Individual optimization education support system, method for supporting individual optimization education, and program for supporting individual optimization education

Abstract

To provide an individual optimization education support system for enabling analysis of a target person on the basis of objective factors and recommendation of a learning method in accordance with characteristics of the target person.SOLUTION: The individual optimization education support system according to the present disclosure includes: behavior information acquisition means for acquiring behavior information relating to behaviors and statements of a target person; analysis means for analyzing behavioral characteristics of the target person on the basis of the behavior information acquired by the behavior information acquisition means; and learning method recommendation means for recommending a learning method in accordance with the behavioral characteristics of the target person analyzed by the analysis means.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an individualized optimal education support system, an individualized optimal education support method, and an individualized optimal education support program. [Background technology]

[0002] Conventionally, a system has been known in which information about subjects, such as children with developmental disabilities, infants, or children in the early grades of elementary school, is shared between the facility that cares for the subjects and their guardians (see Patent Document 1).

[0003] In this regard, it has been pointed out that the above system places emphasis on improving the efficiency of operations at the facility, and there are problems in terms of supporting the development of the target person (see Patent Document 1). Therefore, Patent Document 1 discloses a development support system that can support the development of the target person. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6297712 [Patent Document 2] Utility Model Registration No. 3228155 Summary of the Invention [Problem to be solved by the invention]

[0005] In the development support system disclosed in the above Patent Document 1, the subject's growth information, activity information, etc. are shared between the subject and the facility, and analytical information relating to analysis such as extraction of the types of activities that contributed to the subject's growth is generated based on the growth information and activity information, and advice information relating to advice on future activities, etc. is generated for the subject being analyzed based on the generated analytical information. This is said to enable development support for the subject (see Patent Document 1).

[0006] On the other hand, when it comes to people with developmental disabilities and their guardians or supporters such as instructors, it is often the case that these supporters do not fully understand the people involved, and are often confused about what kind of support is appropriate for them.

[0007] This is thought to be because the supporter subjectively views the words, actions, and behavior of the target person and is unable to evaluate them using objective indicators (elements). This is not only seen between children with developmental disabilities and the facilities that care for them, but can also be seen in all sorts of situations in society, such as between superiors and subordinates in a company, between human resources and internal evaluations, and self-analysis.

[0008] The present invention has been made in consideration of the above-mentioned situation, and aims to provide an individually optimized educational support system that analyzes subjects based on objective factors and makes it possible to recommend learning methods that suit the characteristics of the subjects. [Means for solving the problem]

[0009] In order to solve the above-mentioned problems, the individual optimal educational support system of the present invention comprises a behavioral information acquisition means for acquiring behavioral information regarding the behavior of a subject, an analysis means for analyzing the behavioral characteristics of the subject based on the behavioral information acquired by the behavioral information acquisition means, and a learning method recommendation means for recommending a learning method according to the behavioral characteristics of the subject analyzed by the analysis means. [Effects of the Invention]

[0010] According to the individual optimal education support system of the present invention, it is possible to analyze a subject based on objective factors and recommend a learning method that suits the subject's characteristics. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a diagram showing the overall configuration including an individual optimal education support system according to an embodiment of the present disclosure. [Figure 2]FIG. 2 is a schematic configuration diagram of an individual optimal education support system according to an embodiment of the present disclosure. [Figure 3] FIG. 3 is a schematic diagram showing an example of tags prepared in advance when analyzing behavioral characteristics by the analysis means in the individual optimal education support system according to the embodiment of the present disclosure. [Figure 4] FIG. 4 is a schematic diagram showing an example of tags prepared in advance when analyzing behavioral characteristics by the analysis means in the individual optimal education support system according to the embodiment of the present disclosure. [Figure 5] FIG. 5 is a graph showing the behavioral characteristics of a subject analyzed in the individual optimal education support system according to the embodiment of the present disclosure, divided into cognitive characteristics, personal characteristics, and skill characteristics. [Figure 6] FIG. 6 is a graph showing the mutual association of each element (data) constituting the cognitive characteristics in FIG. [Figure 7] FIG. 7 is a graph showing the mutual association of the elements (data) that make up the personal characteristics in FIG. [Figure 8] FIG. 8 is a graph showing the mutual association of the elements (data) that make up the skill characteristics in FIG. [Figure 9] Figure 9 is a graph showing the mutual associations between the elements (data) that make up the language included in the hard skills in Figure 8. [Figure 10] Figure 10 is a graph showing the interrelationships between the elements (data) that make up the mathematics included in the hard skills in Figure 8. [Figure 11] FIG. 11 is a graph showing the mutual association of each element (data) that constitutes the physics included in the hard skills in FIG. [Figure 12] Figure 12 is a graph showing the mutual relationships between the elements (data) that make up the chemistry included in the hard skills in Figure 8. [Figure 13] FIG. 13 is a graph showing the mutual associations between the elements (data) that make up the organisms included in the hard skills in FIG. [Figure 14]Figure 14 is a graph showing the mutual associations between the elements (data) that make up the geography included in the hard skills in Figure 8. [Figure 15] Figure 15 is a graph showing the interrelationships between the elements (data) that make up the history included in the hard skills in Figure 8. [Figure 16] Figure 16 is a graph showing the interrelationships between the elements (data) that make up civics, which is included in the hard skills in Figure 8. [Figure 17] Figure 17 is a graph showing the interrelationships between the elements (data) that make up the arts included in the hard skills in Figure 8. [Figure 18] FIG. 18 is a graph showing the mutual association of each element (data) constituting the information included in the hard skills in FIG. [Figure 19] FIG. 19 is an overall flow diagram showing the flow of processing by the individual optimal education support system according to the embodiment of the present disclosure. [Figure 20] FIG. 20 is a detailed flowchart showing the specific processing flow included in the "analysis of behavioral characteristics" in FIG. [Figure 21] FIG. 21 is a detailed flow diagram showing the processing after "NO" is selected in step 24 of FIG. [Figure 22] FIG. 22 is a detailed flow diagram showing the processing after "NO" is selected in step 22 of FIG. [Figure 23] FIG. 23 is a detailed flow diagram showing the processing after "NO" is selected in step 41 of FIG. [Figure 24] FIG. 24 is a detailed flowchart showing the specific process flow included in the "study method recommendation" in FIG. [Figure 25] FIG. 25 is a detailed flow diagram showing the processing that takes place after learning using the recommended learning method has ended. [Figure 26] FIG. 26 is a radar chart showing an example of the analysis results of the subject obtained by the skill characteristic analysis processing in FIGS. 20 to 23 and 25. In FIG. [Figure 27] FIG. 27 is a graph showing an example of changes in the self-renewal ability of a subject obtained by the "analysis of self-renewal ability" process in FIG. [Figure 28] FIG. 28 is a pie chart visually showing the proportions of each of the five sensory elements and linguistic elements that make up the cognitive characteristics shown in FIG. [Figure 29] FIG. 29 is a pie chart visually showing the proportion of each of the four elements that make up the personal characteristics shown in FIG. [Figure 30] FIG. 30 is a hardware configuration diagram for realizing the processing performed by the support device constituting the individual optimal education support system according to the embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments described below, and various modifications can be made without departing from the scope of the technical concept.

[0013] [Overall configuration of the individually optimized educational support system disclosed herein] The overall configuration including the individualized optimal education support system 10 (hereinafter sometimes simply referred to as "the system") of the present disclosure will be described with reference to Fig. 1. As shown in Fig. 1, the individualized optimal education support system 10, a subject terminal 50, and a supporter terminal 100 are connected via a network.

[0014] The individually optimized education support system 10 will be described later, but each process by the individually optimized education support system 10 can be realized by, for example, a personal computer (hereinafter referred to as "PC"), a server computer, or a cloud server. The subject terminal 50 and the supporter terminal 100 are each an information processing device such as, for example, a personal computer (hereinafter referred to as "PC"), a notebook PC, a tablet PC, or a smartphone.

[0015] In this embodiment, the subject terminal 50 and the supporter terminal 100 may be owned by the subject and the supporter, respectively, or the subject terminal 50 or the supporter terminal 100 may be shared by the subject or the supporter. In some cases, the individually optimized education support system 10 may have the functions of the subject terminal 50 and the supporter terminal 100, and the subject or the supporter may directly input information to the individually optimized education support system 10 or directly receive information output from the individually optimized education support system 10.

[0016] [Configuration of the individually optimized educational support system disclosed herein] 2 is a schematic configuration diagram of an individualized optimal education support system according to an embodiment of the present disclosure. The schematic configuration of the individualized optimal education support system 10 according to this embodiment will be described below with reference to FIG.

[0017] The individualized optimal education support system 10 comprises speech and behavior information acquisition means 11, physical information acquisition means 12, analysis means 13, learning method recommendation means 14, learning content storage means 15, and behavioral characteristic update means 16. Each of these means will be described below.

[0018] The speech and behavior information acquisition means 11 acquires speech and behavior information related to the speech and behavior of the subject. In this embodiment, the term "subject" refers to a person who needs some kind of support, such as a person with a developmental disorder, a person who needs learning support, a person who needs nursing care, or a person who simply receives help from a third party. In this embodiment, a person who provides support or assistance to these subjects is referred to as a "supporter." Furthermore, "speech and behavior" refers to the subject's words and actions. "Speech and behavior information" includes information related to the subject's words (hereinafter referred to as "speech information") and information related to the subject's actions (hereinafter referred to as "behavior information"). Note that, for a subject who has difficulty speaking for some reason, such as a deaf person, the speech information may be based on the words of a supporter who supports the subject. Furthermore, the behavior information is not limited to information related to the subject's own behavior but may also include information recorded by a supporter observing the subject's behavior (hereinafter referred to as a "behavior record"). Note that comments entered by the subject himself / herself via his / her own information terminal are also included in the behavior record.

[0019] The behavioral information acquisition means 11 acquires speech information and behavioral information by the subject as behavioral information, for example, via the subject terminal 50 described above. The speech information by the subject may be text information input by the subject via the subject terminal 50, or may be text information converted based on speech acquired via a voice input device such as a microphone provided in the subject terminal 50. The behavioral information acquisition means 11 may also acquire, as behavioral information, recordings of observations of the subject's behavior by a supporter on behalf of the subject via the supporter terminal 100 described above. Of the behavioral information, information written in natural language is particularly referred to as "language information."

[0020] As the behavioral information, for example, text information entered by the subject via a diary app or comment information entered via a social networking service (SNS) can be suitably used. The system can acquire data such as text information and comment information via the diary app or SNS used by the subject, for example, by linking with an API (Application Programming Interface). In addition, growth records in which a supporter records the subject's behavior via the diary app or SNS, or minutes in which the subject's remarks are recorded, can also be suitably used as behavioral information.

[0021] At the same time as acquiring the behavioral information, the behavioral information acquisition means 11 acquires information relating to the time required for the subject's behavior identified by the behavioral information (hereinafter referred to as "time information"). This time information can be used for analyzing the subject's cognitive characteristics by the analysis means 13 described later. The time information will be described in detail in the explanation of the analysis means 13 described later.

[0022] The physical information acquisition means 12 acquires physical information of the subject. "Physical information" refers to information obtained from the subject's body, such as the subject's body movements such as gestures and hand movements, the subject's facial expressions and eye movements, or information related to brain waves, heart rate, etc. Examples of suitable physical information acquisition means 12 include a photographing means for photographing the subject's appearance, a wearable electromyography measuring device for detecting the subject's body movements, particularly muscle movements of the entire body, and brain function measurement technologies capable of measuring brain function, such as analytical methods using machine learning for single-molecule measurement to detect single molecules and various brain wave measurement technologies.

[0023] The analysis means 13 analyzes the behavioral characteristics of the subject based on the behavioral information acquired by the behavioral information acquisition means 11. The behavioral characteristics refer to specific properties related to the behavior that the subject may take. These behavioral characteristics include the subject's personal characteristics, cognitive characteristics, and skill characteristics.

[0024] The analysis means 13 analyzes the subject's behavioral information as input information input to the system and the subject's behavioral characteristics as output information output from the system. A desired numerical value is used as the behavioral characteristics (analysis results), which are output information. Specifically, each behavioral characteristic may include multiple elements. For example, cognitive characteristics include six elements: "vision," "hearing," "touch," "taste," "smell," and "language," as described below. If the subject's overall cognitive characteristics are set to 100%, the proportion of each element can be determined based on the analysis results based on the subject's behavioral information so that the sum of the six elements is 100%. For example, if the analysis of the subject's behavioral information shows a tendency toward a strong "visual" element, the numerical value (proportion) for "visual" will be higher than that of other cognitive characteristics. Of course, it is normally impossible for a healthy person to rely solely on "visuality" when perceiving things, but by displaying each element as a percentage, the subject's cognitive tendencies can be understood. Other behavioral traits may be analyzed in a similar manner as the cognitive trait example above.

[0025] Personal characteristics are behavioral characteristics that indicate the individuality of a subject, such as the subject's thinking type and habits, psychological characteristics, emotional type, etc. Personal characteristics include at least the subject's motivation, judgment method, reward characteristics, and behavioral characteristics when no reward is received.

[0026] A judgment method is a characteristic that indicates the criteria a subject uses to make judgments (decisions) when taking a given action. Judgment methods can be classified into two types: judgment methods based on abstract criteria and judgment methods based on concrete criteria. "Judgment methods based on abstract criteria" refer to judgment methods based on the subject's own knowledge. For example, a subject can determine how to confirm an input by understanding and memorizing the statement "Pressing the Enter key will confirm the input" written in a textbook. In contrast, "Judgment methods based on concrete criteria" refer to judgment methods based on the subject's own past experience. For example, a subject can determine that pressing the Enter key will confirm the input because they have a memory of an episode of their brother playing a computer game.

[0027] Reward characteristics, which describe the feelings a subject experiences when engaging in a specific behavior or receiving a specific action, can be analyzed by measuring the levels of neurotransmitters and hormones secreted in the subject's brain. Reward characteristics include stimulation, affection, and happiness. "Stimulation" can be measured by dopamine secretion, "affection" by oxytocin secretion, and "happiness" by serotonin secretion. For example, these neurotransmitters can be detected using machine learning-based analytical techniques for single-molecule measurement (Osaka University research portal "Resou," "World's First! New Method Using Quantum Measurement and AI! Successful High-Speed ​​Detection and Identification of Neurotransmitters," July 9, 2020, https: / / resou.osaka-u.ac.jp / ja / research / 2020 / 20200709_1).

[0028] The behavioral characteristics when no reward is received are characteristics related to the behavior of the subject when no reward is received, and include aggressive and depressive characteristics. The behavioral characteristics when no reward is received can be analyzed from the behavioral records of the subject recorded by the supporter. Specifically, when the supporter prohibits the subject from performing a specific behavior (such as watching television), the behavioral characteristics can be analyzed based on the next behavior (such as anger) that the subject takes.

[0029] Cognitive characteristics refer to characteristics of a subject's perception of things, which can be analyzed from the subject's specific behavior. Cognitive characteristics include at least five-sensory characteristics and language characteristics, and "five-sensory characteristics" include visual characteristics, auditory characteristics, tactile characteristics, taste characteristics, and olfactory characteristics. These five-sensory characteristics can be analyzed based on the subject's statements and behavioral records.

[0030] Skill characteristics refer to the abilities inherent to a subject. Skill characteristics include soft skills and hard skills. Of these, soft skills include at least motor skills, joint attention, memory, control, and self-renewal.

[0031] Athletic ability is a characteristic that indicates the ability of a subject that can be grasped from the subject's physical movements, and can be analyzed from the subject's physical movements when playing a particular sport, or the movements of their hands when taking notes while studying.

[0032] Joint attention is a characteristic that indicates the degree to which a subject can empathize with another person, such as the subject's ability to sense when the other person is sad or when the other person is angry. Joint attention can be analyzed by measuring the subject's brain activity. Brain activity can be measured using any desired brain function measurement technology.

[0033] Memory is a characteristic that can be determined through the subject's learning, and specifically, memory can be analyzed from the percentage of correct answers to predetermined questions and, for example, the subject's note-taking style when studying.

[0034] Control ability is a characteristic that indicates the subject's ability to control their own behavior, and includes inhibition and switching abilities. Control ability is assessed from the supporter's record of the subject's behavior. Specifically, when the supporter prohibits the subject from engaging in a specific behavior (such as watching television), it is assessed by the subject's next behavior (such as obeying and stopping to watch television, then switching to studying).

[0035] The analysis means 13 can analyze athletic ability by taking into account the subject's physical information acquired by the physical information acquisition means 12. The subject's physical information used in this analysis can suitably include the subject's body movements and hand movements. The subject's body movements and hand movements are measured by measuring the subject's whole-body muscle activity. A wearable electromyography device (see: Japan Science and Technology Agency website, "World's First Success in Measuring Palm Muscle Activity During Pitching," https: / / www.jst.go.jp / pr / announce / 20191212 / index.html) is preferably used to measure this muscle activity. The athletic ability analysis involves the following two steps: First, muscle activity is measured in four locations: the hands, arms, legs, and feet when the subject performs a certain action. Next, athletic ability is calculated by comparing the measured muscle activity of the subject with the measurement results of another subject. More specifically, the difference from the average muscle activity is used to determine whether the muscle activity in each location is hypertonic or hypotonic. The closer the muscle activity is to the average, the higher the athletic ability is. In this case, it is desirable to analyze whether the muscle activity tends to be hypertonic or hypotonic. If the difference from the average value of the subject's muscle activity is 0, the athletic ability skill is calculated as, for example, 70. If the difference from the average value of the subject's muscle activity is positive, the athletic ability skill is calculated as, for example, "{1 - (subject's muscle activity value - average muscle activity) ÷ (maximum muscle activity - average muscle activity)} × 70", and if the difference from the average value of the subject's muscle activity is negative, the athletic ability skill is calculated as, for example, "{1 - (subject's muscle activity value - average muscle activity) ÷ (minimum muscle activity - average muscle activity)} × 70". If the calculated result contains a decimal, the skill is expressed as an integer, with the value rounded up.

[0036] Here is a specific example of how motor ability skills are assessed when performing the action of "writing in a notebook." Suppose the average muscle activity of the hands when "writing in a notebook" is 0.3mv, with a maximum of 3mv and a minimum of 0.1mv, the average muscle activity of the arms is 0.6mv, a maximum of 6mv and a minimum of 0.2mv, the average muscle activity of the legs is 0.2mv, a maximum of 2mv and a minimum of 0mv, and the average muscle activity of the feet is 0.2mv, a maximum of 2mv and a minimum of 0mv. In this case, if the subject's muscle activity when "writing in a notebook" is 2mv in the hands, 2.5mv in the arms, 1mv in the legs, and 1mv in the feet, the difference from the average muscle activity of the subject is +1.7mv for the hands, +1.9mv for the arms, +0.8mv for the legs, and +0.8mv for the feet. At this time, the athletic ability [hand] skill is calculated as 26 by "1 - 1.7 ÷ (3 - 0.3)" x 70, the athletic ability [arm] skill is calculated as 45 by "1 - 1.9 ÷ (6 - 0.6)" x 70, the athletic ability [leg] skill is calculated as 39 by "1 - 0.8 ÷ (2 - 0.2)" x 70, and the athletic ability [foot] skill is calculated as 39 by "1 - 0.8 ÷ (2 - 0.2)" x 70. Also, if the target already has an athletic ability skill value, the final athletic ability skill value will be (existing athletic ability skill value + newly calculated athletic ability skill value) ÷ 2.

[0037] The analysis means 13 can analyze joint attention by taking into account the subject's physical information acquired by the physical information acquisition means 12. The subject's brain activity can be suitably used as the subject's physical information used in this analysis. The subject's brain activity is measured using an electroencephalograph (EEG) as the physical information acquisition means 12. The analysis of joint attention involves the following three steps: First, the subject's eye movements are observed using eye tracking when a person who is within the subject's field of vision performs a certain action. If it is not confirmed that the subject's eyes have focused on that person, the joint attention is set to 0. On the other hand, if it is confirmed that the subject's eyes have focused on that person, the process moves to the next step. Next, the response of the mirror neuron system (MNS) when the subject's eyes focus on that person is observed. To observe the MNS response, for example, the μ-wave band (8-13 Hz) near the brain's sensorimotor cortex is measured using an EEG, and the change in power is used as an index of MNS activity intensity. Finally, the joint attention is calculated by comparing the change in MNS activity with the measurement results of another subject. Whether joint attention is high or low is determined from the difference from the average value of the change in MNS activity, and the larger the difference, the higher the joint attention can be determined to be.

[0038] When a person in your field of vision performs action A, if the joint attention skill is not 0, it is calculated as follows: [Formula 1]...{(Pd-Pv)÷Ps×10+50}×(Ap+0.5) Pd: The percentage of decrease in the subject's μ-wave power when the person in their field of vision performed action A Pv: The average rate at which the power of μ waves decreases when a person in your field of vision performs action A Ps: Standard deviation of the rate at which the power of μ waves decreases when a person in view performs action A Ap: The probability that the joint attention of the target is greater than 0 when the person in view performs action A.

[0039] Furthermore, skill is expressed as an integer, rounding up any decimals that may be included in the calculated result. For example, suppose that when a crying friend is within the subject's field of vision, the mean rate of decrease in μ-wave power (Pv) is 15% (0.15), the standard deviation of the rate of decrease in μ-wave power (Ps) is 0.11, and the probability that the subject's joint attention is greater than 0 (Ap) is 90% (0.9). In this case, if eye tracking determines that the subject's gaze is focused on the crying friend, the subject's joint attention is first determined to be greater than 0. Next, if the rate of decrease in the subject's μ-wave power (Pd) is measured to be 20% (0.2), the subject's joint attention skill is calculated as 76, using {(0.2 - 0.15) ÷ 0.11 × 10 + 50} × (0.9 + 0.5). Furthermore, if the subject already has a joint attention skill value, the final joint attention skill value will be (the existing joint attention skill value + the newly calculated joint attention skill value) divided by 2. Furthermore, when calculating skills, it is desirable to also analyze what actions the subject will develop high joint attention skills for.

[0040] The analysis means 13 can analyze reward characteristics by taking into account the subject's physical information acquired by the physical information acquisition means 12. Neurotransmitters and hormones secreted in the subject's brain can be suitably used as the subject's physical information to be used in this analysis. The above-mentioned dopamine, oxytocin, and serotonin can be suitably used as the neurotransmitters and hormones. Neurotransmitters and hormones can be detected by measuring the subject's brain activity. Various brain function measurement techniques can be used as the physical information acquisition means 12 to measure the subject's brain activity. The method for measuring brain activity for reward characteristic analysis involves the following three steps. First, the current waveforms obtained by the brain function measurement techniques are classified into dopamine, serotonin, and oxytocin. Brain function measurement technology for measuring current waveforms can be achieved by using machine learning analysis techniques for single-molecule measurement (Osaka University research portal site "Resou," "World's First! New Method Using Quantum Measurement and AI! Successful High-Speed ​​Detection and Identification of Neurotransmitters," July 9, 2020, https: / / resou.osaka-u.ac.jp / ja / research / 2020 / 20200709_1). Next, the subject's behavior during the measurement is determined from images captured by a camera. Specifically, the subject's behavior at the time of the neurotransmitter detection and hormone secretion measurement is determined using AI image recognition technology. A reward score is then determined based on reward tags and condition tags annotated in advance for the measured behavior. Reward tags and condition tags are described below. For example, consider the case where oxytocin is measured using brain function measurement technology when the subject performs the behavior of "looking at their mother." In this case, AI image recognition technology determines that the subject's behavior is "looking at their mother." If the verb "see" is already annotated with a reward tag with a score of 0.4, the analysis results would determine that oxytocin is 0.4, and the reward (oxytocin) score would be increased by 40, adjusting the overall reward score to 100.

[0041] The analysis means 13 can also analyze the subject's memory ability by taking into account the subject's physical information acquired by the physical information acquisition means 12. The subject's note-taking behavior during study sessions can be suitably used as the subject's physical information. The subject's note-taking behavior can be captured by a camera device serving as the physical information acquisition means 12 and recorded as image data (physical information). In particular, in this embodiment, the subject's note-taking behavior is captured as image data (physical information) by capturing an image of the subject using a camera device in a VR space. An expert, such as a memory athlete, can then assess the subject's note-taking behavior as captured image data to determine whether it facilitates memorization. The assessment data can then be used as training data to assess the subject's memory ability using AI. The subject's memory ability can also be assessed from the subject's answers to a memory test. Specifically, a memory test such as the Japanese Wechsler Memory Scale is used. Since the Japanese Wechsler Memory Scale is a test for measuring memory ability, a memory tag is attached to the test. Furthermore, the tags for each question in the test can be used to determine more detailed analysis items. For example, by assigning a long-term memory tag to questions measuring long-term memory and a short-term memory tag to questions measuring short-term memory, detailed items that can be analyzed from the questions can be determined. More specifically, memory ability can be assessed by having the subject answer memory test questions displayed in a VR space in the same space. For example, if the subject answers six questions tagged only with long-term memory, the subject's long-term memory skill in this test can be calculated using the following formula. [Formula 2]...{(Ptt-Ptv)÷Pts×10+50}×Sw Ptt: Total score of the six questions Ptv: Average of the total score of six questions Pts: Standard deviation of 6 questions Sw: Weight value of the learning level of the teaching material

[0042] Skills are expressed as integers, rounded up if the calculated result contains a decimal point. The learning level weight (Sw) of the teaching material is pre-annotated by a subject expert, such as a university professor, for the tags of the teaching material. It ranges from 0 to 1, with the higher the learning level of the teaching material. For example, if a subject's total score (Ptt) for six questions is 5, the average total score (Ptv) for the six questions is 4, the standard deviation (Pts) for the six questions is 1.41, and the learning level weight for the Japanese version of the Wechsler Memory Test is 0.9, then applying these values ​​to the above [Equation 2], the subject's long-term skill for this test is calculated as {(5 - 4) ÷ 1.41 × 10 + 50} × 0.9, which is 51. If the subject already has a long-term memory skill score, the final long-term memory skill score is calculated as (the existing long-term memory skill score + the newly calculated long-term memory skill score) ÷ 2. For example, if a subject already has a long-term memory skill of 60 and the Japanese version of the Wechsler Memory Test determines that their long-term memory skill is 51, then the subject's final long-term memory skill will be (60 + 51) ÷ 2, which is 56. This method is not limited to long-term memory skill, but also applies to short-term memory and hard skills.

[0043] The analysis means 13 can analyze cognitive characteristics according to the behavioral content of a target person that can be identified by analyzing speech and behavior information using a predetermined natural language analysis. In this case, dependency analysis, predicate-argument structure analysis, discourse structure analysis, etc. can be suitably used as the natural language analysis used.

[0044] Tags, which will be explained below, are used in the analysis of behavioral characteristics by the analysis means 13. "Tags" are used to determine the elements of behavioral characteristics necessary for a subject to perform a specific behavior. When analyzing speech and behavior information using natural language analysis, for example, when analyzing the linguistic information "watch TV" using predicate-argument structure analysis, the case relationship between the noun "television" and the predicate "watch" is obtained. In other words, in this example, the predicate-argument structure analysis determines that the noun "television" is the object of the predicate "watch."

[0045] However, while predicate-argument structure analysis alone can determine the content of a subject's behavior of "watching (or having seen) television," it is difficult to analyze the subject's behavioral characteristics in the act (content) of "watching television." Therefore, this system uses "tags" as one of the tools for analyzing the subject's behavioral characteristics.

[0046] The tags are prepared as individual tags that are assigned to each specific language and as overall tags that are assigned to the content of the subject's actions. These tags are basically prepared in advance when the analysis means 13 analyzes the behavioral characteristics.

[0047] These tags will be described using specific examples shown in Figures 3 and 4. Figures 3 and 4 are schematic diagrams showing examples of tags that are prepared in advance when analyzing behavioral characteristics by the analysis means in the individually optimized education support system according to the embodiment of the present disclosure.

[0048] Individual tags include an action tag ST1 that is assigned to a specific action, a direct object tag ST2 that is assigned to the direct object of the action, and a characteristic tag, condition tag, indirect object tag, etc. that are assigned to characteristics determined from the specific action.

[0049] With reference to Figure 3(A), a detailed explanation will be given below assuming that a supporter has entered, for example, "I watched TV" in a diary or the like. Figure 3(A) illustrates, as characteristic tags, i.e., cognitive characteristic tags in this example, a visual tag ST3, an auditory tag ST4, a taste tag ST5, a tactile tag ST6, an olfactory tag ST7, and a language tag ST8.

[0050] First, when the speech and behavior information acquisition means 11 acquires "I watched TV" as speech and behavior information, the analysis means 13 performs a predicate-argument structure analysis on the verb-predicate sentence of "I watched TV." As a result, the analysis means 13 determines that "television" is a direct object argument and "saw" is a predicate. Thereafter, the analysis means 13 assigns an action tag ST1 to the basic form of "saw," "miru," and an object tag ST2 to "television."

[0051] Furthermore, the five senses used in the behavioral content of "watching TV" are assigned a usage rate for each sense based on a known thesaurus. The analysis means 13 then assigns these usage rates as scores to the visual tag ST3, auditory tag ST4, taste tag ST5, tactile tag ST6, olfactory tag ST7, and language tag ST8, which are the characteristic tags described above. Specifically, the behavior of "watching TV" is considered a "visually dominant" behavior according to the predicate-argument structure thesaurus "Predicate Thesaurus (Takeuchi Lab)" (https: / / pth.cl.cs.okayama-u.ac.jp / testp / pth / vths) as the thesaurus, and the usage rate can be set higher than that of other senses. Here, the visual, auditory, and language scores are set to 0.5, 0.3, and 0.2, respectively, with the other senses assigned 0, resulting in a total of 1. However, the present invention is not limited to this and may be expressed as a percentage (%). The thesaurus is not limited to the predicate-argument structure thesaurus described above, but may be any other thesaurus or an unknown thesaurus that will be introduced in the future, as long as it is capable of annotating the proportion of use of the five senses in a particular action.

[0052] Furthermore, as shown in Figure 3(A), a cognitive characteristic tag TT1 is assigned as an overall tag. The cognitive characteristic tag TT1 is used to determine the overall cognitive characteristics of the subject, taking into consideration the above-mentioned behavior tag ST1, direct target tag ST2, and the proportion of use of each of the five senses assigned to each characteristic tag.

[0053] The time information acquired by the speech and behavior information acquisition means 11 is taken into consideration in determining the final cognitive characteristics. The behavior of a subject identified by the speech and behavior information acquired by the speech and behavior information acquisition means 11 is generally not limited to the above-mentioned behavior of "watching TV," and it is generally considered that other behaviors are also performed at different times. In this case, the time used for each behavior is not necessarily constant and is often considered to vary. In such a case, for example, if a visually characterized behavior is performed twice at different times, each taking one hour, while a single auditory behavior takes three hours, it would not be an accurate determination to determine that the cognitive characteristics of the subject who performed the behavior are visually dominant simply because the number of visually characterized behaviors is greater than the auditory characteristic behaviors. Therefore, multiplying the time required for each behavior (time information) can contribute to a more realistic assessment of cognitive characteristics.

[0054] Next, referring to FIG. 3(A), a detailed explanation will be given below on the assumption that the supporter has entered, for example, "I sent a letter to Grandma" through a diary or the like. In FIG. 3(A), an object tag ST9 is set to be assigned to the target of the action, and a good tag ST 10 and the condition tag ST 11 And, Joy Tag ST 12 It is set as follows.

[0055] First, when the speech and behavior information acquisition means 11 acquires "I sent a letter to Grandma" as speech and behavior information, the analysis means 13 performs a predicate-argument structure analysis on the verb-predicate sentence "I sent a letter to Grandma." As a result, the analysis means 13 determines that "letter" is the direct object argument, "sent" is the predicate, and "grandma" is the indirect object argument (the target of the letter). Thereafter, the analysis means 13 assigns an action tag ST1 to "send," the basic form of "sent," assigns a direct object tag ST2 to "letter," and assigns an indirect object tag ST9 to "grandma."

[0056] Furthermore, when the action of "sending a letter to Grandma" is judged to be a good action, the "good" tag ST 10 is judged as "True." When judging whether an action is good or bad, the system takes into account the prior annotations for the combination of action tag, direct target tag, indirect target tag, and condition tag, or the results output by a machine learning model that judges good or bad from the input of these combinations. Also, when an action is judged to be bad, the "Good" tag ST10 is judged as "False."

[0057] In this example, the condition tag ST 11 Indirect target ≠ stalking target. Condition tags are annotated in advance for pairs of behavior tags and good tags. The condition tag in this example indicates that when the indirect target is not, for example, a stalking target, the behavior of "giving a gift" is judged to be a good behavior. The condition tag specifies the conditions for the direct target and the indirect target. Note that the above condition tag ST 11 The condition "indirect target ≠ stalking target" is just one example, and is not limited to this, and different conditions can be set as appropriate for specific actions.

[0058] "Reward" Tag ST 12 is a tag to which a score is assigned that indicates the degree of reward that the recipient of the letter (in this example, "Grandma") will receive. The degree of reward that the recipient will receive is determined by considering the annotations for the combination of action tag, direct target tag, indirect target tag, condition tag, and good tag, as well as the results output by a machine learning model that scores the degree of reward obtained from the input of these combinations between 0 and 1. In this example, the "reward" tag ST 12 has been given a score of "0.3".

[0059] Furthermore, as shown in Fig. 3(a), a motivation tag TT2 is assigned as an overall tag. The motivation tag TT2 is composed of the above-mentioned action tag ST1, direct target tag ST2, indirect target tag ST9, and good tag ST 10 , condition tag ST 11 , and reward tag ST 12It is used to judge the overall motivation of the subject, including the indirect target tag. If the indirect target tag is the subject and the good tag is True, or if the target tag is someone else and the good tag is False, the motivation score (the subject) is changed. If the indirect target tag is someone else and the good tag is True, or if the target tag is the subject and the good tag is False, the motivation score (the subject) is changed. The degree of score change increases in proportion to the value of the reward tag.

[0060] With reference to Figure 3(c), a detailed description will be given below assuming that a subject inputs "I walked my dog" via SNS or the like. In Figure 3(c), abstraction tags ST 13 is exemplified.

[0061] First, when the behavior information acquisition means 11 acquires "I walked my dog" as behavior information, the analysis means 13 performs a predicate-argument structure analysis on the verb-predicate sentence "I walked my dog." As a result, the analysis means 13 determines that "dog" is a direct object argument and "I walked my dog" is a predicate. After that, the analysis means 13 assigns an action tag ST1 to the basic form "I walked my dog," and assigns a direct object tag ST2 to "dog."

[0062] Furthermore, a known thesaurus is used to determine whether the action "walked the dog" is concrete or abstract and to assign a score to it. For example, the "Japanese Vocabulary System (https: / / www.iwanami.co.jp / book / b265812.html)" is used as this thesaurus. Based on the classification of "concrete" and "abstract" in the Japanese Vocabulary System, the noun is determined to be concrete or abstract, and the score is determined based on the depth of the thesaurus. For example, the abstraction score for the "concrete" classification is a number between 0 and 0.5, and the abstraction score for the "abstract" classification is a number between 0.5 and 1, with the deeper the depth, the lower the abstraction score. Another method is to use the Japanese Abstraction Dictionary (AWD-J core Social Computing Laboratory https: / / sociocom.naist.jp / awd-j / ), which records sentences and languages ​​with abstraction scores. Using the thesaurus described above, the score obtained as the abstraction of the action "walked the dog" in this example is used as the abstraction tag ST. 13 is granted to.

[0063] The abstraction score is a numerical value ranging from 0 to 0.5. In this example, the most specific action (sentence) is assigned a numerical value of "0", and the most abstract action (sentence) is assigned a numerical value of "1". In this example, the abstraction score is assigned a value of "0.3".

[0064] Furthermore, as shown in FIG. 3(c), a determination method tag TT3 is assigned as an overall tag. The determination method tag TT3 is a tag that is generated by combining the above-mentioned action tag ST1, direct target tag ST2, and abstraction level tag ST 13 It is used to determine the subject's overall judgment method, including

[0065] By using each of the above-mentioned tags, the analysis means 13 can analyze the cognitive characteristics, motivation, and judgment methods of the subject according to the subject's behavior, which can be identified by analyzing speech and behavior information using a predetermined natural language analysis.

[0066] By using the above-mentioned tags, the analysis means 13 can analyze the control ability and behavioral characteristics when no reward is present, taking into account the results of natural language analysis. The control ability and behavioral characteristics when no reward is present as the analysis results are expressed as numerical values ​​obtained by adding or subtracting points.

[0067] The analysis means 13 can analyze the control ability according to the subject's behavior, which can be identified by analyzing the behavioral information using a predetermined natural language structure analysis. The control ability is analyzed based on the supporter's behavioral record regarding the subject. For example, a case will be explained with reference to FIG. 4(A) assuming that the supporter has used a diary or the like to enter information about the subject, such as "When I forbade him from watching TV, he got angry, but then he switched to studying." In this case, the analysis means 13 outputs the analysis results as a set of behavior tag, direct target tag, characteristic tag, and numerical value.

[0068] In this example, the behavior tag ST1 is assigned to "watching," the direct target tag ST9 is assigned to "television," the behavior set tag ST1' is assigned to the pair "watching, studying," and the direct target set tag ST9' is assigned to the pair "television, null." In addition, as a characteristic tag corresponding to control power, the inhibition power tag ST 14 and switching force tag ST 15 In this example, for the characteristic tag representing "inhibitory power," the inhibitory power required to inhibit the prohibited behavior is annotated in advance with a score α between 0 and 1 based on the information on the prohibited behavior (behavior tag and direct target tag). Also, for the characteristic tag representing "switching power," the switching power required to switch behavior is annotated in advance with a score β between 0 and 1 based on the information on the behavior (behavior tag and direct target tag) performed after switching from the prohibited behavior (behavior tag and direct target tag).

[0069] The above α is the score required to refrain from "watching TV" in this example, and the score 1-α indicates the "probability of being able to endure." The "probability of being able to endure" indicates the probability that a person will be able to refrain from an action when prohibited from doing so. This score α may be a score calculated in advance based on the results of an analysis by a specialist such as a clinical psychologist, or may be a score calculated in advance based on the results of an experiment conducted on multiple testers. In this example, the subject became "angry" when prohibited from "watching TV," so the inhibitory force tag ST 14 If the child complies with the prohibited behavior without getting angry or sulking, the score is α. If the child does not comply with the prohibited behavior because they get angry or sulky, the score is "α-0.5" if α is 0.5 or greater, and 0 if α is less than 0.5.

[0070] The above β is the score required for the behavior of "studying after being prohibited from watching TV" in this example, and the score 1-β indicates the "probability of being able to switch." The "probability of being able to switch" indicates the probability that, for example, when a person is prohibited from an action, they can switch from that action to another action. This score β may also be a score calculated in advance based on the results of analysis by the above-mentioned expert, or may be a score calculated in advance based on the results of experiments on multiple testers, for example. In this example, since the subject was able to switch from the behavior of "watching TV" to the behavior of "studying," the switching ability tag ST 15 If the behavior can be switched, the score is β, if the behavior cannot be switched and β is 0.5 or more, the score is β-0.5, and if β is less than 0.5, the score is 0.

[0071] Furthermore, as shown in FIG. 4(A), a control force tag TT4 and a control force tag TT4' are assigned as overall tags. The control force tag TT4 is a tag that includes the above-mentioned action tag ST1, direct target tag ST9, and restraint tag ST 14The control ability tag TT4' is used to determine the control ability (restraint ability) of the subject. In addition, the control ability tag TT4' is used to determine the control ability (restraint ability) of the subject. 15 It is used to assess the subject's overall control ability (switching ability), including

[0072] The analysis means 13 can analyze the behavioral characteristics when no reward is received according to the subject's behavioral content, which can be identified by analyzing the behavioral information using a predetermined natural language analysis. The behavioral characteristics when no reward is received can be analyzed from the supporter's behavioral record of the subject regarding the subject's behavior when no reward is received. As the behavioral information of the subject when analyzing the behavioral characteristics when no reward is received, for example, the subject's emotions and attitude when a certain behavior is prohibited by the supporter are preferably used. Typical examples of behavioral information (behavioral record) that indicate the subject's emotions and attitude include "angry" and "sulky." The analysis means 13 outputs the analysis results as a set of behavior tags, direct object tags, characteristic tags, and numerical values.

[0073] In this example, the behavior tag ST1 is assigned to "watching" and the direct target tag ST9 is assigned to "television." The attack tag ST 16 and Utsu Tag ST 17 In this example, if the target person is "angry," the attack tag ST 16 γ is scored for "sulky" and the depression tag ST 17For trait tags representing "attack" and "depression," the probability of not following the prohibited behavior is pre-annotated using a score between 0 and 1, γ, or ω, based on the information about the prohibited behavior (behavior tag and direct target tag). The score γ indicates the "probability of getting angry," and the score ω indicates the "probability of sulking." The "probability of getting angry" indicates the probability that a person will get angry when prohibited from performing a behavior, and the "probability of sulking" indicates the probability that a person will sulk when prohibited from performing a behavior. These scores may be pre-calculated based on the results of an analysis by a specialist such as a clinical psychologist, or may be pre-calculated based on the results of an experiment conducted on multiple testers.

[0074] Furthermore, as shown in FIG. 4(A), a no-reward behavior characteristic tag TT5 and a no-reward behavior characteristic tag TT5' are assigned as overall tags. The no-reward behavior characteristic tag TT5 is a tag that includes the above-mentioned behavior tag ST1, direct target tag ST9, and attack tag ST 16 The unrewarded behavior characteristic tag TT5' is used to determine the subject's overall unrewarded behavior characteristic (aggression), including the above-mentioned behavior tag ST1, direct target tag ST9, and depression tag ST 17 It is used to assess the subject's overall behavioral characteristics (depression) when no reward is given, including:

[0075] Among the soft skill characteristics, "self-renewal ability" refers to a characteristic that indicates the degree to which a subject's skill characteristics have changed (rate of change) after a predetermined period of time has passed since the subject began learning using a learning method recommended by the learning method recommendation means 14 described below. The rate of change in self-renewal ability can be determined, for example, from test results using indicators on a learning material basis, based on the difference between the increase or decrease in the numerical value of the skill characteristics and a predetermined threshold value. The threshold value is determined by the average increase or decrease in the numerical value of the skill characteristics when subjects belonging to the same community take the same learning material.

[0076] Skill characteristics are learning indicators obtained as a result of a subject's learning using a learning method recommended to the subject by the learning method recommendation means 14 described below, and typically include learning subjects such as "language" skills and "mathematics" skills. Hard skills can be analyzed retrospectively by analysis means. The above-mentioned tags (hereinafter referred to as "hard skill tags") are also used in the analysis of hard skill characteristics. Hard skill tags are assigned to each learning material or test question selected by learning material selection means 144 possessed by the learning method recommendation means 14 described below. Skills consist of, for example, a combination of skills acquired by studying learning materials or skills required to answer test questions, and numerical values. Skill tags can clarify the skills acquired through studying learning materials or skills required to answer test questions.

[0077] The learning method recommendation means 14 recommends a learning method according to the behavioral characteristics of the subject analyzed by the analysis means 13. The learning methods recommended by the learning method recommendation means 14 are learning methods recommended as a result of analyzing the subject's own behavior, so the subject can learn the knowledge and skills he or she needs to supplement efficiently and without taking detours, thereby achieving a high learning effect.

[0078] The learning method recommendation means 14 further includes and is configured to include a basic information acquisition means 141, a target skill calculation means 142, a learning category selection means 143, a learning material selection means 144, a teaching style selection means 145, and a goal change recommendation means 146.

[0079] The basic information acquisition means 141 acquires basic information including at least the target person's goals. "Basic information" refers to the target person's attributes such as the target person's name, gender, age, and the community to which the target person belongs. The basic information also includes the target person's behavioral characteristics (cognitive characteristics, personal characteristics, skills) as the analysis results obtained by the analysis means 13.

[0080] "Community" refers to organizations and communities such as schools (elementary and junior high school students), high schools (high school students), and universities (university students), as well as companies. Furthermore, communities also include more specialized organizations. Specifically, in the case of schools, free schools for absentee students and homes are included in communities. In the case of high schools, general education, science and mathematics, or specialized departments such as industrial and commercial departments are included in communities. In the case of companies, roles and positions within organizations, such as salaried workers, self-employed individuals, and managers, are also included in communities. Side jobs in addition to main jobs are also included in communities. Furthermore, the target person's goals are also included in basic information. For example, a goal could be "achieving an annual income of 10 million yen within two years."

[0081] The target skill calculation means 142 calculates a target skill characteristic as a predetermined numerical value required to achieve the goal acquired by the basic information acquisition means 141. Specifically, the target skill characteristic is calculated by the following formula as the difference (z) between a numerical value (x) that is the goal for achieving the goal and a numerical value (y) that calculates the current skill of the subject. [Formula 3] Target skill characteristic (z) = Achieved skill characteristic (x) - Current skill characteristic (y)

[0082] The learning category selection means 143 selects a learning category according to the goal. "Learning categories" are broadly divided into major categories such as humanities and science, medium categories such as Japanese, mathematics (arithmetic), science, social studies, and English, which are mainly representative of elementary and junior high school subjects, and minor categories included in each subject. For example, the minor learning categories included in the medium category of Japanese include reading, writing, and reading comprehension, and further subcategories include grammar, vocabulary, and idioms (reading and writing) and modern Japanese and classics (reading comprehension). In addition, the categories are not limited to the above-mentioned subjects, but also include curricula for learning specialized knowledge at universities, etc. Furthermore, categories such as curricula for working adults and categories for qualification studies are also included.

[0083] The learning material selection means 144 selects learning materials corresponding to the target skill characteristics calculated by the skill calculation means 142 from learning materials corresponding to the selected learning category. Here, in the case of the learning category of the above-mentioned class subject, for example, learning materials corresponding to the learning category include study guides and workbooks prepared for each subject. From among these, the learning material corresponding to the target skill specification is selected by the learning material selection means 144. If the numerical value of the achieved skill characteristic among the target skill characteristics calculated by the target skill calculation means 142 is high, the learning material selection means 144 selects, for example, learning materials with a high learning level from learning materials corresponding to the selected learning category. Typically, texts with subtitles such as advanced text or high-level text are cited as learning materials with a high learning level.

[0084] The learning material selection means 144 may select learning materials according to the cognitive characteristics of the subject analyzed by the analysis means 13. For example, if the cognitive characteristics of the subject analyzed by the analysis means 13 are visually dominant, video learning materials are selected. This can improve the learning effect of subjects who are not good at text learning, for example.

[0085] The instruction style selection means 145 selects an instruction method and an instructor according to the subject's personal characteristics and soft skill characteristics analyzed by the analysis means 13. Personal characteristics include the subject's motivation, judgment method, reward characteristics, and non-reward behavior characteristics, as described above. Soft skill characteristics include the subject's motor skills, joint attention, memory, control, and self-renewal, as described above. The instruction style selection means 145 selects an instruction method and an instructor according to any one of these or a combination of two or more of these analyzed by the analysis means 13. For example, if motivation (self) is higher than a threshold (motivation (others) is lower than the threshold), the instruction style selection means 145 suggests respecting the subject's own approach to problems and being conscious of structuring information and explaining it in an easy-to-understand manner so that the subject can see the path to solving the problem, and selects an instructor who can do this appropriately. This allows the subject to receive instruction from an instructor and instruction method that suits their personal characteristics, resulting in a much higher learning effect than with a uniform instruction method.

[0086] If the target's current skill characteristics exceed all of the target skill characteristics calculated by the target skill calculation means 142 that would be required if the target chooses to achieve the goal in a desired community, the goal change recommendation means 146 recommends a new goal in place of the target. For example, the following assumes that the target's goal is to "earn 8 million yen per year." Suppose the target chooses to belong to community α, and if the target's joint attention score is 50 in community α, the target can achieve the goal. However, if the target's current joint attention score is 40, the learning material selection means 144 simply selects (recommends) learning material that increases the joint attention score by 10. On the other hand, if the target chooses to belong to community β, and if the target's physical skill (hard skill) is 40 in community β, the target can achieve the goal. However, if the target's current physical skill score is 50, the target has already acquired sufficient skills, so the goal change recommendation means 146 recommends a new goal in place of the target. This allows the target person to study using materials that will improve their current skills if they are lacking in the selected community. On the other hand, if their current skills exceed their target skills in the selected community, they will be recommended to change their goals, allowing them to identify goals that match their current skills.

[0087] The analysis means 13 compares the target skill characteristics calculated by the target skill calculation means 142 before and after learning using the learning material selected by the learning material selection means 144, and analyzes whether the subject has the ability to self-update based on a comparison of the changes in the compared target skill characteristics with the changes of people who studied in the same community as the subject using the same learning material. This allows the subject to understand the rate of growth in their own target skill characteristics and their relative ability to self-update within the community to which they belong.

[0088] Furthermore, in this system, the learning content recording means 15 can record the learning content of the subject who studied based on the learning method recommended by the learning method recommendation means 14. By recording the learning content, not only the subject himself / herself but also his / her supporter can understand the learning effect of the recommended learning method. Then, the behavioral characteristic update means 16 can update the subject's behavioral characteristics according to the learning content recorded by the learning content recording means 15. In this way, the behavioral characteristics are updated according to the subject's current situation, i.e., the results of learning based on the recommended learning method, so that the learning method recommended based on the updated behavioral characteristics can further raise the subject's level.

[0089] [About the behavioral characteristics analyzed in the individually optimized educational support system disclosed herein] FIG. 5 is a graph showing the behavioral characteristics of subject X analyzed in the individually optimized education support system according to an embodiment of the present disclosure, divided into cognitive characteristics, personal characteristics, and skill characteristics. As shown in this graph, the cognitive characteristics N, personal characteristics P, and skill characteristics S of subject X analyzed by the analysis means 13 are organized into a hierarchical data structure. Furthermore, the cognitive characteristics N are weighted (quantified) by a predetermined value, such as "cognitive (100)," the personal characteristics P are weighted by "personal (400)," and the skill characteristics S are weighted by "skill (65)." This allows the learning method recommendation means 14 to accurately grasp the behavioral characteristics of subject X when determining a learning method suitable for subject X. Below, the hierarchical structure of each of the cognitive characteristics N, personal characteristics P, and skill characteristics S will be described in detail with reference to FIGS. 6 to 18.

[0090] Figure 6 is a graph showing the interrelationships between the elements (data) that make up cognitive characteristic N in Figure 5. The elements that make up cognitive characteristic N are arranged in the first layer: vision, hearing, touch, taste, smell, and language. Note that "language" here refers to the linguistic characteristics that allow the subject to recognize things, and is a different concept from "language" in hard skills, which will be discussed later.

[0091] Each of the above elements is weighted numerically. Here, a numerical value corresponding to the percentage of each element is assigned to each element, assuming that the overall cognitive characteristic is 100. Here, numerical values ​​are assigned to each element as follows: visual (40), auditory (20), tactile (10), taste (5), olfactory (5), and linguistic (20). This is, of course, merely an example. Each numerical value is calculated by multiplying the numerical value assigned to the dominant element (e.g., visual dominance) determined based on the analysis of cognitive characteristics by the analysis means 13 by the time spent performing the behavior that is the basis for that dominant element. Specifically, in the above example, the behavior of "watching television" has a usage ratio of 0.5 for visual, 0.3 for auditory, and 0.2 for linguistic. Each of these percentages is multiplied by the time spent performing the behavior of "watching television." For example, if the behavior lasts for two hours, visual is 0.5 x 2 = 1, auditory is 0.3 x 2 = 0.6, and linguistic is 0.2 x 2 = 0.4. Adding these together, we calculate the percentage of each element by dividing them by 2: visual = 1 / 2 = 0.5, auditory = 0.6 / 2 = 0.3, and language = 0.4 / 2 = 0.2. Expressing these as percentages, we get 50% for visual, 30% for auditory, and 20% for language. This is how the weighting for each element is calculated. However, since the actual activities a subject may perform in a day include activities other than "watching TV," performing the above calculation for all of the subject's activities in a day can derive a value that supports the subject's cognitive characteristics for that day's activities. Note that, to increase the accuracy of determining a subject's cognitive characteristics, behavioral information related to activities performed in a single day is usually insufficient; therefore, in practice, it is desirable to derive the above value from all of the subject's activities over at least one month.

[0092] The second layer shows the elements included in each of the visual, auditory, and linguistic senses. Visual sense includes still images and video, with numerical values ​​assigned to each element: still images (10) and video (30). In other words, the visual sense value of "40" is allocated to each element so that it is the sum of the still image and video values. While the above numerical values ​​are merely examples, for example, in the case of the aforementioned behavior of "watching television," television images themselves are composed of video, and if only the behavior of "watching television" were considered, the proportion of video would be higher. Assuming the basis for the above numerical values, for example, if three hours out of four hours were spent on the behavior of "watching television" and the remaining hour was spent on the behavior of appreciating paintings and illustrations, the values ​​would be calculated as still images (10) and video (30).

[0093] The third layer shows the elements contained in still images and videos. A still image is divided into two dimensions and three dimensions, with each element assigned a numerical value: two-dimensional (10) and three-dimensional (0). That is, the numerical value "10" of a still image is allocated to each element so that it equals the sum of the two-dimensional and three-dimensional numerical values. A video is also divided into two dimensions and three dimensions, with each element assigned a numerical value: two-dimensional (20) and three-dimensional (10). That is, the numerical value "30" of a video is allocated to each element so that it equals the sum of the two-dimensional and three-dimensional numerical values. While two-dimensional and three-dimensional images are conceivable for still images, and two-dimensional and three-dimensional videos are conceivable for videos, for example, in the above-mentioned action of "watching television," television is a two-dimensional video, and so the above elements are judged to be two-dimensional. On the other hand, in the case of an action such as "playing soccer," the above elements may be judged to be three-dimensional.

[0094] Next, the elements of hearing are assigned to sound and language, with numerical values ​​assigned to each element as sound (10) and language (10). For example, from the behavior of "listening to classical music," it can be inferred that "sound" is dominant because classical music does not contain linguistic information. On the other hand, from the behavioral information of "watching dramas on TV," it can be inferred that "language" is dominant because in dramas, it is considered normal to recognize the voices of actors (people) speaking linguistically.

[0095] Furthermore, language is divided into concrete and abstract elements, with each element assigned a numerical value: concrete (10) and abstract (10). For example, in the case of the action of "reading a paper," it can be inferred that "abstract" is dominant because papers contain many abstract concepts. In contrast, the action of "reading a manga" can be inferred that "concrete" is dominant because it also contains concrete images linked to linguistic information.

[0096] Figure 7 is a graph showing the interrelationships between the elements (data) that make up the personal traits in Figure 5. Motivation, judgment method, reward, and behavioral characteristics when no reward are present are arranged in the first layer as elements that make up personal trait P. Here, a numerical value equivalent to the percentage of each element is assigned to each element when the overall personal trait is set to 400. Here, numerical values ​​are assigned to each element as motivation (100), judgment method (100), reward (100), and behavioral characteristics when no reward is present (100), but it goes without saying that this is just an example.

[0097] The second layer shows the elements included in motivation, judgment method, reward characteristics, and behavioral characteristics when no reward is present. Motivation includes self and others, with numerical values ​​assigned to each element as self (70) and others (30). Judgment method includes abstract and concrete, with numerical values ​​assigned to each element as abstract (40) and concrete (60). Reward characteristics include love, stimulation, and happiness, with numerical values ​​assigned to each element as love (20), stimulation (30), and happiness (50). Behavioral characteristics when no reward is present include aggression and depression, with numerical values ​​assigned to each element as aggression (50) and depression (50).

[0098] Figure 8 is a graph showing the mutual relationships between the elements (data) that make up the skill characteristics in Figure 5. As elements that make up the skill characteristics S, soft skill characteristics and hard skill characteristics are arranged in the first layer, and numerical values ​​are assigned to each element as soft skill characteristics (60) and hard skill characteristics (70), respectively. Unlike the cognitive characteristics and personal characteristics described above, the numerical values ​​assigned to skill characteristics are values ​​calculated based on standard deviation.

[0099] The second layer contains the soft skill characteristics of motor skills, joint attention, memory, control, and self-renewal. Here, numerical values ​​are assigned to each element: motor skills (55), joint attention (70), memory (70), and control (50), but it goes without saying that this is just an example.

[0100] The third layer contains the working memory, short-term memory, and long-term memory components of memory, each assigned a numerical value: working memory (75), short-term memory (75), and long-term memory (60). The third layer also contains the inhibition and switching components of control, each assigned a numerical value: inhibition (45) and switching (55).

[0101] The second layer includes the hard skill characteristics: Language, Mathematics, Physics, Chemistry, Biology, Geography, History, Civics, Arts, and Information. Here, numerical values ​​are assigned to each element as Language (65), Mathematics (70), Physics (70), Chemistry (70), Biology (70), Geography (70), History (70), Civics (70), Arts (70), and Information (70), but this is only an example. The elements in the third layer and below, which are subordinate to each element in the second layer, will be explained with reference to Figures 9 to 18, which are provided for each element in the second layer.

[0102] Figure 9 is a graph showing the interrelationships between the elements (data) that make up the language component of hard skills in Figure 8. The third layer contains Japanese and English, which are included in language, and each element is assigned a numerical value, such as Japanese (65) and English (65), respectively. The fourth layer contains reading, writing, and reading comprehension, which are included in Japanese, and each element is assigned a numerical value, such as Reading and Writing (65) and Reading Comprehension (65). Furthermore, the fifth layer contains grammar, words, phrases, and kanji, which are included in reading and writing, and each element is assigned a numerical value, such as Grammar (65), Words (65), Phrases (65), and Kanji (65). Furthermore, modern and classical, which are included in reading comprehension, are included, and each element is assigned a numerical value, such as Modern (65) and Classical (65).

[0103] The fourth layer includes the English elements of grammar, vocabulary, phrases, reading, listening, writing, and speaking, with each element assigned a numerical value: Grammar (65), Vocabulary (65), Phrases (65), Reading (65), Listening (65), Writing (65), and Speaking (65).

[0104] Figure 10 is a graph showing the interrelationships between the elements (data) that make up the mathematics included in the hard skills in Figure 8. The third layer includes the mathematics components of mathematics: Foundations of Mathematics, Algebra, Geometry, Analysis, Finite and Discrete Mathematics, Mathematical Science, and Mathematics Education and History. Numerical values ​​are assigned to each component as Foundations of Mathematics (70), Algebra (70), Geometry (70), Analysis (70), Finite and Discrete Mathematics (70), Mathematical Science (70), and Mathematics Education and History (70), but needless to say, these are just examples. Note that the numerical values ​​assigned to each hard skill below are also just examples.

[0105] Figure 11 is a graph showing the interrelationships between the elements (data) that make up the physics included in the hard skills in Figure 8. The third layer contains mechanics, thermodynamics, continuum mechanics, and electromagnetism, which are included in physics, and each element is assigned a numerical value as Mechanics (70), Thermodynamics (70), Continuum Mechanics (70), and Electromagnetism (70), respectively.

[0106] Figure 12 is a graph showing the interrelationships between the elements (data) that make up the chemistry included in the hard skills in Figure 8. The third level includes physical chemistry, inorganic chemistry, organic chemistry, polymer chemistry, biochemistry, analytical chemistry, instrumental analytical chemistry, synthetic organic chemistry, applied chemistry, and environmental chemistry, which are all included in chemistry, and each element is assigned a numerical value as follows: physical chemistry (70), inorganic chemistry (70), organic chemistry (70), polymer chemistry (70), biochemistry (70), analytical chemistry, instrumental analytical chemistry, synthetic organic chemistry (70), applied chemistry (70), and environmental chemistry (70).

[0107] Figure 13 is a graph showing the mutual relationships between the elements (data) that make up the organisms included in the hard skills in Figure 8. The third layer contains physiology and ecology, which are included in organisms, and each element is assigned a numerical value as physiology (70) and ecology (70), respectively.

[0108] Figure 14 is a graph showing the interrelationships between the elements (data) that make up geography, which is included in hard skills in Figure 8. Japan and the world, which are included in geography, are placed in the third layer, with numerical values ​​assigned to each element as Japan (70) and the world (70), respectively. Furthermore, for Japan, systematic geography and geography are placed in the fourth layer, with numerical values ​​assigned to each element as systematic geography (70) and geography (70), respectively. Systematic geography is further placed in the fifth layer, with physical geography and human geography, with numerical values ​​assigned to each element as physical geography (70) and human geography (70), respectively. As with Japan, for the world, systematic geography and geography are placed in the fourth tier, with each element assigned a numerical value as systematic geography (70) and geography (70), respectively. Physical geography and human geography are placed in the fifth tier, with each element assigned a numerical value as physical geography (70) and human geography (70), respectively.

[0109] Figure 15 is a graph showing the interrelationships between the elements (data) that make up the history included in the hard skills in Figure 8. The third layer contains Japanese history and world history, both of which are included in history, and each element is assigned a numerical value, such as Japanese history (70) and world history (70), respectively.

[0110] Figure 16 is a graph showing the interrelationships between the elements (data) that make up civics, which is included in hard skills in Figure 8. The third layer contains Japan and the world, which are included in civics, with numerical values ​​assigned to each element as Japan (70) and the world (70), respectively. Furthermore, for Japan, the fourth layer contains politics, economics, and society, with numerical values ​​assigned to each element as politics (70), economics (70), and society (70), respectively. Similarly to Japan, for the world, the fourth layer contains politics, economics, and society, with numerical values ​​assigned to each element as politics (70), economics (70), and society (70), respectively.

[0111] Figure 17 is a graph showing the interrelationships between the elements (data) that make up the arts, which are included in the hard skills in Figure 8. The third layer contains the elements included in the arts, such as literature, fine arts, music, comprehensive arts, design, and others, and each element is assigned a numerical value: literature (70), fine arts (70), music (70), comprehensive arts (70), design (70), and others (70).

[0112] Figure 18 is a graph showing the interrelationships between the elements (data) that make up the information contained in the hard skills in Figure 8. The third layer contains the elements included in the information: principles, computers, design, society, and systems, with numerical values ​​assigned to each element as principles (70), computers (70), design (70), society (70), and systems (70). Furthermore, for systems, the fourth layer contains technology, institutions, and organizations, with numerical values ​​assigned to each element as technology (70), institutions (70), and organizations (70).

[0113] [Overall processing flow of the individually optimized educational support system disclosed herein] 19 is an overall flow diagram showing the flow of processing by an individual optimal education support system (hereinafter referred to as "this system") according to an embodiment of the present disclosure. This system acquires behavior information related to the behavior of a subject (step S1).

[0114] The method of acquiring the behavioral information may be via a communication network from an information and communication terminal used by the subject or their supporter, or by directly inputting the information into an operation terminal associated with the system. Furthermore, the timing of acquiring the behavioral information is not particularly limited, and the information may be acquired in real time or periodically at a predetermined timing. The acquired behavioral information may be stored in the memory unit of the system, or may be recorded on an external network server and acquired as needed. Alternatively, the contents of a diary app or messages posted on a social networking site may be acquired via API integration.

[0115] Next, the system analyzes the behavioral characteristics of the subject based on the acquired behavioral information (step S2). Finally, the system recommends a study method according to the analyzed behavioral characteristics of the subject (step S3). The detailed procedures for analyzing behavioral characteristics and recommending a study method will be described later. According to the above-described procedures in the system, the subject can be analyzed based on objective factors and a study method according to the subject's characteristics can be recommended.

[0116] [Flow of analysis process of behavioral characteristics by analysis means] FIG. 20 is a detailed flow diagram showing the specific processing flow included in "Analysis of Behavioral Characteristics" in FIG. 19. The system receives acquisition of speech and behavior information in step 1 of FIG. 19 (step S21, START), and determines whether the acquired speech and behavior information is information related to language (step S22). If the system determines that the acquired speech and behavior information is language information (step S22, YES), it then performs natural language analysis (step S23). On the other hand, if the acquired speech and behavior information is not information related to language (step S23, NO), the system proceeds to the processing of step S41 in FIG. 22.

[0117] Next, if the system determines that the acquired speech and behavior information is behavioral information (YES in step S24) through natural language syntax analysis (step S23), it analyzes motivation (step S25). On the other hand, if the system determines that the acquired speech and behavior information is not behavioral information (NO in step S24), it proceeds to the processing of step S31 in FIG.

[0118] Next, if the system determines that the subject's behavior is prohibited based on the acquired behavioral information through the motivation analysis (step S25) (step S26, YES), it performs a control ability analysis (step S27).On the other hand, if the system determines that the subject's behavior is not prohibited based on the acquired behavioral information (step S26, NO), it proceeds to step S29.

[0119] After analyzing the control ability, the system analyzes the behavioral characteristics when no reward is received (step S28), then analyzes the cognitive characteristics (step S29), and ends the process (step S30). Through this series of processes, the system completes the analysis of the subject's behavioral characteristics.

[0120] Figure 21 is a detailed flow diagram showing the processing after "NO" is selected in step 24 of Figure 20. After determining that the input speech and behavior information is not behavior information in step 24 of Figure 20 (step S24, NO), the system analyzes the determination method (step S31) and ends the processing (step S31).

[0121] FIG. 22 is a detailed flow diagram showing the processing after "NO" is selected in step 22 of FIG. 20. After determining that the input information is not linguistic information in step 22 of FIG. 19 (step S22, NO), the system determines whether the input information is physical information (step S41). If the system determines that the input information is physical information (step S41, YES), it then analyzes motor skills (step S42). On the other hand, if the system determines that the input information is not physical information (step S41, NO), it proceeds to step S61 of FIG. 23.

[0122] After analyzing motor skills (step S42), the system analyzes rewards (step S43) and determines whether the input physical information is information from communication with others (step S44). If the system determines in step S44 that the input physical information is information from communication with others (step S44, YES), it then analyzes joint attention (step S45) and terminates the process (step S46). On the other hand, if the system determines in step S44 that the input physical information is not information from communication with others (step S44, NO), it terminates the process without analyzing joint attention (step S46).

[0123] Fig. 23 is a detailed flow diagram showing the processing after "NO" is selected in step S41 of Fig. 22. If the system determines in step S41 of Fig. 21 that the input information is not physical information (step S41, NO), it then analyzes the subject's hard skills (step S61), then analyzes memory ability (step S62), and then terminates the processing (step S63).

[0124] [About the process flow for recommending learning methods by the learning method recommendation unit] FIG. 24 is a detailed flow diagram showing the specific process flow included in "recommending a learning method" in FIG. 19. The detailed flow of the learning method recommendation process by this system will be described below with reference to FIG. 24. In the learning method recommendation process, this system first acquires basic information about the target person (step S71). The basic information includes information such as the target person's name, gender, age, community to which they belong, goals, and deadlines for achieving those goals. A goal is a specific behavioral goal, such as "achieving an annual income of 10 million yen within two years." Note that if information about the target person's cognitive characteristics, personal characteristics, and skill characteristics has been accumulated through the above-mentioned analysis process, this information may also be included in the basic information.

[0125] Next, the system determines whether or not it is possible to calculate the skills required to achieve the goal (step S72). If it is possible to calculate the skills (numerical values) required to achieve the goal (step S72, YES), the system performs a calculation process for the skills required to achieve the goal (step S73). On the other hand, if it is not possible to calculate the skills required to achieve the goal (step S72, NO), the system asks the subject for missing information required for skill calculation (step S74), completes the missing information (step S75), and then returns to step S72.

[0126] After calculating the skills (target skill characteristics) required to achieve the goal in step S73, the system then asks the subject or supporter to select a community in which to achieve the goal (step S76). In this case, the system provides, for example, via a network, an interface for community selection together with a message requesting the subject to select a community to the information terminal used by the subject.

[0127] Next, the system determines whether the subject's current skill characteristics exceed all of the calculated target skill characteristics that will be required if the subject chooses to achieve their goal in the desired community, i.e., whether there is a difference between the target skill characteristics and the current skill characteristics (step S77).

[0128] If the system determines that there is a difference in step S77 (step S77, YES), it calculates the numerical value of the skill characteristic to be improved (step S78).On the other hand, if the system determines that there is no difference in step S77 (step S77, NO), it recommends a new goal instead of the above goal (step S79).

[0129] Then, the system selects a learning category that the subject should study according to the goal (step S80), and selects learning materials to be used from the learning materials corresponding to the learning category according to the calculated target skill characteristics (step S81). In this case, the system may select the learning materials to be used according to the cognitive characteristics of the subject analyzed in the analysis process.

[0130] Thereafter, the system goes through a process of selecting a teaching style (teaching method and instructor) according to the personal characteristics of the subject analyzed in the above analysis process (step S82), and ends the process (step S83).

[0131] [About the flow of analysis process of self-renewal power using analytical means] Figure 25 is a detailed flow diagram showing the processing that occurs after the completion of learning using the recommended learning method. The system compares the skill characteristics calculated in the skill calculation process (step S73 in Figure 24) before and after learning using the learning material selected in the learning material selection process (step S81) and obtains changes in the compared skill characteristics and changes in people who studied in the same community as the subject using the same learning material (step S91). The system then analyzes whether the subject has the ability to self-update based on the changes in the skill characteristics obtained through the comparison (step S92) and ends the processing (step S93).

[0132] [About radar charts] FIG. 26 is a radar chart showing an example of the skill characteristic analysis results of a subject obtained by the skill characteristic analysis process in FIGS. 20 to 23 and 25. In the radar chart shown in FIG. 26, motor skills, joint attention, inhibition, switching ability, self-renewal, and hard skills are each plotted based on a five-point evaluation. Here, it can be seen that hard skills are the highest, followed by self-renewal. By showing each skill characteristic in this way using a radar chart, it is possible to easily visually grasp the skill characteristics of the subject. In this system, numerical values ​​for the sub-elements of hard skills can also be confirmed using a radar chart.

[0133] [About line graphs] Figure 27 is a graph showing an example of changes in the subject's self-renewal ability obtained by the "analysis of self-renewal ability" process in Figure 25. In this graph, the horizontal axis represents the month, and the vertical axis represents the numerical value (%) of self-renewal ability. By plotting self-renewal ability by month on such a graph, it is possible to grasp the rate of change in the subject's self-renewal ability. In this system, changes in other soft skill characteristics and hard skill characteristics can also be grasped using line graphs in a similar manner.

[0134] [About the pie chart (cognitive characteristics)] Figure 28 is a pie chart visually showing the proportions of the five senses and language elements that make up the cognitive characteristics shown in Figure 6. By using such a pie chart to represent the cognitive characteristics of a subject, it is possible to visually grasp which of the five senses and language the subject is dominant in. Here, it can be easily understood that for this subject, vision is the most dominant.

[0135] [About the pie chart (personal characteristics)] Figure 29 is a pie chart visually showing the proportion of each of the four elements that make up the personal characteristics shown in Figure 7. As with the cognitive characteristics in Figure 28, it is easy to grasp the dominant elements for each of (A) motivation, (B) judgment method, (C) reward, and (D) behavioral characteristics when no reward is present. It is easy to grasp that (A) motivation is self-dominant, (B) judgment method is concrete-dominant, (C) reward is happiness-dominant, and (D) behavioral characteristics when no reward is present are roughly equal in aggression and depression.

[0136] As described above, according to this embodiment, it is possible to analyze a subject based on objective factors and recommend a learning method that suits the subject's characteristics. Furthermore, the individually optimized learning support system of this embodiment makes it possible to significantly improve accuracy compared to conventional systems. One indicator of this is the "individual optimization level." The individual optimization level of this system will be explained with reference to Table 1 below.

[0137] [Table 1]

[0138] As shown in Table 1, the individual optimum level for the traditional educational method of "not changing the learning content depending on the learner" is 0. The individual optimum level for the method of "changing the learning content depending on the learner's community" is 1, because the learning content takes into account the community to which the learner belongs. The individual optimum level for the method of "changing the learning content depending on the learner's community and skill characteristics" is 2. A typical example of this would be a cram school. The individual optimum level for the method of "changing the learning content depending on the learner's community and skill characteristics, and the teaching method depending on the learner's personal characteristics" is 3. A typical example of this would be a free school. Note that free schools are likely to make subjective judgments rather than use actual data. The individual optimum level for the method of "changing the learning content depending on the learner's community and skill characteristics, the teaching method depending on the learner's personal characteristics, and the teaching materials used depending on the learner's cognitive characteristics" is 4. A typical example of this would be a free school run by experts.

[0139] On the other hand, in the method of "changing the learning content in the optimal community for achieving the goals selected by the learner according to the learner's community, goals, and skill characteristics, changing the teaching method according to personal characteristics, and changing the teaching materials used according to cognitive characteristics," that is, in the individually optimal learning support system in this embodiment, this is realized by accumulating behavioral information about the subject, so the individually optimal level can be 5.

[0140] The level of automation of educational support provided by this system will be explained with reference to Tables 2 and 3 below. Here, the automation level is determined based on whether it is performed by a person or an AI, and six levels are shown, from level 0 to level 5.

[0141] [Table 2]

[0142] [Table 3]

[0143] Level 0 (no educational automation) is when education is entirely (100%) done by humans, with educators designing the curriculum and teaching independently. This is the so-called traditional educational method, and is the educational method that is still traditionally used, at least in Japan as of 2024.

[0144] Next, at Level 1 (educational support), 90% of the instructors are humans and 10% are AI, and statistical natural language processing is typically used. In this case, the educator designs the curriculum and is able to obtain the statistically most likely answer to questions that students don't know the answer to. Next, at Level 2 (partially automated education), 80% are humans and 20% are AI, and statistical natural language processing is typically used. In this case, the educator designs the curriculum and is able to obtain the optimal answer to questions that students don't know the answer to themselves.

[0145] Next, at Level 3 (conditional educational automation), where 70% of the instructors are human and 30% are AI, statistical natural language processing, image processing, audio processing, and video processing are typically used. In this case, instruction is based on a curriculum recommended by the AI, and the trainee can obtain the optimal answer to questions that they themselves do not know the answer to.

[0146] Next, at level 4 (advanced educational automation), 50% of the instructors are human and 50% are AI, and this is the level of automation that can be achieved by the individual optimal educational support system of this embodiment. In addition to the statistical machine learning of level 3, automation based on a proprietary rule base is realized. In other words, instruction is based on educational content (curriculum and educational method) recommended by the AI, and optimal answers can be obtained for questions that the student does not know the answer to by themselves.

[0147] And at level 5 (complete educational automation), 100% of the instructors are AI, and this is the ultimate level of automation that this invention aims for. While the basic technology is the same as level 4, level 5 differs from level 4 in that it uses voice synthesis and avatar generation technology. In other words, the AI ​​(avatar) provides instruction based on the educational content (curriculum and teaching method) recommended by the AI.

[0148] [Hardware configuration] Fig. 30 is a hardware configuration diagram for realizing the processing performed by the support system constituting the individualized optimal education support system according to an embodiment of the present disclosure. The support system 10 according to this embodiment is realized by a computer having the configuration shown in Fig. 30. The computer related to the support system 10 has a CPU 310, a RAM 320, a ROM 330, a HDD 340, a media interface 350, a communication interface 360, and an input / output interface 370 (see Fig. 30).

[0149] The CPU 310 operates and controls each unit based on programs stored in the ROM 330 or the HDD 340. The ROM 330 stores a boot program executed by the CPU 310 when the computer related to the support system 10 starts up, programs dependent on the hardware of the computer related to the support system 10, and the like.

[0150] The HDD 340 stores programs executed by the CPU 310, data used by such programs, etc. The media interface 350 reads programs or data stored in a recording medium (not shown) and provides the programs or data to the CPU 310 via the RAM 320. The CPU 310 loads the programs from the recording medium onto the RAM 320 via the media interface 350 and executes the loaded programs. The recording medium may be, for example, an optical recording medium such as a DVD (Digital Versatile Disk) or a PD (Phase Change Rewritable Disk), a magneto-optical recording medium such as an MO (Magneto Optical Disk), a tape medium, a magnetic recording medium, or a semiconductor memory.

[0151] The communication interface 360 ​​receives data from other devices via the network and sends it to the CPU 310, and transmits data generated by the CPU 310 to other devices via the network.

[0152] The CPU 310 controls output devices such as a display and a printer, and input devices such as a keyboard and a mouse, via the input / output interface 370. The CPU 310 acquires data from the input devices via the input / output interface 370. The CPU 310 also outputs generated data to the output devices via the input / output interface 370.

[0153] For example, the CPU 310 of the computer related to the assistance system 10 executes a program loaded onto the RAM 320 to realize the functions of each processing means (see FIG. 2) of the assistance system 10. The HDD 340 stores various data handled by each processing means (see FIG. 2). The CPU 310 of the computer reads and executes these programs from a recording medium, but as another example, these programs may be acquired from another device via a network.

[0154] The above-described embodiments have been described to facilitate understanding of the present invention, and are not intended to limit the present invention. Therefore, each element disclosed in the above embodiments is intended to include all design modifications and equivalents that fall within the technical scope of the present invention.

[0155] For example, with regard to the above-mentioned automation levels, the individualized optimal education support system according to this embodiment can achieve Level 4 (highly automated education), but can also be applied to a system capable of achieving Level 5 (fully automated education). In this embodiment, the instructor who will teach the learning method recommended by the learning method recommendation means is selected by the teaching style selection means, and this instructor is assumed to be a human instructor. Instead, adopting an avatar (AI) is considered to be able to achieve Level 5 automation. Even if the instructor selected by the system (teaching style selection means) in this embodiment is suitable for the subject, it is possible that the subject's personality may not be a good fit for the human instructor. Furthermore, since a human instructor is ultimately involved, it cannot be said that full automation has been achieved. Therefore, the emergence of an avatar as the optimal instructor for the subject is anticipated. The following processing is considered to be necessary to generate this avatar. First, a human instructor who is most compatible with the analyzed behavioral characteristics of the subject is determined. Furthermore, the optimal gender (male / female), age (e.g., from teens to 80s), face, voice quality, and speaking speed for the subject are each determined. As a result, an optimal avatar for the subject can be generated. In addition, implementing a support system using an avatar requires not only avatar output, but also advanced voice recognition technology and general-purpose AI technology that can communicate naturally with humans. Furthermore, in order to reproduce human-like qualities in the avatar of the instructor mentioned above, it is necessary to break down human-like qualities and human characteristics into more detailed terms than the analysis of the subject, and to generate an avatar that is most compatible with the results of the subject's analysis. In this case, it is desirable to break down and analyze the characteristics that represent human-like qualities, including appearance, speaking style, and listening style. [Explanation of symbols]

[0156] 10...Individually optimized educational support system 11...Means of acquiring behavioral information 12…Physical information acquisition means 13…Analysis means 14...Study method recommendation means 141…Basic information acquisition means 142...Target skill calculation method 143...Learning category selection means 144...Means for selecting teaching materials 145...Instruction method selection means 146...Goal change recommendation method 15...Means for memorizing learning content 16…Behavior characteristics update means

Claims

1. A speech and behavior information acquisition means for acquiring speech and behavior information relating to the speech and behavior of a subject; an analysis means for analyzing the behavioral characteristics of the subject based on the behavioral information acquired by the behavioral information acquisition means; a learning method recommendation means for recommending a learning method according to the behavioral characteristics of the subject analyzed by the analysis means; An individually optimized educational support system.

2. The behavioral characteristics include the subject's personal characteristics, cognitive characteristics, and skill characteristics. The individualized optimal education support system according to claim 1.

3. The personal characteristics include at least motivation, judgment method, reward characteristics, and behavioral characteristics when no reward is received, the cognitive characteristics include at least five sense characteristics and language characteristics, and the skill characteristics include soft skill characteristics and hard skill characteristics. The individualized optimal education support system according to claim 2.

4. The individualized optimal educational support system according to claim 3 , wherein the soft skill characteristics include at least motor ability, joint attention, memory, and control.

5. further comprising a physical information acquisition means for acquiring physical information of the subject; The analysis means analyzes the motor ability, the joint attention ability, the memory ability, and / or the reward characteristics in consideration of the physical information of the subject acquired by the physical information acquisition means. The individual optimal education support system according to claim 4.

6. The analysis means analyzes the cognitive characteristics, the judgment method, the motivation, the control ability, and / or the non-reward behavioral characteristics according to the subject's behavioral content that can be identified by analyzing the speech and behavior information using a predetermined natural language analysis. The individual optimal education support system according to claim 4.

7. In the analysis of the cognitive characteristics, the judgment method, and the motivation by the analysis means, an individual tag assigned to a language obtained by analyzing the speech and behavior information using the natural language analysis and an overall tag assigned to the subject's behavior content are used. The individualized optimal education support system according to claim 6.

8. The learning method recommendation means a basic information acquisition means for acquiring basic information including at least the target of the subject; a target skill calculation means for calculating a target skill characteristic required for the subject to achieve the goal acquired by the basic information acquisition means using a predetermined numerical value; a learning category selection means for selecting a learning category according to the goal; a learning material selection means for selecting learning materials according to the target skill characteristics calculated by the target skill calculation means from the learning materials corresponding to the selected learning category; 8. The individualized optimal education support system according to claim 5 or 7, comprising:

9. The learning material selection means selects the learning material according to the cognitive characteristics of the subject analyzed by the analysis means. The individualized optimal education support system according to claim 8.

10. The learning method recommendation means further includes a teaching style selection means for selecting a teaching method and / or a teacher according to the personal characteristics of the subject analyzed by the analysis means. The individualized optimal education support system according to claim 8.

11. The learning method recommendation means further includes a goal change recommendation means for recommending a new goal in place of the target goal when the current skill characteristics of the subject exceed all of the target skill characteristics calculated by the target skill calculation means that will be required when the subject chooses to achieve the goal in a desired community. The individualized optimal education support system according to claim 8.

12. The analysis means compares the target skill characteristics calculated by the target skill calculation means before and after learning using the learning material selected by the learning material selection means, and analyzes whether the subject has the ability to self-renew based on a comparison of the change in the compared target skill characteristics with a change in the target skill characteristics of a person who studied in the same community as the subject using the same learning material. The individualized optimal education support system according to claim 8.

13. further comprising a learning content recording means for recording the learning content of the subject who has studied based on the learning method recommended by the learning method recommendation means, The analysis means includes a behavioral characteristic update means for updating the behavioral characteristics of the subject in accordance with the learning content recorded by the learning content recording means. The individualized optimal education support system according to claim 8.

14. The behavioral characteristic update means updates the behavioral characteristics of the subject taking into consideration the physical information of the subject acquired by the physical information acquisition means during learning of the subject. The individualized optimal education support system according to claim 13.

15. The behavioral characteristics include the subject's ability to self-renew. The individualized optimal education support system according to claim 14.

16. a step in which a receiving means receives speech and behavior information regarding the speech and behavior of a target person via a communication network; an analysis step of analyzing the behavioral characteristics of the subject based on the received speech and behavior information; A step in which a learning method recommendation means recommends a learning method according to the analyzed behavioral characteristics of the subject. An individually optimized educational support method consisting of:

17. A process of accepting input of behavioral information regarding the behavior of the subject; A process of analyzing the behavioral characteristics of the subject based on the received speech and behavior information; A process of recommending a learning method according to the analyzed behavioral characteristics of the subject; An individually optimized educational support program that runs on a computer.

Images