to respond logically and appropriately to the evaluator and other speakers without departing from the nature of the dialogue.

4) Ability to use language materials: Use appropriate vocabulary and grammar to convey meaning, and use paraphrase strategy to determine the ability to continue the conversation if the other party of the conversation does not understand.

5) Pronunciation: The examinee's words are clear, and it is determined whether or not the meaning can be properly delivered to the conversation partner.

6) Fluency: Determines whether the examinee's speech, rhythm and hesitation are natural.

The six items listed above are further refined to the following seven items.

① Whether the given task is realized or not

② Task realization time: Compared with the average native language realization time

③ Use of core vocabulary: Adjust the expression of the expression to grade by using the core vocabulary or not using the core vocabulary.

④ Accuracy of sentences (with or without grammar errors): Grammar accuracy is calculated by dividing the error into grades according to the importance of grammar (whether it affects communication).

⑤ Pronunciation accuracy: Different grades depending on whether native speaker pronunciation or communication affects communication compared to native speaker's pronunciation

⑥ Fluency: Differential grades are compared with the data of native speakers based on the speed, rhythm and hesitation of the examinee's words.

⑦ Logic: As a criterion to judge whether the conversation has been conducted consistently and smoothly, a helper appeared to judge whether the examinee continued the conversation or whether the conversation was performed well by himself. The score is deducted based on the number of appearances of the assistant.

The detailed evaluation items presented above are scored in order of native speakers in order and input into the evaluation criteria database 47. For example, if the task is not realized, the score is calculated as 0, and if the task is realized, the score is returned to the sub-category item, and according to ②③④⑤, the score is 0 to the specific score (1 to 10 or 1 to the specific score). Differentiate to (the steps may vary depending on the nature and purpose of the test). Through the evaluation calculation criteria, the speaking content input by the examinee is verified in the evaluation criteria database 47 and calculated as a score.

Therefore, the present invention uses a computer to speak (attempt to speak) a language in a general conversational manner, and in particular, limits the number of conversations that are predicted, thereby evaluating speaking ability by inducing natural conversational speech using limited resources of the computer. By comparing the examinee's speaking input with the evaluation criteria recorded in advance in the database of the data section, the examinee's speaking ability of the specific language can be assessed on the spot and objectively and with high reliability. As the overall scoring criteria stored in the database of the data department, it considers the feelings, naturalness, and impressions of the examinee's attempts to speak. Clarity, fluency, etc.) To as evaluation criteria based on the ignition and enter data into the database summing the partial credit it can be quantified as possible.

In the above description, the specific embodiments have been shown and described, but the present invention is not limited thereto, for example, by those skilled in the art without departing from the concept of the present invention. Of course, the design can be changed in many ways.

1 is a block diagram schematically illustrating an apparatus for evaluating computer-driven speaking language learning ability using speech recognition according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a detailed structure of the processing unit and the data unit in FIG. 1.

FIG. 3 is a diagram schematically illustrating a tree structure based on questions and answers of a learning conversation in a computer-driven speech evaluation apparatus using voice recognition according to an embodiment of the present invention.

4 is a flowchart schematically illustrating a method for evaluating computer-driven speaking ability using speech recognition according to an embodiment of the present invention.

          <Explanation of symbols for main parts of the drawings>

10: program unit 20: control unit

30: processing unit 31: voice recognition unit

32: sound output unit 33: background sound unit

34: background image part 35: predictable conversation part

36: helper 40: data

41: voice recognition database 42: voice output database

43: background sound database 44: background image database

45: predictive answer database 46: helper database

47: evaluation criteria database 50: microphone

48: Conversation subject matter topic 60: Display part

70: input unit 80: speaker

90: communication unit

Claims (7)

A program storage unit 10 for storing a program provided to the examinee who wants to receive an interactive language evaluation by the computer; A control unit 20 for inducing an answer by a helper stored in the helper unit 36 of the processing unit 30 and comparing the searched answer if the examinee's answer is an unexpected response; A processing unit 30 for recognizing a voice signal, converting an answer voice signal, a background sound signal, and a background image signal from corresponding data, determining whether the answer is a predictable answer, and outputting a designated helper signal; A data unit 40 for recording the recognized voice signal, the voice signal to be output, the background sound signal, the background image signal, the predicted answer signal, and the helper information in a database; A microphone 50 for inputting the examinee's voice response signal or voice question signal; A display unit 60 for outputting a background image and text for learning in color; An input unit 70 for inputting an evaluation request command and data information for an evaluation request for speaking; A speaker 80 for outputting a speech signal and a background sound signal for evaluating speech; In the computer-driven speaking language learning ability evaluation device using the speech recognition consisting of a communication unit 90 for communicating with other computers by wire or wireless, The data portion 40 is A voice recognition database 41 which stores voice signal data as data; A voice output database 42 for storing voice output data and outputting the retrieved voice data to the controller 20 for speech evaluation; A background sound database (43) for storing background sound data for speech, retrieving the stored background sound data, and outputting the stored background sound data to the controller (20); A background image database 44 for storing background image data for speaking, retrieving the stored background image data, and outputting the stored background image data to the controller 20; Predictive answer storing the predicted answer signal of the examinee for speech evaluation and receiving the answer signal recognized by the voice recognition unit 31 of the processing unit 30 from the control unit 20 to search whether there is a similar answer signal A database 45; A helper database 46 for storing characters, characters, sounds, and colors as helper data, and searching and providing the designated helper data to the controller 20; While storing the evaluation criteria (evaluation criteria) for evaluating the examinee's speech utterance ability, receiving the voice output data stored in the voice output database 42 and comparing the answer signal to output the comparison result to the control unit 20 An evaluation criteria database 47; A dialogue subject / subject database 48 which stores dialogue subject data or subject data; A computer-driven speaking ability evaluation apparatus using speech recognition, characterized in that the evaluation result database for storing the test result data of the examinee. delete delete A background image, text, and background sound output process of outputting a language learning background image, text, and background sound to the display unit 60 by a computer; An evaluating dialogue voice signal output process for outputting a dialogue speech signal for speech evaluation after the background image / text / background acoustic output process; After the voice signal is recognized in the evaluation dialogue voice signal output process, an answering process for answering the evaluation dialogue voice signal by the microphone 50 and whether or not the voice signal answered by the examinee in the answering process are recognized. A voice recognition determination process of discriminating; An answer discrimination process for determining whether the recognized speech signal answered by the examinee in the speech recognition determination process is a predicted answer when the speech signal recognized by the examinee in the speech recognition determination process is recognized; An answer data search process for searching the answer data stored in the predictive answer database 45 of the data part 40 under the control of the controller 20 when the answer is a predicted result of the determination of the answer discrimination process; An answer similarity determination step of determining whether the answer data searched in the answer data search process and the examinee's answer are similar; An appropriate answer evaluation process of evaluating an appropriate answer when the examinee's answer is similar to the answer data retrieved from the predicted answer database 45 of the data part 40 in the process of determining whether the answer is similar or not; An evaluation result storing step of storing the evaluation result evaluated in the appropriate answer evaluation step in the evaluation result database part 49 of the data part 40; A result of determining whether the answer is similar or not, when the answer data searched in the answer data search process and the answer of the examinee are not similar, an incorrect answer evaluation process of evaluating a wrong answer; In the computer-driven speaking ability evaluation method using speech recognition comprising the evaluation result storage step of proceeding to the evaluation result storage process and storing the evaluation result in the evaluation result database 49 of the data unit 40, A helper help process for deriving a question limited to a designated helper stored in the helper database 46 of the data unit 40 when the recognized voice signal answered by the examinee is not a predicted answer; After the helper help process, a questionnaire process for the examinee to question the microphone 50 in accordance with the helper help; A question voice recognition determining step of determining whether the examinee has recognized a voice signal asked by the examinee; A question discrimination process of determining whether the recognized question sound signal queried by the examinee in the question voice recognition determination process is a predicted question when the voice signal asked by the examinee is recognized in the question speech recognition determination process; A question material retrieval process of retrieving question data stored in the helper database 46 of the data unit 40 under the control of the control unit 20 when the question is a predicted result of the determination of the question determination process; A similar question determining step of determining whether the question material searched in the question material searching process and the examinee's question are similar; An appropriate question evaluating step of evaluating an examinee's question as an appropriate question when the examinee's question is similar to the questionnaire retrieved from the helper database 46 of the data part 40; As a result of determining whether the similar question is determined or not, if the question material searched in the question material search process and the examinee's question are not similar to each other, the method further includes an incorrect question evaluating process which evaluates as a wrong question. Computer-Driven Speaking Assessment. delete delete The helper of claim 4, wherein when the examinee does not recognize the voice signal asked by the examinee in the question recognition process or when the recognized question voice signal asked by the examinee in the question discrimination process is not a predicted question. A method of evaluating computer-driven speech using speech recognition, characterized in that to return to the help process to perform the following actions of the helper help process.

Images