Question submission system and method

Abstract

This system and method for receiving questions provides answers to questions in a format suitable for learners. [Solution] The system includes a storage unit for storing educational material data, an artificial intelligence unit that receives questions about the educational material data from a learner's information processing terminal using the educational material data and generates answers to the questions based on the educational material data, and an answer unit that outputs the answers generated by the artificial intelligence unit to the information processing terminal, reflecting the learner's characteristics.

Description

Technical Field

[0001] The present disclosure relates to a technique for assisting learning using a computer system.

Background Art

[0002] Patent Document 1 discloses a learning support system having a learning support device that supports a user's learning by generating an explanation about the answer to a problem. When the learning support device receives a question input by the user regarding the solution to the problem from the user terminal, it has a transmission / reception unit that transmits the received question to a generation AI, and a generation unit that generates an explanation based on the answer when receiving the answer from the generation AI. The explanation is composed of an answer to the question and a plurality of key points. Patent Document 1 describes that the plurality of key points are knowledge necessary for understanding the problem and the answer.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the learning support system disclosed in Patent Document 1, in addition to the answer to the question, the user can acquire the knowledge necessary for understanding the problem and the answer. However, depending on the user, there may be an excess or deficiency in the knowledge necessary for understanding the problem and the answer. For example, a user with a high learning level may feel that obtaining the answer is sufficient and that the knowledge necessary for understanding the problem and the answer is unnecessary. On the other hand, for a user with a low learning level, the knowledge included in the explanation may be insufficient for understanding. The uniformly created explanation is not necessarily suitable for every user.

[0005] One of the purposes included in this disclosure is to provide a question submission system and a question submission method that provides answers to questions in a format suitable for learners. [Means for solving the problem]

[0006] A question-receiving system in one aspect included in this disclosure comprises: a storage unit for storing educational material data; an artificial intelligence unit that, upon receiving a question regarding the educational material data from an information processing terminal of a learner who is learning using the educational material data, generates an answer to the question based on the educational material data; and an answer unit that transmits the answer generated by the artificial intelligence unit to the information processing terminal, reflecting the characteristics of the learner. [Effects of the Invention]

[0007] According to one aspect of this disclosure, it is possible to immediately provide answers that are appropriate to the learner's characteristics. [Brief explanation of the drawing]

[0008] [Figure 1] This is a block diagram showing an example configuration of a question reception system according to the first embodiment. [Figure 2] This is a block diagram showing the hardware configuration of the question reception system. [Figure 3] Figure 1 shows a table illustrating an example of the learning record data stored in the memory unit. [Figure 4] This flowchart shows an example of the operation procedure of the question reception system according to the first embodiment. [Figure 5] Figure 4 is a flowchart illustrating an example of the procedure for providing responses. [Figure 6] Figure 4 is a flowchart illustrating an example of the procedure for the learning process to overcome weaknesses. [Figure 7] Figure 4 is a flowchart illustrating an example of the data collection process procedure. [Figure 8] Figure 1 is a flowchart showing an example of the procedure for the additional learning process performed by the learning unit. [Figure 9] This flowchart shows another example of the procedure for the additional learning process performed by the learning unit shown in Figure 1. [Figure 10] This is a block diagram showing an example configuration of a question reception system according to the second embodiment. [Figure 11] This flowchart shows an example of the operation procedure of the question reception system according to the second embodiment. [Modes for carrying out the invention]

[0009] The question reception system of this embodiment will be described below with reference to the drawings.

[0010] <First Embodiment> The question-receiving system of this embodiment is a computer system that provides students with lecture content consisting of video and audio through a virtual person (hereinafter referred to as "virtual instructor") that has the appearance of a lecturer. In this embodiment, the computer system will be described in the case where it is used, for example, for a learner to study subjects for a qualification exam with the aim of passing the exam. The configuration of the question reception system of this embodiment will be described with reference to the figures. Figure 1 is a block diagram showing an example configuration of the question reception system according to the first embodiment.

[0011] As shown in Figure 1, the question reception system 10 includes an immediate response unit 11, an AI (artificial intelligence) unit 12, a learning unit 13, a setting unit 14, and a storage unit 15. The storage unit 15 stores pre-prepared data necessary for performing the service and information registered through the performance of the service. The immediate response unit 11, the AI ​​unit 12, the learning unit 13, and the setting unit 14 have their processing defined by software programs, which are implemented by a processor executing these software programs. Details of each unit, data, and information will be described later.

[0012] Figure 2 is a block diagram showing the hardware configuration of the question reception system. Referring to FIG. 2, the question reception system 10 is composed of a server 20 and an information processing terminal 30 such as a personal computer or a smartphone that can be connected to the server 20 via a communication network such as the Internet. The information processing terminal 30 is an information processing device operated by a learner 90 who takes lecture content.

[0013] The information processing terminal 30 has a camera 31, a microphone 32, a speaker 33, a display 34, a memory (not shown), and a processor (not shown). The memory stores a browser 35 which is a browsing application software program. The processor executes the browser 35 stored in the memory. The learner 90 connects to the server 20 using the browser 35 on the information processing terminal 30 and uses the services provided by the server 20. The services are services that provide content related to learning. The services are realized by each part shown in FIG. 1.

[0014] As shown in FIG. 2, as hardware, the server 20 has a processing device 21, a main memory 22, a storage device 23, a communication device 24, an input device 25, and a display device 26, and they are connected to a bus 27.

[0015] The storage device 23 stores data in a writable and readable manner. The storage unit 15 shown in Figure 1 is realized by this storage device 23. The storage device 23 is, for example, an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The processing unit 21 is a processor that reads the data stored in the storage device 23 into the main memory 22 and uses the main memory 22 to execute the processing of the software program. The processing unit 21 is, for example, a CPU (Central Processing Unit). The processing unit 21 realizes the instant response unit 11, AI unit 12, learning unit 13, and setting unit 14 shown in Figure 1. The communication device 24 transmits information processed by the processing unit 21 via a communication network including wired, wireless, or both, and transmits information received via the communication network to the processing unit 21. The received information is used by the processing unit 21 for software processing. The input device 25 is a device that receives information from operator input such as a keyboard or mouse, and the input information is used by the processing unit 21 for software processing. The display device 26 is a device that displays image or text information on a display screen in conjunction with software processing by the processing device 21. The input device 25 and the display device 26 are provided primarily for use by administrators (not shown), not learners 90.

[0016] The data stored in the memory unit 15 shown in Figure 1 will be explained below. The storage unit 15 stores teaching material data, learning record data, learning assistance data, and an answer model. The teaching material data includes lecture content, data of textbooks used in the lecture, and explanatory content that is content for explaining a target item for each of a plurality of items described in the textbook. Further, the teaching material data may include a confirmation test for confirming a plurality of items described in the textbook, and answers to a plurality of questions included in the confirmation test. Furthermore, the teaching material data may include problem exercises with a higher level of difficulty than the confirmation test, answers to a plurality of questions included in the problem exercises, and explanatory content that is content for explaining the answers to the plurality of questions, regarding the content described in the textbook.

[0017] Figure 3 is a table showing an example of the learning record data stored in the storage unit shown in Figure 1. In the table, question-and-answer data, evaluations, learning levels, strong / weak areas, learning processes, and examination experiences are recorded corresponding to a learner ID, which is a different identifier for each learner.

[0018] The question-and-answer data is data in which questions that learner 90 corresponding to the learner ID has made in the past and answers generated by the AI unit 12 are combined. For the case where the learner's examination qualification is a real estate broker, two specific examples of combinations of questions and answers will be described.

[0019] (Question 1) · Question target: <Teaching material> Textbook "Real Estate Business Law", <Question number or page, etc.> Pages 114 - 116 · Question content: "Regarding the business guarantee deposit and the settlement business guarantee deposit that appear in the text, it is stated that a notification is required to the licensing authority everywhere, but who is the licensing authority?"

[0020] (Answer 1) "To conduct real estate business, its license is required. The explanation of the licensing authority is on pages 98 - 101 of the real estate business law textbook. Briefly speaking, as shown in the figure on page 99 of the textbook, when the office is in only one prefecture, the governor of that prefecture is the licensing authority, and when the office is in two or more prefectures, it is the Minister of Land, Infrastructure, Transport and Tourism."

[0021] (Question 2) • Question subject: <Textbook> "Real Estate Brokerage Law Textbook 1", <Question number or page number, etc.> page 118 Question: "I have a question regarding the difference between recovering business guarantee deposits and compensation guarantee deposits. For business guarantee deposits, if the deposit exceeds a certain amount due to the closure of a branch, 'public notice is required.' For compensation guarantee deposits, if the amount exceeds a certain amount due to the closure of some offices, 'public notice is not required.' What is the reason for this difference in requirements?"

[0022] (Answer 2) "For example, let's say Company A has its head office and one branch office, and has deposited 15 million yen as a business guarantee. In this case, the maximum amount that customers can receive back is 15 million yen. If Company A were to close the branch office and immediately recover the business guarantee (5 million yen), the maximum amount that customers can receive back would drop to 10 million yen. In order to prevent the maximum amount that customers can receive back from immediately decreasing even if a branch office is closed, and to inform customers of this, it is necessary to set a period of six months or more and make a public announcement stating that they should claim their rights. What the public announcement means is, 'Those who are entitled to a refund should exercise their rights as soon as possible. If you do not act quickly, you may not be able to receive a refund from the business guarantee.'" On the other hand, let's say Company B also has a head office and one branch, and has paid 900,000 yen as its contribution to the compensation fund. In this case, the maximum amount that customers can receive as a refund is 15 million yen, just like with Company A. The reason why the maximum refund amount is 15 million yen, even though only 900,000 yen has been paid, is because of the compensation fund contributions paid by other members. For example, if the guarantee association has 1,000 member companies (real estate agents), and each company pays a compensation fund contribution of 900,000 yen, the guarantee association will have a total of 900 million yen deposited. Even if Company B closes its branch and its compensation fund contribution (300,000 yen) is immediately recovered, it will not have a significant impact on the total of 900 million yen. Therefore, if Company B simply closes its branch and remains a member of the guarantee association, it is possible to recover the funds immediately without any public announcement.

[0023] Figure 3 shows a case where the text of the question and answer are recorded in a table, but the location where the question and answer data is recorded is not limited to the table shown in Figure 3. For example, the setting unit 14 may assign an identifier to the text data of the question and answer, record the assigned identifier in the learning record data, and store the text data of the question and answer in a separate storage area in the storage unit 15. As mentioned above, the text of the question may include information about the subject of the question, not just the content of the question.

[0024] The evaluation column in the table shown in Figure 3 records the learner's evaluation of the response. The evaluation is a five-point scale, for example, from rank 1 to rank 5. Rank 5 is the highest evaluation, and rank 1 is the lowest. Specific examples of ranks 5 through 1 are as follows: Rank 5: Very easy to understand Rank 4: Fairly easy to understand Rank 3: Normal Rank 2: A little difficult to understand Rank 1: Very difficult to understand

[0025] Referring to Figure 3, for example, in the case of learner ID Gs2, two questions were asked during the learning process, and of the two answers, one answer was given a rank 5 evaluation and the other answer was given a rank 4 evaluation. In this embodiment, the explanation is given using a 5-point evaluation scale, but the ranking of evaluations is not limited to 5 levels.

[0026] The learning level column records the learning level of each of the 90 learners. Figure 3 shows the case where the learning level is divided into three stages: upper, middle, and lower, but the division of learning levels is not limited to three stages. There may be two stages, or even four or more stages. The "Strengths / Weaknesses" column lists the subjects in which learner 90 excels and struggles. In this embodiment, there are five subjects: T, J, S, K, and F. Referring to Figure 3, for example, learner 90 with learner ID Gs1 is shown to be strong in subjects T and J, but weak in subject S. The subjects of strengths and weaknesses are not limited to subjects, but may also be learning areas that constitute a part of a single subject. Hereafter, the information on strengths and weaknesses will be referred to as strengths and weaknesses information. The learning process indicates the portion of the entire lecture content that the 90 learners are currently viewing. For example, if the entire lecture content consists of 30 sessions, it would be divided into three parts: the first half (sessions 1-10), the middle (sessions 11-20), and the second half (sessions 21-30). In the "Exam Experience" column, "None" is recorded if Learner 90 has never taken the actual exam, and "Yes" is recorded if Learner 90 has taken the actual exam at least once. Learning support data includes information on virtual instructors, serving as supplementary data to the learning materials. Virtual instructors act as intermediaries, answering learners' questions on behalf of real instructors. Virtual instructors are created to match the learners' preferences. This information is linked to the learner ID and registered in the learning support data.

[0027] The configuration of the instant response unit 11, AI unit 12, learning unit 13, and setting unit 14 shown in Figure 1 will be briefly explained. The instant response unit 11 inputs the questions received from the learner's information processing terminal 30 into the AI ​​unit 12. The instant response unit 11 reflects the learner's characteristics in the answers obtained from the AI ​​unit 12 and transmits them to the information processing terminal 30. Learner characteristics are a concept that comprehensively represents the individuality and traits of each learner, and include elements such as emotional aspects, learning progress, learning level, and subjects in which they excel / weak. The AI ​​unit 12 generates answers to questions using an answer model. The learning unit 13 updates the answer model using training data. The setting unit 14 records the evaluation of the answers in the learning record data stored by the storage unit 15. The functions of each part will be explained in accordance with the operation of the question reception system 10, and a detailed explanation will be omitted here.

[0028] Next, the operation of the question reception system 10 of this embodiment will be described. Figure 4 is a flowchart showing an example of the operation procedure of the question reception system according to the first embodiment.

[0029] In step S101, the server 20 receives a question from the learner 90's information processing terminal 30. The data format of the question is not limited to text; it may also be audio, a still image, or a video. The question format is not limited to text; it may also be an audio dialogue using the learner 90's voice, or a face-to-face meeting using facial images and audio. If the question format is an audio dialogue, the information processing terminal 30 transmits the learner 90's audio data, captured by the microphone 32, to the server 20. If the question format is a face-to-face meeting, the information processing terminal 30 transmits the learner 90's facial image data, captured by the camera 31, and the learner 90's audio data, captured by the microphone 32, to the server 20. The image data may be a still image or a video.

[0030] In step S102, the server 20 performs the response provision process. The response provision process will be explained in detail with reference to Figure 5. In step S103, the server 20 performs the weakness overcoming learning process. The weakness overcoming learning process will be explained in detail with reference to Figure 6. In step S104, the server 20 performs the data collection process. The data collection process will be explained in detail with reference to Figure 7.

[0031] Figure 5 is a flowchart showing an example of the procedure for providing responses as shown in Figure 4. In step S201, when the instant response unit 11 receives selection information from the learner 90 via the information processing terminal 30 to select a virtual instructor, it selects a virtual instructor from the learning support data according to the selection information. The instant response unit 11 obtains information about the selected virtual instructor from the learning support data. The selection information is, for example, the learner ID. In step S202, the instant response unit 11 inputs the question to the AI ​​unit 12. If the question format is text via email, the instant response unit 11 inputs the text data to the AI ​​unit 12. If the question format is voice dialogue, the instant response unit 11 inputs the voice data to the AI ​​unit 12. If the question format is a face-to-face meeting, the AI ​​unit 12 inputs the learner's facial image data and voice data.

[0032] Regarding step S202, we will explain the case of question 2, one of the two specific examples of the questions mentioned above. In step S101, the email received from the information processing terminal 30 contains the following questions in addition to the description of the subject of the question. "thank you always. I have a question regarding the difference between recovering a business guarantee deposit and a compensation business guarantee deposit. Regarding the business guarantee deposit, it states that "public notice is required" if the business guarantee deposit exceeds the limit due to the closure of a branch. However, regarding the compensation business guarantee deposit, it states that "public notice is not required" if the amount of the contribution exceeds the limit due to the closure of some offices. What is the reason for this difference in requirements? In this case, the instant response unit 11 takes the above question content, removes the part "Thank you for your continued support," which is not directly related to the question, and inputs it into the AI ​​unit 12.

[0033] In step S203, when the AI ​​unit 12 receives a question from the instant response unit 11, it analyzes the learner's emotions from the question. If the question format is a voice dialogue, the AI ​​unit 12 determines whether the learner is in a good mood, for example, from the volume and / or frequency of the learner's voice. When a person is in a bad mood, their voice tends to become quieter and their voice frequency lower. The AI ​​unit 12 determines that the learner is in a bad mood if the volume of the learner's voice is lower than a predetermined volume threshold. The AI ​​unit 12 also determines that the learner is in a bad mood if the frequency of the learner's voice is lower than a predetermined frequency threshold. The AI ​​unit 12 may also determine the learner's mood based on the speed of the learner's speech. When a person is in a bad mood, their speech speed tends to increase, so the AI ​​unit 12 determines that the learner is in a bad mood if their speech speed is faster than a predetermined threshold speed. An example of when a learner is in a bad mood is when the actual exam is approaching. If the question format is a face-to-face meeting, the AI ​​unit 12 determines whether the learner is highly motivated to learn based on the size of the learner's facial outline. When a learner is highly motivated to learn, they tend to lean forward, so the outline of the facial image captured by the camera of the information processing terminal 30 becomes larger. Therefore, if the size of the learner's facial outline is larger than a predetermined standard size, the AI ​​unit 12 determines that the learner is highly motivated. The image of the learner's face may be a still image or a moving image. The AI ​​unit 12 outputs the results of the sentiment analysis to the immediate response unit 11. The AI ​​unit 12 also generates an answer to the question using the response model.

[0034] In step S204, the instant response unit 11 obtains the answer from the AI ​​unit 12. The instant response unit 11 obtains the results of the sentiment analysis from the AI ​​unit 12. If the sentiment analysis results indicate that the learner is in a bad mood, such as having low motivation to learn, the instant response unit 11 requests the AI ​​unit 12 to make the answer more empathetic to the learner's feelings. Specifically, along with the "analysis results" and "answer" received from the AI ​​unit 12, the instant response unit 11 inputs a request to the AI ​​unit 12 to make the "answer" more respectful of the learner's mood. For example, along with the answer, the instant response unit 11 inputs a request to the AI ​​unit 12 saying, "The learner receiving the answer is in a bad mood, so please make the answer respectful of the learner's mood." The AI ​​unit 12 creates a response that corresponds to the learner's emotions based on the generated response. The instant response unit 11 may, instead of the AI ​​unit 12, request that the response be made to the learner using expressions that resonate with the learner's emotions, instead use an external general-purpose generation AI.

[0035] Here, we will explain specific examples of how to create answers that address the learner's emotions when they are in a bad mood, such as having low motivation to learn. To increase the motivation of learners who are not enthusiastic about learning, several effective answering methods can be considered. Four methods, (1) to (4), are described below as specific examples of effective answering methods. The AI ​​unit 12 applies one or more of the methods (1) to (4) to the answer it generates. If we were to prioritize (1) to (4), (1) would be the most important.

[0036] (1) Show empathy for the learner. By showing understanding for the learner's anxieties and questions and using phrases like, "I completely understand how you feel," the learner will feel understood and will be more likely to open up. (2) Encourage learners to set specific goals. Ask learners, "What do you want to achieve through this lesson?" and help them clarify their own goals. When goals become clear, their motivation to learn increases. Also, by setting small, achievable goals, they can gain successful experiences, which builds confidence. (3) Emphasize the practicality and / or relevance of the learning content to the learner. By explaining to the learner with specific examples how this knowledge will be useful in the future, you can make them feel that learning is worthwhile. For example, you can pique their interest by presenting a specific scenario such as, "This formula is very important in real business situations." (4) Give learners positive feedback. Praising learners for the questions they ask and the effort they put in can help them gain confidence. It is important to acknowledge learners' thoughts and actions with phrases such as, "That's a very good point of view."

[0037] Of the two specific examples of the questions mentioned above, for example, if (4) above is applied to answer 1 to question 1, AI unit 12 will create an answer that adds the message "That's a very good point of the question" before answer 1. Specifically, AI unit 12 will create an answer that says, "That's a very good point of the question. A license is required to operate a real estate business. An explanation of the licensing authority can be found on pages 98-101 of the Real Estate Business Act textbook. Simply put, as shown in the diagram on page 99 of the textbook, if the office is located in only one prefecture, the governor of that prefecture is the licensing authority, and if the office is located in two or more prefectures, the Minister of Land, Infrastructure, Transport and Tourism is the licensing authority."

[0038] In step S205, the immediate response unit 11 determines whether the learner 90's learning level is high or low. If the assessment in step S205 indicates that the learner's learning level is high, then the learner's strengths and weaknesses are assessed (step S206). If the result of the determination in step S206 indicates that the subject of the question belongs to the learner's area of ​​expertise, the immediate response unit 11 provides an answer to the information processing terminal 30 (step S207). Specifically, the immediate response unit 11 has the virtual instructor selected by the learner 90 in step S201 read the answer aloud, displays the virtual instructor's video on the display 34 via the browser 35 of the information processing terminal 30, and outputs the answer as audio from the speaker 33.

[0039] On the other hand, if the determination in step S205 indicates that the learner's learning level is not high, the immediate response unit 11 proceeds to the process in step S208. If the determination in step S206 indicates that the subject of the question belongs to an area in which the learner is weak, the immediate response unit 11 proceeds to the process in step S208. In step S208, the immediate response unit 11 obtains explanatory content of basic matters from the teaching material data and provides the obtained explanatory content to the information processing terminal 30. Specifically, the immediate response unit 11 has the virtual instructor selected by the learner 90 in step S201 read the explanation aloud, displays the virtual instructor's video on the display 34 via the browser 35 of the information processing terminal 30, and outputs the explanation as audio from the speaker 33. After the process in step S208, the immediate response unit 11 has the virtual instructor output the answer as described above (step S207).

[0040] Thus, if the assessment in step S205 indicates a high learning level, and the assessment in step S206 indicates that it is a strong area, the instant response unit 11 should send the answer to the information processing terminal 30 as content to be provided to the learner in step S207. On the other hand, if the assessment in step S205 indicates that the learning level is not high, or if the assessment in step S206 indicates that it is a weak area, the instant response unit 11 should send explanatory content in addition to the answer to the information processing terminal 30 as content to be provided to the learner.

[0041] Note that the server 20 does not have to perform all of the steps S201, S203, S205, and S206 shown in Figure 5. It only needs to perform at least one of these steps. For example, if the instant response unit 11 does not receive selection information from the information processing terminal 30 to select a virtual instructor, it does not need to perform the process in step S201. Also, if the result of the determination in step S205 indicates that the learning level is high and the determination in step S206 is not performed, the instant response unit 11 may proceed to the process in step S207.

[0042] This section explains how the aforementioned question reception system 10 can solve conventional problems. Conventional problems will be explained using cases 1 through 5. (Problem 1: It takes time from receiving a question to providing an answer.) When instructors receive questions from learners via email and then prepare and send answers back to learners via email, the replies may not arrive until the day after the question was received. One reason for this is the limited number of instructors who can answer questions from a large number of learners. In particular, questions tend to concentrate just before the actual exam, making it difficult for instructors to reply to all questions by the day after they are received. On the other hand, learners can maintain their motivation to learn if they receive answers to their questions quickly, as their doubts are resolved immediately. Conversely, the longer the response is delayed, the more the learner's motivation to learn diminishes. Therefore, it is necessary to provide answers to questions to learners as quickly as possible.

[0043] (This embodiment addresses Problem 1: AI-powered instant response) In this embodiment, the question reception system 10 has an AI unit 12 that automatically generates answers to questions, and an immediate response unit 11 outputs the answers generated by the AI ​​unit 12 to the learner's information processing terminal 30. Instead of a real instructor, the AI ​​unit 12 on the server 20 generates answers to questions based on teaching material data, so answers can be provided to learners quickly, regardless of when the question was received.

[0044] In this embodiment, if an answer to the same question received from the learner has been generated in the past, the previously generated answer may be used. This method will be explained in detail. Assume that the learning record data contains question answer data and evaluation information for answers given in the past to the same question as the newly received question. In this case, the immediate response unit 11 outputs to the learner's information processing terminal 30 any answers given in the past to the same question that have an evaluation of or higher than a predetermined threshold value. The threshold value is, for example, rank 5. Since previously generated answers can be used, answers can be provided to the learner quickly. Also, since the answers have an evaluation of rank 5, high-rated answers can be provided to the learner. However, depending on the learning scope, there may be few answers with a rank of 5. Therefore, the threshold value may be rank 4. The threshold value may be changed according to the learning scope to which the question is applied.

[0045] (Challenge 2: Learners cannot choose who they answer.) Learners tend to be more motivated to learn and more likely to ask questions when they receive answers from instructors who match their preferences. The number of instructors who teach a single exam is limited. As with any human interaction, there are differences in compatibility, so learners may not necessarily find that they are compatible with the instructor who answers their questions.

[0046] (This embodiment addresses Problem 2: A virtual instructor provides answers tailored to the learner's preferences.) Many people have the experience of diligently studying subjects taught by their favorite teachers during their school days. While it's not always possible for learners to find an instructor that suits their preferences when dealing with real-life instructors, it's highly likely to be possible with virtual instructors. The question-receiving system 10 of this embodiment allows learners to obtain answers from virtual instructors that match their preferences.

[0047] The learner inputs their preferences for each of several conditions into the server 20 via the information processing terminal 30. The setting unit 14 creates a virtual instructor that matches the learner's preferences according to the instructions entered for each of the multiple conditions. The setting unit 14 stores the information of the created virtual instructor along with the learner ID in the storage unit 15. The multiple conditions include, for example, the virtual instructor's appearance, fashion, voice, and personality.

[0048] Appearance refers to information describing the appearance of the virtual instructor. For example, the virtual instructor will be assigned an appearance selected by the learner from several pre-prepared appearances. Fashion refers to information describing the clothing of the virtual instructor. For example, the fashion will be assigned an outfit selected by the learner from several pre-prepared outfits. The voice is information that represents the voice emitted by the virtual instructor. For example, the virtual instructor is assigned a voice selected by the learner from several pre-prepared voices. Personality is information that represents the personality of the virtual instructor. For example, a virtual instructor may be assigned a personality selected by the learner from several pre-prepared personalities.

[0049] This allows learners to create virtual instructors with appearances, fashion sense, voices, and personalities that they prefer. Learners can then receive answers from the virtual instructors they have created. While the virtual instructor is providing answers, learners may change the settings for multiple criteria. In this case, learners can update to create a more ideal virtual instructor.

[0050] (Challenge 3: Inability to provide answers that take into account the learner's emotions when asking questions) Learners are human beings, and they have good days and bad days. Naturally, a learner's mental state varies from time to time. In particular, many people become mentally unstable right before the actual exam. Therefore, when an instructor receives a question right before the exam, it is desirable for them to anticipate the learner's emotions and respond appropriately to those emotions. When an instructor receives a question from a learner via email, they reply with the answer via email. In this case, it is difficult for the instructor to anticipate the learner's emotions from the content of the email text and take those emotions into consideration when answering each time a question is received.

[0051] (This embodiment addresses Problem 3: Sentiment Analysis) The question reception system 10 of this embodiment infers the learner's emotions from their questions and provides an appropriate response based on those emotions. Furthermore, the question reception system 10 of this embodiment is not limited to text-based questions, but also supports voice dialogues and face-to-face meetings. The AI ​​unit 12 analyzes the questioner's emotions in accordance with the question format and generates a response that is in line with the emotions derived from the analysis. The question format can be, for example, text, voice dialogue, or face-to-face meeting.

[0052] If the question format is text, the AI ​​unit 12 infers the learner's emotions from the words used. If the question format is voice dialogue, the AI ​​unit 12 infers the learner's emotions from the state of the learner's voice and the words used in the conversation. In this case, the AI ​​unit 12 can obtain information such as the strength, speed, and / or intonation of the learner's voice when the question is asked, so it can infer emotions more accurately than when the question format is text. If the question format is a face-to-face meeting, the AI ​​unit 12 analyzes the learner's facial expressions and infers the learner's emotions from the results of the facial expression analysis and the words used. In this case, the AI ​​unit 12 can obtain information on the learner's facial expressions in addition to the learner's voice when the question is asked, so it can infer emotions more accurately. The question receiving system 10 of this embodiment can provide answers that correspond to the learner's emotions at the time of the question.

[0053] (Challenge 4: Questions that are difficult to express in writing are difficult to ask via email alone.) The questions from learners cover a wide range of topics, including textbooks and lectures. Some questions concern the content of the textbook, while others concern what the instructor explained orally during lectures. Consider a scenario where a student asks a question via email about something written in a textbook. In this case, the student must include "the '...' written on page XX, line YY of the textbook" as the subject of their question in the email message field, followed by their question. As a prerequisite for their question, the student must transcribe the information written in the textbook. Let's also consider the case where a student asks a question via email about something the instructor explained orally during a lecture. In this case, the student must include in the email message field the subject of the question, "In the lecture for the XXth session, the instructor said '...' at △△ minutes from the start of the lecture," followed by the question itself. As a prerequisite for the question, the student must transcribe what the instructor said orally. Thus, if learners have to ask questions about all the learning materials via email, it places a heavy burden on them.

[0054] (This embodiment addresses issue 4: multimodal question reception) When learners ask questions, various appropriate methods exist depending on the subject, the content of the question, the level of the question, and the learner's level of study. For example, if a learner asks a question about something in a mathematics textbook, a face-to-face meeting is ideal. Asking questions about mathematical content via email is quite burdensome for the learner. Therefore, the question reception system 10 of this embodiment is designed to handle questions from learners in any of three question formats: text, voice dialogue, and face-to-face meeting. Learners can choose the format that is least burdensome and easiest for them to ask questions in from the three options.

[0055] (Challenge 5: Even with the same answer, evaluations differ among learners.) Even when the same question is answered with the same response, learners will evaluate the answer differently. One reason for this is the difference in academic ability among learners. If the same answer is provided to all learners, those whose answers do not match their learning level will not be satisfied. Therefore, it is desirable to provide answers that are appropriate to the learner's learning level. Another reason is that learners have different areas of strength and weakness. Even if a learner's learning level is low, they are likely to be satisfied with the answer if it relates to their area of ​​strength. On the other hand, even if a learner's learning level is high, they are less likely to be satisfied with the answer if it relates to their area of ​​weakness. Therefore, it is desirable to provide answers that are appropriate to both the learner's areas of strength and weakness.

[0056] (This embodiment addresses Problem 5: Generating answers that match the learner's learning level) The question reception system 10 of this embodiment generates answers according to the learner's learning level. Each learner's learning level is recorded in the learning record data, as explained with reference to Figure 3. An example of how to record each learner's learning level is described below. For each course, a test is administered within the curriculum to all learners taking that course, and each learner's test score is recorded in the learning record data. When learners have a low level of understanding, they often lack a grasp of fundamental knowledge. Therefore, when answering questions from low-level learners, it is desirable to provide information on basic concepts to help them confirm the necessary background knowledge before answering their questions. Conversely, when answering questions from high-level learners, it is desirable to answer them concisely without providing information on basic concepts. This is because prompting high-level learners to confirm basic concepts before answering is likely to lead to dissatisfaction, with them wishing for a more concise answer. Therefore, this embodiment takes the above-mentioned learner characteristics into consideration and changes the way of answering depending on the learner's learning level. For example, each learner's learning level is classified into either an upper or lower rank, and the style of answering is changed according to the learner's learning level. When providing learners with information on basic matters, the basic matters may be provided together with the answers.

[0057] (This embodiment addresses Task 5: generating answers that match the learner's strengths and weaknesses) Most people have areas they excel at and areas they struggle with when it comes to learning. Even high-achieving students inevitably have areas they struggle with. Only a small number of people have absolutely no areas they are weak in. The question reception system 10 of this embodiment generates an answer depending on whether the question belongs to the learner's area of ​​expertise or area of ​​weakness. Information on each learner's area of ​​expertise and area of ​​weakness is recorded in the learning record data, as explained with reference to Figure 3. An example of how to record information on each learner's area of ​​expertise and area of ​​weakness is described below. A questionnaire is distributed to all learners who apply for the course, asking them to fill in their area of ​​expertise and area of ​​weakness, and each learner's questionnaire results are recorded in the learning record data. Alternatively, the AI ​​unit 12 may determine whether a learner's area of ​​expertise or area of ​​weakness is determined based on the learner's test results and record the determination result in the learning record data. For example, if a learner's test score in a certain learning area is significantly worse than that in other learning areas, the AI ​​unit 12 will determine that the learning area with the poor score is a weak area. This embodiment provides learners with information on fundamental points to assist in answering questions that fall under areas in which the learner is weak, in addition to providing answers to the questions themselves.

[0058] Next, we will explain the weakness-overcoming learning process shown in Figure 4. Figure 6 is a flowchart showing an example of the steps for the weakness-overcoming learning process shown in Figure 4. In step S301, the AI ​​unit 12 determines the learner's proficiency level from the learner's questions. There are three levels of proficiency: A, B, and C. The relationship between these three levels of proficiency is A <B<Cである。

[0059] A proficiency level of A indicates that learner 90 does not yet possess basic skills in the subject matter of the question. If the AI ​​unit 12 determines that the question concerns fundamental aspects of the subject matter, it assigns learner 90 a proficiency level of A. A proficiency level of B indicates that learner 90 has a basic understanding of the learning scope covered by the question, but lacks the ability to apply that understanding. If the AI ​​unit 12 determines from the question that learner 90 understands the basic concepts but not the applied concepts, it assigns learner 90 a proficiency level of B. A proficiency level of C indicates that the learner 90 possesses considerable ability in the target learning area. The AI ​​unit 12 determines from the questions that the learner 90 has both basic and applied skills, and based on these basic and applied skills, if it determines that the learner can apply the knowledge in the target learning area to other fields, it assigns a proficiency level of C to the learner 90. In this embodiment, the explanation is given by classifying the learner's proficiency in the learning area into three levels: A, B, and C, but the number of levels is not limited to three.

[0060] If, as a result of the judgment in step S301, the AI ​​unit 12 determines that the learner's proficiency level is A, the instant response unit 11 provides explanatory content for the subject area to be studied to the information processing terminal 30 (step S302). In step S303, the instant response unit 11 conducts a confirmation test for the learner 90.

[0061] If, as a result of the judgment in step S301, the AI ​​unit 12 determines that the learner 90's proficiency level is B, the immediate response unit 11 provides practice problems in the subject area to be learned to the information processing terminal 30 and prompts the learner 90 to complete the practice problems (step S304). When the immediate response unit 11 receives the answers to the practice problems from the information processing terminal 30, it compares the correct answers with the learner 90's answers and extracts the problems that were answered incorrectly from the practice problems. In step S305, the immediate response unit 11 provides the information processing terminal 30 with explanatory content for the problems that were answered incorrectly.

[0062] This section explains how the question reception system 10 of this embodiment can solve the problems of the conventional system. The conventional problem will be referred to as Problem 6. (Challenge 6: Answers questions but does not go so far as to provide learning guidance to the students.) Instructors are often overwhelmed with answering questions from many students, making it difficult for them to provide adequate instruction. Therefore, even if instructors notice a student's weak areas based on their questions, they often lack the time to offer further guidance. Many people lack the motivation to address their weak areas, making it difficult to overcome them on their own. Leaving weak areas unaddressed will prevent academic improvement. Therefore, it is desirable to encourage students to engage in learning activities to overcome their weak areas at the same time as answering their questions.

[0063] (This embodiment addresses Problem 6: It proposes a learning plan for overcoming weaknesses for each learner.) The question reception system 10 of this embodiment determines the learner's level of proficiency in the subject matter based on the content of the question, proposes a learning plan to the learner to overcome their weaknesses according to their level of proficiency, and guides them through the learning process to overcome their weaknesses. Since the learner receives a learning plan to overcome their weaknesses at the same time they receive an answer to their question, they can smoothly work on overcoming their weaknesses according to the learning plan. In this embodiment, the immediate response unit 11 proposes the following learning plan to learners with proficiency levels A and B, respectively, and has them immediately begin learning according to the learning plan. For proficiency level A, the learning plan involves having the learner immediately watch lecture videos on areas they struggle with to review their basic skills, followed by a short test to ensure knowledge retention. For learners at proficiency level B, a combination of foundational skills and applied skills is necessary to solve problems equivalent to those on the actual exam. Therefore, the learning plan for proficiency level B involves having learners first solve multiple problems, and then having them watch lectures on how to solve the problems they were unable to solve.

[0064] Next, we will explain the data collection process shown in Figure 4. Figure 7 is a flowchart showing an example of the procedure for the data collection process shown in Figure 4. In step S401, the setting unit 14 receives an evaluation of the answer from the information processing terminal 30 and accepts the evaluation. In step S402, the setting unit 14 stores the question and answer data, which combines the question and answer, in the learning record data. The immediate response unit 11 also records the received evaluation result information in the learning record data. In this way, the evaluation of the answer is stored in the learning record data.

[0065] Next, we will explain the additional learning process performed by the learning unit 13. First, we will explain the case where the learning unit 13 performs fine tuning. Figure 8 is a flowchart showing an example of the procedure for the additional learning process performed by the learning unit shown in Figure 1. In step S501, the learning unit 13 refers to the learning record data and extracts question answer data for which the answer is rated 5. In step S502, the learning unit 13 adds the extracted question answer data to the training data of the answer model. In step S503, the learning unit 13 fine-tunes the response model using the training data.

[0066] Next, we will explain the case where the learning unit 13 performs context learning. Figure 9 is a flowchart showing another example of the procedure for the additional learning process performed by the learning unit shown in Figure 1. In step S601, the learning unit 13 refers to the learning record data and extracts question answer data for which the answer is rated 5. In step S602, the learning unit 13 adds the extracted question answer data to the training data of the answer model. In step S603, the learning unit 13 updates the prompt with the training data.

[0067] This section explains how the question reception system 10 of this embodiment can solve the problems of the conventional system. The conventional problem will be referred to as Problem 7. (Issue 7: The quality of responses is inconsistent) For example, during periods when many questions are concentrated, such as immediately before the actual exam, in addition to the instructors, those who have passed the exam may join the answering staff to help answer questions. In this case, differences in knowledge and experience between the instructors and the answering staff can lead to variations in the quality of the answers provided by the instructors and the answering staff.

[0068] (This embodiment addresses Problem 7: Generating responses of uniform quality) In this embodiment, the question receiving system 10 has a learning unit 13 that updates the answer model for generating answers from questions using question-answer data, which consists of combinations of questions and answers from past responses that have an evaluation of or higher than a predetermined threshold value, as training data. Therefore, the quality of answers is maintained to improve. Since the AI ​​unit 12 generates answers using the answer model, the quality of the generated answers becomes uniform. The threshold value is, for example, rank 5 on a 5-point scale.

[0069] The learning unit 13 updates its answer model using past answers that received a rank of 5. This modifies the algorithm that generates answers from questions, resulting in more optimal answers. To continue improving the quality of answers, the AI ​​unit 12 also asks the learner to rate newly generated answers on a 5-point scale. The learning unit 13 then continuously adds answers with a rank of 5 as training data. This maintains the improvement in the quality of answers.

[0070] (Main configuration and effects of the system of this embodiment) The question reception system 10 of this embodiment includes a storage unit 15 for storing teaching material data, an AI unit 12 that receives questions about the teaching material data from an information processing terminal 30 of a learner 90 who is learning using the teaching material data, generates answers to the questions based on the teaching material data, and an immediate response unit 11 that outputs the answers generated by the AI ​​unit 12 to the information processing terminal 30, reflecting the characteristics of the learner 90.

[0071] According to this embodiment, it is possible to immediately provide answers that are suitable for the characteristics of the learner 90.

[0072] <Second Embodiment> This embodiment allows learners at the same learning level or stage of learning to share question and answer data. In this embodiment, the same reference numerals are used for components as in the first embodiment, and their detailed descriptions are omitted. The configuration of the question receiving system of this embodiment will be described with reference to Figure 10. Figure 10 is a block diagram showing an example configuration of the question receiving system according to the second embodiment.

[0073] As shown in Figure 10, the question reception system 10 of this embodiment has a question sharing unit 16 in addition to the immediate response unit 11, AI unit 12, learning unit 13, setting unit 14, and storage unit 15 shown in Figure 1.

[0074] Next, the operation of the question reception system 10 of this embodiment will be described. Figure 11 is a flowchart showing an example of the operation procedure of the question reception system according to the second embodiment. In step S701, the question sharing unit 16 extracts question answer data with an evaluation rank of 5 from the learning record data. In step S702, the question sharing unit 16 refers to the learning record data and determines the learner's learning process. In step S703, the question sharing unit 16 refers to the learning record data and determines the learner's learning level. In step S704, based on the learner's learning process and learning level, it selects question and answer data to share with the learner from the learning record data.

[0075] This section explains how the question reception system 10 of this embodiment can solve the problems of the conventional system. The conventional problem will be referred to as Problem 8. (Issue 8: Learners cannot share questions with other students) When people begin studying for certification exams, they tend to have similar questions or doubts about the same topics. If answers to these common questions or doubts can be shared among multiple students, each student's learning efficiency will improve. On the other hand, if we try to share the answers to all questions among the students, the sheer volume of accumulated questions becomes enormous. Learners would have to search through this vast number of questions to find those relevant to their learning, which would be time-consuming. Furthermore, if learners assume that all the accumulated "questions and answers" are useful and try to read through them all, they will end up reading unnecessary information as well. As a result, they waste time. Therefore, it is desirable to provide information shared among students in a way that is tailored to the individual characteristics of each learner.

[0076] (This embodiment addresses Task 8: Students sharing questions with each other who are in the same learning process.) The question reception system 10 of this embodiment provides the learner with question and answer data from other learners who are in the same learning process as the learner. The learner can share common questions or doubts and their answers with other learners, according to their own learning progress. Typically, the content of questions changes as learners progress through their studies, even if they are the same individual. Let's illustrate this with a specific example. When learners begin studying for a certification exam using a computer system, they often ask questions about how to use the materials and understand the basic concepts of the subject matter. Later, after completing all the lessons in a course, learners begin to grasp the overall picture of the exam subjects. At this stage, the content of their questions shifts from general basic learning to specific details. Finally, in the last stage of studying the certification exam materials, learners primarily focus on practicing past exam questions. At this stage, the content of their questions changes to how to solve past questions or how to study before the actual exam. As the content of learners' questions changes as they progress through their studies, it is desirable for learners to share questions and answers from other learners at a similar stage of learning. Therefore, in principle, this embodiment allows learners at similar learning stages to share information about questions and answers from other learners at similar learning stages. However, learners with exam experience do not exhibit the above-mentioned changes in the content of their questions. For this reason, as an exception, questions from learners with exam experience are shared with other learners regardless of their learning stage. In this embodiment, the learning process was described as being divided into three parts: the first half, the middle, and the second half. However, the number of divisions in the learning process is not limited to three.

[0077] (This embodiment addresses Task 8: Students with the same learning level share questions with each other.) The question reception system 10 of this embodiment provides learners with question and answer data from other learners at a similar learning level to the learner's own learning level. Learners can share common questions or doubts and their answers that are appropriate to their own learning level with other learners. The content of learners' questions varies depending on their learning level. Learners with higher learning levels tend to ask high-quality questions. Learners with lower learning levels may feel inferior when they read high-quality questions, thinking, "My learning level is still low." On the other hand, learners with lower academic ability tend to ask low-quality questions. Low-quality questions are often not helpful to learners with higher learning levels. Therefore, it is desirable for learners with similar learning levels to share information about questions and answers. Therefore, in this embodiment, learning levels are divided into three ranks: upper, middle, and lower, and learners of the same rank share questions and answers with each other. Each learner's learning level is ranked, for example, based on their performance on tests administered within the learning curriculum, and registered in the learning record data. In this embodiment, the learning levels were described as being divided into three ranks, but the number of learning level ranks is not limited to three.

[0078] <Note> The embodiments described above are illustrative for explaining the present invention and are not intended to limit the scope of the invention to those embodiments only. Those skilled in the art can implement the present invention in various other forms without departing from the scope of the invention.

[0079] Furthermore, the embodiments described above include the following items. However, the items included in these embodiments are not limited to those listed below.

[0080] (Item 1) A memory unit that stores teaching material data, When an artificial intelligence unit receives a question about the aforementioned teaching material data from a learner's information processing terminal using the aforementioned teaching material data, the artificial intelligence unit generates an answer to the question based on the aforementioned teaching material data. A response unit that outputs the response generated by the artificial intelligence unit to the information processing terminal, reflecting the characteristics of the learner, A question reception system that has the following features. According to this, it is possible to immediately provide answers that are appropriate to the learner's characteristics.

[0081] (Item 2) The memory unit stores information on evaluations by other learners, different from the learner, of past answers to the same question received from the information processing terminal. The aforementioned response section is, Of the responses given in the past, those whose evaluation is above a predetermined threshold are output to the information processing terminal. The question submission system described in item 1. According to this, it is possible to instantly provide highly-rated answers to the same question from the past.

[0082] (Item 3) The memory unit stores information about a virtual instructor, a virtual person modeled after an instructor, which reflects the learner's preferences. The response unit causes the virtual instructor to output the response to the information processing terminal. The question submission system described in item 1 or 2. According to this, learners can get answers to their questions from virtual instructors of their choice, which increases their motivation to learn.

[0083] (Item 4) The memory unit stores information about the learner's learning level as the learner's characteristics. The aforementioned response section is, Select content corresponding to the learning level as the answer to be output to the information processing terminal. A question submission system as described in any one of items 1 to 3. According to this method, learners can easily understand the answers because they receive responses that match their learning level.

[0084] (Item 5) The memory unit stores information about the learner's strengths and weaknesses regarding the subject of study, The aforementioned response section is, As the response to be output to the information processing terminal, select content corresponding to the strengths and weaknesses information. A question submission system as described in any one of items 1 to 3. According to this method, learners can receive answers tailored to their strengths and weaknesses, thus enabling them to understand the content of the answers.

[0085] (Item 6) The memory unit stores learning plans corresponding to each of several levels of proficiency for the subject to be learned. The artificial intelligence unit determines the learner's proficiency level from the multiple proficiency levels based on the aforementioned questions, When the response unit outputs the response to the information processing terminal, it outputs the learning plan corresponding to the proficiency level determined by the artificial intelligence unit to the information processing terminal. A question submission system as described in any one of items 1 to 5. According to this, learners can not only get answers to their questions but also obtain a learning plan that suits their learning level, thereby improving their learning proficiency.

[0086] (Item 7) The answer model, which generates answers from questions, has a learning unit that updates it based on training data. The artificial intelligence unit generates the answer to the question using the answer model, The memory unit stores question-answer data, which is a combination of questions received in the past and the answers to those questions, and information on the evaluation of the answers included in the question-answer data. The aforementioned learning unit, The question-answer data stored in the memory unit, including answers whose evaluation is equal to or greater than a predetermined threshold, is used as training data to update the answer model. A question submission system as described in any one of items 1 to 6. According to this, the answer model is updated based on high-rated answers from the past, improving the quality of answers generated from questions. Learners can then obtain higher-quality answers.

[0087] (Item 8) The aforementioned artificial intelligence department, When the system receives a question from the learner that includes data in at least one of the following data formats: text, audio, still images, and video, it generates the answer corresponding to the question based on the teaching material data. A question submission system as described in any one of items 1 through 7. According to this, learners can ask questions in various data formats.

[0088] (Item 9) The aforementioned artificial intelligence department, Upon receiving the question from the learner, which includes data in at least one data format from text, audio, still images, and video, the system analyzes the learner's emotions from the question. The aforementioned response section is, The artificial intelligence unit is instructed to create a response based on the aforementioned response that corresponds to the results of the analysis as the characteristics of the learner. A question submission system as described in any one of items 1 through 8. According to this, learners can obtain answers that are in line with their own emotions.

[0089] (Item 10) The memory unit stores learning record data for each of the multiple learners, which includes the learner's learning process and question-answer data, which is a combination of the question and the answer. The aforementioned response section is, When transmitting the answer to the learner's information processing terminal, the learning record data is referenced to determine the learner's learning process. The question answer data of other learners in the same learning process as the determined learning process is read from the learning record data. The retrieved question and answer data is transmitted to the information processing terminal. A question submission system as described in any one of items 1 through 9. According to this, learners can share information about questions and their answers with other learners who are in the same learning process as them.

[0090] The memory unit stores learning record data for each of the multiple learners, which includes the learner's learning level and question-answer data, which is a combination of the question and the answer. The aforementioned response section is, When transmitting the answer to the learner's information processing terminal, the learning record data is used to determine the learner's learning level. The question answer data of other learners at the same learning level as the determined learning level is read from the learning record data. The retrieved question and answer data is transmitted to the information processing terminal. A question submission system as described in any one of items 1 through 9. According to this, learners can share information about questions and their answers with other learners at the same learning level as themselves. [Explanation of Symbols]

[0091] 10 Question reception system, 11 Instant response unit, 12 AI (artificial intelligence) unit, 13 Learning unit, 14 Configuration unit, 15 Memory unit, 16 Question sharing unit, 20 Server, 21 Processing unit, 22 Main memory, 23 Storage device, 24 Communication device, 25 Input device, 26 Display device, 27 Bus, 30 Information processing terminal, 31 Camera, 32 Microphone, 33 Speaker, 34 Display, 35 Browser.

Claims

1. A memory unit that stores teaching material data, When an artificial intelligence unit receives a question about the aforementioned teaching material data from a learner's information processing terminal using the aforementioned teaching material data, the artificial intelligence unit generates an answer to the question based on the aforementioned teaching material data. A response unit that outputs the response generated by the artificial intelligence unit to the information processing terminal, reflecting the characteristics of the learner, A question reception system that has the following features.

2. The memory unit stores information on evaluations by other learners, different from the learner, of past answers to the same question received from the information processing terminal. The aforementioned response section is, Of the responses given in the past, those whose evaluation is above a predetermined threshold are output to the information processing terminal. The question reception system according to claim 1.

3. The memory unit stores information about a virtual instructor, a virtual person modeled after an instructor, which reflects the learner's preferences. The response unit causes the virtual instructor to output the response to the information processing terminal. The question reception system according to claim 1.

4. The memory unit stores information about the learner's learning level as the learner's characteristics. The aforementioned response section is, Select content corresponding to the learning level as the answer to be output to the information processing terminal. The question reception system according to claim 1.

5. The memory unit stores information about the learner's strengths and weaknesses regarding the subject of study, The aforementioned response section is, As the response to be output to the information processing terminal, select content corresponding to the strengths and weaknesses information. The question reception system according to claim 1.

6. The memory unit stores learning plans corresponding to each of several levels of proficiency for the subject to be learned. The artificial intelligence unit determines the learner's proficiency level from the multiple proficiency levels based on the aforementioned questions, When the response unit outputs the response to the information processing terminal, it outputs the learning plan corresponding to the proficiency level determined by the artificial intelligence unit to the information processing terminal. The question reception system according to claim 1.

7. The answer model, which generates answers from questions, has a learning unit that updates it based on training data. The artificial intelligence unit generates the answer to the question using the answer model, The memory unit stores question-answer data, which is a combination of questions received in the past and answers to those questions, and information on the evaluation of the answers included in the question-answer data. The aforementioned learning unit, The question-answer data stored in the memory unit, including answers whose evaluation is equal to or greater than a predetermined threshold, is used as training data to update the answer model. The question reception system according to claim 1.

8. The aforementioned artificial intelligence department, When the system receives a question from the learner that includes data in at least one data format from text, audio, still images, and video, it generates the answer corresponding to the question based on the teaching material data. The question reception system according to claim 1.

9. The aforementioned artificial intelligence department, Upon receiving the question from the learner, which includes data in at least one data format from text, audio, still images, and video, the system analyzes the learner's emotions from the question. The aforementioned response section is, The artificial intelligence unit is instructed to create a response based on the aforementioned response that corresponds to the results of the analysis as the characteristics of the learner. The question reception system according to claim 1.

10. The memory unit stores learning record data for each of the multiple learners, which includes the learner's learning process and question-answer data, which is a combination of the question and the answer. The aforementioned response section is, When transmitting the answer to the learner's information processing terminal, the learning record data is referenced to determine the learner's learning process. The question answer data of other learners in the same learning process as the determined learning process is read from the learning record data. The retrieved question and answer data is transmitted to the information processing terminal. The question reception system according to claim 1.

11. The memory unit stores learning record data for each of the multiple learners, which includes the learner's learning level and question-answer data, which is a combination of the question and the answer. The aforementioned response section is, When transmitting the answer to the learner's information processing terminal, the learning record data is used to determine the learner's learning level. The question answer data of other learners at the same learning level as the determined learning level is read from the learning record data. The retrieved question and answer data is transmitted to the information processing terminal. The question reception system according to claim 1.

12. A question reception method performed by an information processing device, When a question regarding the learning material data is received from the learner's information processing terminal while the learner is studying using the learning material data, the system generates an answer to the question based on the learning material data. The generated response is output to the information processing terminal, reflecting the characteristics of the learner. How to submit questions.

Images