Information processing method and information processing system

Abstract

The psychological state evaluation device selects one clustering model suitable for determining the psychological state of the user from among a plurality of clustering models used for determining the psychological state of the user, acquires a first biological feature amount associated with the biological information of the user in a first period, and determines the psychological state of the user in the first period on the basis of the first biological feature amount and the one clustering model.

Description

Technical Field

[0001] This disclosure relates to techniques for determining a user's mental state. Background Technology

[0002] For example, the system in Patent Document 1 uses the prediction result closest to the correct answer from multiple prediction results using multiple prediction models as the prediction result for mental state and / or emotional state.

[0003] Additionally, for example, in Patent Document 2, there are stored comfort models representing the relationship between biological information and comfort, wakefulness models representing the relationship between biological information and alertness, and fatigue models representing the relationship between biological information and fatigue. The obtained biological information is applied to the comfort models to estimate comfort, the obtained biological information is applied to the alertness models to estimate alertness, and the obtained biological information is applied to the fatigue models to estimate fatigue.

[0004] Existing technical documents Patent documents Patent Document 1: Japanese Patent Publication No. 2021-508518 Patent Document 2: Japanese Patent Application Publication No. 2013-27570 Summary of the Invention

[0005] The existing technologies described above are insufficient for accurately determining a user's psychological state and require further improvement.

[0006] This disclosure provides a technique that can accurately determine a user's psychological state.

[0007] The information processing method disclosed herein is an information processing method executed by a computer, comprising: selecting a clustering model suitable for determining the user's psychological state from a plurality of clustering models used to determine the user's psychological state; obtaining a first biometric feature associated with the user's biometric information in a first period; and determining the user's psychological state in the first period based on the first biometric feature and the clustering model.

[0008] According to this disclosure, it is possible to determine the user's psychological state with high accuracy. Attached Figure Description

[0009] Figure 1 This is a diagram showing the structure of the information processing system in Embodiment 1.

[0010] Figure 2 This is a flowchart illustrating the evaluation process of the psychological state evaluation device in Embodiment 1.

[0011] Figure 3 This is a flowchart illustrating the evaluation process of the psychological state evaluation device and server in Modification 1 of Embodiment 1.

[0012] Figure 4 This is a graph showing the creative evaluation results with and without a clustering model selected.

[0013] Figure 5 This is a graph showing the relationship between biological features input into two clustering models and Shap values.

[0014] Figure 6 This is a graph showing the correlation between the measured values ​​of divergent thinking states and the estimated values ​​of divergent thinking states based on the first clustering model.

[0015] Figure 7 This is a graph showing the correlation between the measured values ​​of divergent thinking states and the estimated values ​​of divergent thinking states based on the second clustering model.

[0016] Figure 8 This is a graph showing the evaluation results of comfort and alertness in Variation 2 of Embodiment 1, where a clustering model was selected and no clustering model was selected.

[0017] Figure 9 This is a graph representing the relationship between biological features and weights in two clustering models used to determine comfort levels.

[0018] Figure 10 This is a graph showing the relationship between biological features and weights in two clustering models used to determine alertness.

[0019] Figure 11 This is a graph showing the correlation between measured comfort values ​​and estimated comfort values ​​based on the first clustering model.

[0020] Figure 12 This is a graph showing the correlation between measured comfort values ​​and estimated comfort values ​​based on the second clustering model.

[0021] Figure 13 This is a graph showing the correlation between measured values ​​of alertness and estimated values ​​of alertness based on the third clustering model.

[0022] Figure 14 This is a graph showing the correlation between measured values ​​of alertness and estimated values ​​of alertness based on the fourth clustering model.

[0023] Figure 15 This is a diagram showing the structure of the information processing system in Embodiment 2.

[0024] Figure 16This is a diagram illustrating an example of a questionnaire survey item in Embodiment 2.

[0025] Figure 17 This is a flowchart illustrating the evaluation process of the psychological state evaluation device in Embodiment 2.

[0026] Figure 18 This is a flowchart illustrating the evaluation process of the psychological state evaluation device and server in a modified example of Embodiment 2.

[0027] Figure 19 This is a graph showing the relationship between the number of feature quantities (questionnaire survey items) and accuracy in Implementation Method 2.

[0028] Figure 20 This is a graph representing the classification accuracy of the selection model when the responses to eight questionnaire items were input.

[0029] Figure 21 This is a graph showing the top 8 questionnaire survey items with the highest Shap values ​​among multiple survey items.

[0030] Figure 22 This is a graph representing the average responses of subjects belonging to the first and second clusters to two questionnaire items.

[0031] Figure 23 This is a graph representing the average responses of subjects belonging to the first and second clusters to the four questionnaire items.

[0032] Figure 24 This is a diagram showing an example of the evaluation result screen displayed in Embodiment 1 and Embodiment 2.

[0033] Figure 25 This is a diagram showing the structure of the information processing system in Embodiment 3. Detailed Implementation

[0034] (The insights that form the basis of this disclosure) Previously, it was known that biological characteristics and psychological states, which are associated with human biological information, influence each other. However, due to individual differences in the relationship between biological characteristics and psychological states, it is difficult to infer psychological states based on biological characteristics.

[0035] Both Patent Document 1 and Patent Document 2 disclose obtaining multiple prediction results or multiple inference results from multiple models, but they do not consider selecting a suitable model from multiple models to determine the user's psychological state and using the prediction result of a single model to determine the psychological state. Therefore, it is difficult to determine the user's psychological state with high accuracy based on the user's biometric characteristics.

[0036] The following technologies are disclosed in this disclosure.

[0037] (1) The information processing method involved in one aspect of this disclosure is an information processing method executed by a computer, comprising: selecting a clustering model suitable for determining the user's psychological state from a plurality of clustering models used to determine the user's psychological state; obtaining a first biometric feature associated with the user's biometric information in a first period; and determining the user's psychological state in the first period based on the first biometric feature and the clustering model.

[0038] According to this structure, a clustering model suitable for determining a user's psychological state is selected from multiple clustering models used to determine the user's psychological state. The user's psychological state is determined based on the first biometric feature associated with the user's biometric information and the selected clustering model. Therefore, the user's psychological state can be determined with high accuracy.

[0039] (2) In the information processing method described in (1) above, it may also include: obtaining task evaluation information representing the evaluation result of a task performed by the user; and obtaining a second biometric feature associated with the biometric information of the user during a second period in which the user performs the task, wherein the selection of a clustering model includes: selecting the clustering model from the plurality of clustering models based on the correlation between the second biometric feature and the evaluation result shown by the task evaluation information.

[0040] According to this structure, by having the user perform a task in advance, a clustering model suitable for determining the user's psychological state can be selected with high precision based on the correlation between the second biometrics associated with the user's biometric information obtained during the second period of the user's task performance and the evaluation results of the task.

[0041] (3) In the information processing method described in (2) above, the task may be a task for enabling the user to engage in divergent thinking, and the psychological state may be the divergent thinking state of the user engaging in the divergent thinking.

[0042] Based on this structure, a clustering model can be selected to determine the appropriate state of divergent thinking for users by performing tasks that enable users to engage in divergent thinking.

[0043] (4) In the information processing method described in (3) above, the biological information may include information associated with at least one of brain waves, cerebral blood flow, facial images, heart rate indicators, seat pressure, skin potential and skin temperature.

[0044] According to this structure, brain waves, cerebral blood flow, facial images, heart rate indicators, seat pressure, skin potential, and skin temperature are affected by the user's psychological state. Therefore, the user's psychological state can be determined with high precision based on information associated with at least one of brain waves, cerebral blood flow, facial images, heart rate indicators, seat pressure, skin potential, and skin temperature.

[0045] (5) In the information processing method described in (3) above, the biological information may also include brain waves, facial images, or information associated with the brain waves and the facial images. The plurality of clustering models includes a first clustering model, which determines that the degree of divergent thinking is proportional to at least one of the magnitude of the activity of the γ wave, a first biometric quantity of the brainwave, and the magnitude of the facial movement, a first biometric quantity of the facial image.

[0046] Based on this structure, the psychological state of a user can be determined as follows: the greater the activity of the γ wave, which is a first biological characteristic quantity of brainwaves, and / or the greater the facial movement, which is a first biological characteristic quantity represented by facial images, the greater the degree of divergent thinking.

[0047] (6) In the information processing method described in (3) above, the biological information may also include heart rate indicators, facial images, or information associated with the heart rate indicators and the facial images. The plurality of clustering models includes a second clustering model, which determines that the degree of divergent thinking is proportional to at least one of the magnitude of the reciprocal of the value of the heart rate index, which is a first biometric quantity representing the heart rate interval, and the magnitude of the facial movement shown in the facial image, which is a first biometric quantity.

[0048] Based on this structure, the psychological state of a user can be determined as follows: the shorter the heartbeat interval (as a primary biometric) and / or the greater the facial movement (as a primary biometric) represented by the facial image, the greater the degree of divergent thinking.

[0049] (7) In the information processing method described in (1) above, it may also include: obtaining answer information containing the user's answer to the question, the selection of a clustering model includes: selecting the clustering model from the plurality of clustering models based on the answer information, and the psychological state is the divergent thinking state of the user.

[0050] Based on this structure, by having users answer questions in advance, a clustering model suitable for determining the user's psychological state can be selected with high precision based on the answer information containing the user's answers to the questions.

[0051] (8) In the information processing method described in (7) above, the question may also include at least one question item related to demographic attributes, positive emotional characteristics, behavioral characteristics, physical and mental health and psychological characteristics.

[0052] Based on this structure, users can be classified with high accuracy by answering questions related to at least one of the following: demographic attributes, positive emotional characteristics, behavioral characteristics, physical and mental health, and psychological characteristics. A clustering model suitable for determining the psychological state of users can be selected.

[0053] (9) In the information processing method described in (8) above, the questions related to the demographic attributes may include at least one of the following: questions related to the number of times the user has moved, questions related to the user's feelings toward the company, and questions related to the standardization of the user's work.

[0054] Based on this structure, by answering at least one of the following questions—related to the number of times a user has moved, related to the user's feelings toward the company, and related to the standardization of the user's work—users can be classified with higher accuracy, and a clustering model suitable for judging the user's psychological state can be selected.

[0055] (10) In the information processing method described in (8) above, the question items related to the positive emotional characteristics may include at least one of the question items related to the user's joy and the question items related to the user's pride.

[0056] Based on this structure, users can be classified with greater accuracy by answering at least one of the questions related to user joy and the questions related to user pride, and a clustering model suitable for judging the user's psychological state can be selected.

[0057] (11) The information processing method described in (8) above may also include questions related to the intention to gain closeness in the adjustment focus theory.

[0058] Based on this structure, users can be classified with greater accuracy by answering questions related to their intention to receive benefits in the moderating focus theory, and a clustering model suitable for judging the user's psychological state can be selected.

[0059] (12) In the information processing method described in (8) above, the questions related to physical and mental health may include questions related to the degree of loneliness of the user.

[0060] Based on this structure, users can be classified with higher accuracy by answering questions related to their level of loneliness, and a clustering model suitable for determining the user's psychological state can be selected.

[0061] (13) In the information processing method described in (8) above, the questions related to psychological characteristics may include questions related to the degree of self-esteem of the user.

[0062] Based on this structure, users can be classified with greater accuracy by answering questions related to their level of self-esteem, and a clustering model suitable for determining the user's psychological state can be selected.

[0063] (14) In the information processing method described in (7) above, the question may also include question items related to the user’s level of self-esteem.

[0064] Based on this structure, users can be classified with greater accuracy by answering questions related to their level of self-esteem, and a clustering model suitable for determining the user's psychological state can be selected.

[0065] (15) In the information processing method described in (2) above, the task may include appreciating content that changes at least one of the user's comfort level and alertness level, and the psychological state may include at least one of a comfort state representing the user's comfort level and an alertness state representing the user's alertness level.

[0066] Based on this structure, a clustering model can be selected that is suitable for determining at least one of the user's comfort and alertness by having the user appreciate content that changes the user's comfort and alertness.

[0067] (16) In the information processing method described in (15) above, the biological information may also include information associated with at least one of the facial image and heart rate indicators.

[0068] According to this structure, facial images and heart rate indicators are influenced by the user's comfort and wakefulness states, thus enabling the determination of at least one of the user's comfort and wakefulness states with high accuracy based on information associated with at least one of the facial images and heart rate indicators.

[0069] (17) In the information processing method described in (15) above, the psychological state may include the user's comfort state, and the biological information may include information associated with the facial image. The plurality of clustering models includes a first clustering model, which determines that the comfort level of the comfort state increases proportionally to at least one of the following: the magnitude of the rotation of the face about the vertical direction as a first biometric quantity shown in the facial image; the reciprocal of the value of the deviation representing the amount of movement of the face about the vertical direction as a first biometric quantity shown in the facial image; and the reciprocal of the value representing the deviation representing the opening of the eyes as a first biometric quantity shown in the facial image.

[0070] Based on this structure, the user's comfort state can be determined, that is, the greater the rotation of the face around the vertical direction, which is represented by the facial image as a first biological characteristic, and / or the smaller the deviation of the amount of movement of the face around the vertical direction, which is represented by the facial image as a first biological characteristic, and / or the smaller the deviation of the eye opening, which is represented by the facial image as a first biological characteristic, the greater the comfort level.

[0071] (18) In the information processing method described in (15) above, the psychological state may include the user's comfort state, the biometric information may include a facial image, a heart rate index, or information associated with the facial image and the heart rate index, and the plurality of clustering models may include a second clustering model, which determines that the comfort level of the comfort state is proportional to at least one of the following: the magnitude of the rotation of the face around the vertical direction as the first biometric quantity shown by the facial image, the magnitude of the reciprocal of the value of the deviation representing the amount of movement of the face around the vertical direction as the first biometric quantity shown by the facial image, and the magnitude of the HF component of the heart rate fluctuation as the first biometric quantity shown by the heart rate index.

[0072] Based on this structure, the user's comfort state can be determined, that is, the greater the rotation of the face around the vertical direction, which is represented by the facial image as a first biological characteristic, and / or the smaller the deviation of the amount of movement of the facial rotation around the vertical direction, which is represented by the facial image as a first biological characteristic, and / or the greater the HF component of the heart rate fluctuation, which is represented by the heart rate index as a first biological characteristic, the greater the comfort level.

[0073] (19) In the information processing method described in (15) above, the psychological state may include the user's awake state, the biological information may include a facial image, a heart rate index, or information associated with the facial image and the heart rate index, and the plurality of clustering models may include a first clustering model, which determines that at least one of the following increases proportionally: the difference between the awakeness of the awake state and the maximum and minimum values ​​of the heart rate index as the first biological characteristic quantity; the median value of the heart rate index as the first biological characteristic quantity; and the magnitude of the deviation of the eye movement as the first biological characteristic quantity shown by the facial image.

[0074] According to this structure, the greater the difference between the maximum and minimum values ​​of the heart rate, which is the first biological characteristic quantity of the heart rate index, and / or the greater the median value of the heart rate, which is the first biological characteristic quantity of the heart rate index, and / or the greater the deviation of the eye movement, which is the first biological characteristic quantity represented by the facial image, the more likely the user's awake state is to be determined.

[0075] (20) In the information processing method described in (15) above, the psychological state may include the user's awake state, the biological information may include a facial image, a heart rate index, or information associated with the facial image and the heart rate index, and the plurality of clustering models may include a second clustering model, wherein the second clustering model determines that at least one of the following increases proportionally: the awakeness of the awake state and the magnitude of the average heart rate of the heart rate index as the first biological characteristic quantity, the magnitude of the difference between the maximum and minimum values ​​of the heart rate index as the first biological characteristic quantity, and the magnitude of the vertical length of the cornea of ​​the eye shown by the facial image as the first biological characteristic quantity.

[0076] Based on this structure, the following can be used to determine the user's level of consciousness: the greater the average heart rate (a first biometric indicator), and / or the greater the difference between the maximum and minimum values ​​of the heart rate (a first biometric indicator), and / or the greater the vertical length of the cornea of ​​the eye (a first biometric indicator represented by a facial image), the greater the level of consciousness.

[0077] (21) In the information processing method described in (15) above, the task evaluation information may also represent the user's evaluation result related to the comfort level or the alertness level.

[0078] Based on this structure, a clustering model can be selected with high precision to determine the user's comfort or alertness state, based on the correlation between second biometric parameters associated with the user's biometric information obtained during a second period of the user's task and the evaluation results of tasks related to comfort or alertness. The correlation can utilize any one of Pearson's product-moment correlation coefficient, Spearman's rank correlation coefficient, Kendall's rank correlation coefficient, or an average combination of multiple correlation coefficients.

[0079] Furthermore, this disclosure can be implemented not only as an information processing method that performs the aforementioned characteristic processing, but also as an information processing system having a characteristic structure corresponding to the characteristic processing performed by the information processing method. It can also be implemented as a computer program that causes a computer to execute the characteristic processing included in such an information processing method. Therefore, in the following other embodiments, the same effect as the information processing method described above can also be achieved.

[0080] (22) Another aspect of the present invention relates to an information processing system comprising: a selection unit that selects a clustering model suitable for determining the user's psychological state from a plurality of clustering models used for determining the user's psychological state; an acquisition unit that acquires a first biometric feature associated with the user's biometric information during a first period; and a determination unit that determines the user's psychological state during the first period based on the first biometric feature and the clustering model.

[0081] (23) Another aspect of the present invention relates to an information processing method executed by a computer, which includes the following steps: obtaining task evaluation information representing the evaluation result of a task performed by a user; obtaining a first biometric feature associated with the biometric information of the user during a first period in which the user performs the task; and selecting a clustering model suitable for determining the user's psychological state from a plurality of clustering models for determining the user's psychological state based on the correlation between the first biometric feature and the evaluation result shown by the task evaluation information.

[0082] According to this structure, by having the user perform a task in advance, a clustering model suitable for determining the user's psychological state can be selected with high precision based on the correlation between the second biometrics associated with the user's biometric information obtained during the second period of the user's task performance and the evaluation results of the task.

[0083] (24) The information processing method involved in other aspects of this disclosure is an information processing method executed by a computer, including: obtaining answer information containing the user's answer to a question; and selecting, based on the answer information, a clustering model suitable for determining the user's divergent thinking state from a plurality of clustering models used to determine the user's divergent thinking state.

[0084] Based on this structure, by having users answer questions in advance, a clustering model suitable for determining the user's psychological state can be selected with high precision based on the answer information containing the user's answers to the questions.

[0085] (25) In the information processing method described in (3) above, if it is determined that the degree of divergent thinking is low, a control signal is sent to a control device that includes at least one of a lighting device that illuminates the space where the user is located and an air conditioning device that regulates the air in the space where the user is located, so as to change the output content of the control device. If it is not determined that the degree of divergent thinking is low, the control signal is not sent, and the output content of the control device is maintained.

[0086] (26) In the information processing method described in (25) above, the control device may be the lighting device, and the control signal may include a control signal that reduces the color temperature of the lighting light and reduces the illuminance of the lighting light.

[0087] (27) In the information processing method described in (25) above, it may also include: obtaining a state of alertness representing the user's level of alertness, wherein the control device is the lighting device, and when it is determined that the degree of divergent thinking is low and the state of alertness represents the user's low level of alertness, sending a control signal to increase the illuminance of the lighting light, and when it is determined that the degree of divergent thinking is low and the state of alertness represents the user's high level of alertness, sending a control signal to decrease the illuminance of the lighting light.

[0088] (28) In the information processing method described in (25) above, the control device may be the air conditioning device, and the control signal may include a control signal that increases the room temperature of the space.

[0089] (29) In the information processing method described in (15) above, if the comfort level is determined to be low in the comfort state, a control signal is sent to the control device to change the output content of the control device. The control device includes at least one of a lighting device that illuminates the space where the user is located and an air conditioning device that adjusts the air in the space where the user is located. If the comfort level is not determined to be low in the comfort state, the control signal is not sent and the output content of the control device is maintained.

[0090] (30) In the information processing method described in (29) above, the control device may be the lighting device, and the control signal may include a control signal that reduces the color temperature of the lighting light.

[0091] (31) In the information processing method described above (29), the control device may be the air conditioning device, and the control signal may include a control signal that lowers the room temperature of the space.

[0092] (32) In the information processing method described in (15) above, if the level of alertness is determined to be low in the state of wakefulness, a control signal is sent to a control device that includes at least one of a lighting device that illuminates the space where the user is located and an air conditioning device that adjusts the air in the space where the user is located, so that the output content of the control device changes. If the level of alertness is not determined to be low in the state of wakefulness, the control signal is not sent, and the output content of the control device is maintained.

[0093] (33) In the information processing method described in (29) above, the control device may be the lighting device, and the control signal may include a control signal that increases the color temperature of the lighting light.

[0094] (34) In the information processing method described in (29) above, the control device may be the air conditioning device, and the control signal may include a control signal that lowers the room temperature of the space.

[0095] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. Furthermore, the embodiments described below represent specific examples of the present disclosure. The numerical values, shapes, constituent elements, steps, and order of steps shown in the following embodiments are examples and are not intended to limit the present disclosure. Additionally, any constituent elements in the following embodiments that are not described in the independent claims representing the highest-level concept will be described as arbitrary constituent elements. Furthermore, various contents can be combined in all embodiments. In this disclosure, "at least one selected from the group consisting of A1, A2, ..., and An" can be interpreted as "A1, A2, ..., An or any combination of A1, A2, ..., An". n is an integer greater than or equal to 2. For example, at least one of A1 and A2 can be interpreted as "A1", "A2", or "A1 and A2". For example, at least one of A1, A2, and A3 can be interpreted as “A1”, “A2”, “A3”, “A1 and A2”, “A1 and A3”, “A2 and A3”, or “A1, A2, and A3”. In this disclosure, “A and / or B” can be interpreted as “A”, “B”, or “A and B”.

[0096] (Implementation Method 1) Figure 1 This is a diagram showing the structure of the information processing system 101 of Embodiment 1.

[0097] Figure 1 The information processing system 101 includes a psychological state evaluation device 1 and a measurement device 2.

[0098] The measuring device 2 measures the user's biometric information. The biometric information includes information associated with at least one of brain waves, cerebral blood flow, facial images, heart rate indicators, seat pressure, skin potential, and skin temperature.

[0099] The measuring device 2 may also include a brainwave measuring device that acquires multiple electrical signals from multiple electrodes worn on the user's head and measures brainwaves based on the acquired electrical signals. The brainwave measuring device generates and outputs information related to the user's brainwaves based on the multiple electrical signals.

[0100] Alternatively, the measuring device 2 may also include an NIRS brain measurement device that uses near-infrared spectroscopy (NIRS) to acquire signals representing the state of the user's cerebral blood flow. The NIRS brain measurement device generates a detection signal representing the temporal changes in cerebral blood flow by repeatedly illuminating the user's forehead with near-infrared light and detecting the components of the light scattered within the forehead. The NIRS brain measurement device outputs this detection signal representing the temporal changes in cerebral blood flow as information associated with the user's cerebral blood flow.

[0101] Alternatively, the measuring device 2 may also include a camera that acquires an image of the user's face by capturing the user's face. The camera outputs information related to the user's facial image.

[0102] Alternatively, the measuring device 2 may also include an electrocardiogram (ECG) waveform measuring device, which acquires multiple electrical signals generated by the user's heartbeat from multiple electrodes worn on the user's chest, and measures the ECG waveform (also known as an electrocardiogram) based on the acquired multiple electrical signals. The ECG waveform measuring device outputs the measured ECG waveform as information associated with the user's heart rate indicators.

[0103] In addition, the measuring device 2 can also illuminate the user with near-infrared light or light in the green wavelength range, and measure the heart rate based on the reflected light. The measuring device 2 may also have a light source that illuminates the opposite side of the display surface of a wearable device such as a smartwatch, and a light detection unit that detects the reflected light.

[0104] Alternatively, the measuring device 2 may also include a pressure sensor disposed on the seat of the chair where the user sits, detecting seat pressure based on the change in resistance or capacitance caused by the pressure applied by the user sitting. The pressure sensor outputs information related to the user's seat pressure.

[0105] Alternatively, the measuring device 2 may include a sensor that receives electrical signals from electrodes attached to the user's palm to measure the skin potential of the user's palm. Alternatively, the measuring device 2 may include a temperature sensor attached to the user's fingertips to measure the skin temperature of the user's fingertips. The measuring device 2 outputs information related to the user's skin potential and / or skin temperature.

[0106] The measuring device 2 transmits biological information to the psychological state evaluation device 1. The measuring device 2 and the psychological state evaluation device 1 are communicatively connected via a network, such as the Internet.

[0107] The psychological state evaluation device 1 includes, for example, a computer system. This computer system includes a control program, processing circuitry such as a processor or logic circuitry that executes the control program, and a recording device such as internal memory or accessible external memory that stores the control program. Furthermore, the psychological state evaluation device 1 can also be implemented, for example, through hardware installation based on processing circuitry, or through the execution of software programs stored in processing circuitry-based memory, distributed from an external server, or through a combination of these hardware and software installations.

[0108] The psychological state evaluation device 1 includes a data acquisition unit 11, a feature quantity calculation unit 12, a model selection and processing unit 13, a psychological state evaluation processing unit 14, and an output unit 15.

[0109] The data acquisition unit 11 acquires the user's biometric information sent by the measuring device 2. The biometric information includes information associated with at least one of brain waves, cerebral blood flow, facial images, heart rate indicators, seat pressure, skin potential, and skin temperature.

[0110] The feature calculation unit 12 acquires a first biometric feature associated with the user's biometric information during the first period. The first period is the period for determining the user's mental state. Furthermore, the feature calculation unit 12 acquires a second biometric feature associated with the user's biometric information during a second period in which the user performs a task. The second period is the period preceding the first period. During the second period, the user performs a task, and a clustering model suitable for determining the user's mental state is selected from multiple clustering models. Conversely, the user's mental state is determined during the first period.

[0111] During the first period of determining the user's psychological state, the feature calculation unit 12 calculates a first biometric feature associated with the user's biometric information. Furthermore, during the second period of the user performing a task, the feature calculation unit 12 calculates a second biometric feature associated with the user's biometric information.

[0112] The feature calculation unit 12 can also calculate a first or second biometric feature associated with the user's brainwaves. Brainwaves include alpha waves (8-13 Hz), beta waves (14-30 Hz), theta waves (4-8 Hz), delta waves (0.5-4 Hz), and gamma waves (30-50 Hz). For example, the first or second biometric feature associated with the brainwaves is the activity level of gamma waves, theta waves, delta waves, or beta waves during a first or second period. The activity level is represented by the magnitude of the median or average value of the brainwaves.

[0113] Additionally, the feature calculation unit 12 can also calculate a first biological feature or a second biological feature associated with the user's cerebral blood flow. For example, the first biological feature or second biological feature associated with cerebral blood flow is brain activity data representing the state or amount of hemoglobin in the blood within the brain during a first or second period. Brain activity data includes, for example, at least one of the concentrations of oxyhemoglobin, deoxyhemoglobin, and total hemoglobin in the cerebral blood. Hemoglobin that binds to oxygen is called oxyhemoglobin. Hemoglobin that does not bind to oxygen is called deoxyhemoglobin. The total hemoglobin concentration is the sum of the oxyhemoglobin concentration and the deoxyhemoglobin concentration. The feature calculation unit 12 can also calculate at least one of the concentrations of oxyhemoglobin, deoxyhemoglobin, and total hemoglobin in the blood within the brain during the first or second period as a first biological feature or a second biological feature.

[0114] Additionally, the feature calculation unit 12 can also calculate a first biometric feature or a second biometric feature associated with the user's facial image. The first biometric feature or the second biometric feature associated with the facial image is, for example, the magnitude of facial movement or the magnitude of eye movement. Facial movement includes movement that rotates the face around the Z-axis (vertical direction), movement that moves the face in the vertical direction (up-down direction), and movement that moves the face in the horizontal direction (left-right direction). The feature calculation unit 12 can also calculate the magnitude of the facial rotation around the Z-axis (vertical direction), the deviation (standard deviation) of the vertical (up-down) movement of the face, and the deviation (standard deviation) of the horizontal (left-right) movement of the face during the first or second period. Furthermore, eye movement includes movement that moves the eyes (pupils) in the horizontal direction (left-right direction). The eye movement can be either the left or right eye, or it can be the average of the movements of both eyes. The feature calculation unit 12 can also calculate the deviation (standard deviation) of the horizontal (left-right) movement of the eyes (pupils) during the first or second period. In addition, the deviation in the amount of eye movement can be a deviation in the amount of movement of either the left or right eye, or a deviation in the amount of movement of both eyes.

[0115] Additionally, the feature calculation unit 12 can also calculate a first or second biometric feature associated with the user's heart rate index. The electrocardiogram (ECG) waveform consists of a P wave reflecting the electrical excitation of the atria, a Q wave reflecting the electrical excitation of the ventricles, an R wave and an S wave, and a T wave reflecting the repolarization process of myocardial cells in the excited ventricles. Among these waves, the R wave has the largest wave height (i.e., potential difference) and is the most robust to noise such as myocardial potentials. The feature calculation unit 12 detects two consecutive R waves in the ECG waveform. Furthermore, the R wave detection can also use known methods such as the Pan-Tompkins method. Then, the feature calculation unit 12 determines the interval between the peak values ​​of the two R waves as the heart rate interval (RRI: RR intervals). Thus, the feature calculation unit 12 obtains the RRI in the first or second period as the first or second biometric feature. If the heart rate interval is set to t... RPI This means that the reciprocal of the heartbeat interval can also be |1 / t|. RPI |

[0116] In addition, the characteristic calculation unit 12 can calculate the heart rate, CvRR (Coefficient of Variation of RR Intervals), LF (Low Frequency) component, HF (High Frequency) component, LF+HF and RSPI (Respiration Signal Peak Interval) in the first period or the second period as the first biological characteristic quantity or the second biological characteristic quantity.

[0117] Heart rate, for example, is the number of beats per minute, calculated by dividing 60 seconds by the number of seconds in the RRI. CvRR is the coefficient of variation of heart rate fluctuations. The feature quantity calculation unit 12 calculates CvRR by standardizing the standard deviation of the RRI over any time period using the average of the RRI over any time period. The feature quantity calculation unit 12 performs frequency analysis on the time series data of the RRI using Fast Fourier Transform, and calculates the LF component and HF component based on the power spectrum obtained through the frequency analysis. The LF component is the integral value of the power spectrum in the low-frequency region of 0.04Hz to 0.14Hz, which is considered to reflect the activity of the sympathetic and parasympathetic nervous systems. The HF component is the integral value of the power spectrum in the high-frequency region of 0.14Hz to 0.4Hz, which is considered to reflect the activity of the parasympathetic nervous system. LF+HL is calculated by adding the HF component to the LF component. The RSPI respiratory cycle is extracted from the electrocardiogram waveform.

[0118] Note that the feature calculation unit 12 can estimate the heart rate based on the changes in the color of the user's skin obtained from the facial image.

[0119] Additionally, the feature calculation unit 12 can also calculate a first biological feature or a second biological feature associated with the user's seat pressure. The first biological feature or the second biological feature associated with seat pressure is the change in seat pressure during a first period or a second period.

[0120] Additionally, the feature calculation unit 12 can also calculate a first biometric feature or a second biometric feature associated with the user's skin potential. The first biometric feature or the second biometric feature associated with the skin potential is the change in skin potential during a first period or a second period.

[0121] Additionally, the feature calculation unit 12 can also calculate a first biological feature or a second biological feature associated with the user's skin temperature. The first biological feature or the second biological feature associated with skin temperature is the change in skin temperature during a first period or a second period.

[0122] The model selection processing unit 13 performs model selection processing to select a clustering model suitable for determining the user's psychological state from multiple clustering models used to determine the user's psychological state.

[0123] The model selection processing unit 13 includes a task execution unit 131, a task evaluation unit 132, a correlation calculation unit 133, a model selection unit 134, a task prompting unit 135, a task acceptance unit 136, a task database 137, an evaluation benchmark storage unit 138, and a selection model storage unit 139.

[0124] It is known that some mental tasks require convergent thinking, while others require divergent thinking. Convergent thinking is defined as thinking that arrives at an answer by logically repeating and reasoning based on existing information. For example, thinking about a question with a known correct answer, or thinking that synthesizes multiple ideas to arrive at one or more conclusions, is equivalent to convergent thinking. Convergent thinking is also known as key thinking.

[0125] Divergent thinking is the process of exploring various ideas based on known information to generate new creative concepts. Specifically, it involves thinking that produces new ideas without a definitive solution. Brainstorming and other creative thinking techniques also correspond to divergent thinking. Divergent thinking is also known as creative thinking.

[0126] Divergent thinking refers to the state in which a person engages in divergent thinking. In this context, divergent thinking can also include the state in which a person is exercising creativity.

[0127] The task database 137 pre-stores task data representing topics in tasks designed to encourage divergent thinking in users. Furthermore, the task database 137 can also store multiple task data items.

[0128] The task execution unit 131 instructs the user to perform a task. The user performs a task to provide an original answer to a given topic within a specified time. The task execution unit 131 reads the topic data stored in the task database 137 and outputs the read topic data to the task prompt unit 135.

[0129] The task prompting unit 135, such as a display device, touch panel, or speaker, acquires task data output from the task execution unit 131 and prompts the user with the task. The task prompting unit 135 can display the task via text data or output the task via sound data.

[0130] The task receiving unit 136, such as a keyboard, mouse, touch panel, or microphone, receives user input for answers to questions prompted by the task prompting unit 135. The user inputs an original answer to the question.

[0131] The task processing unit 136 can accept responses based on text data or audio data. If it accepts a response based on audio data, the task processing unit 136 converts the audio data into text data. Furthermore, the task processing unit 136 can accept either one response for a single topic or multiple responses for a single topic.

[0132] The task execution unit 131 obtains the user's answer input by the task acceptance unit 136 and outputs the obtained answer to the task evaluation unit 132.

[0133] The evaluation benchmark storage unit 138 pre-stores word vectors associated with the topic as benchmark word vectors for answers with low originality. The benchmark word vectors are word vectors in a vector space corresponding to the benchmark answer text representing the low-originality answer associated with the topic. The word vectors are calculated using methods such as Word2vec to vectorize words. The evaluation benchmark storage unit 138 stores benchmark word vectors representing low-originality answers for multiple topics.

[0134] The task evaluation unit 132 acquires task evaluation information representing the evaluation results of the task performed by the user. The task evaluation unit 132 converts the user's answer output from the task execution unit 131 into a word vector. The task evaluation unit 132 vectorizes the answer using, for example, a vectorization method such as Word2vec. The task evaluation unit 132 reads the benchmark word vector associated with the topic of the task performed by the user from the evaluation benchmark storage unit 138.

[0135] The task evaluation unit 132 evaluates the originality of the task performed by the user based on the similarity between the reference word vector read from the evaluation reference storage unit 138 and the word vector representing the user's answer received by the task acceptance unit 136. For example, the similarity is represented by the cosine similarity between the word vector representing the user's answer and the reference word vector. When the object word vector is set as 'a' and the reference word vector is set as 'b', the cosine similarity is... [Mathematical Expression 1] The cosine similarity is used to represent the originality of a response. A higher cosine similarity indicates a greater similarity between the target word vector and the baseline word vector, resulting in lower originality. Therefore, the evaluation score is defined as a higher value for a lower cosine similarity. The task evaluation unit 132 evaluates the originality of the user's task using evaluation scores ranging from 1 (lowest originality) to 5 (highest originality). The task evaluation unit 132 outputs task evaluation information representing the evaluation result of the user's task to the relevance calculation unit 133. The evaluation result is represented by an evaluation score corresponding to the level of originality of the response.

[0136] Furthermore, when multiple answers are obtained from the user, the task evaluation unit 132 can also calculate the word vector corresponding to the centroid position determined by the multiple word vectors generated by converting the multiple answers into word vectors respectively.

[0137] Furthermore, in Embodiment 1, the task evaluation unit 132 evaluates the originality of the task performed by the user based on the similarity between the reference word vector and the word vector representing the user's answer; however, this disclosure is not particularly limited to this. The mental state evaluation device 1 may also include: an answer prompting unit that prompts the user's answer to the evaluator; and an evaluation receiving unit that receives the evaluator's evaluation of the user's answer. The evaluator evaluates the originality of the prompted user's answer using a five-level evaluation score. The evaluation receiving unit receives the evaluator's evaluation score for the user's answer as input. The evaluation receiving unit outputs the evaluation result input by the evaluator to the task evaluation unit 132. Furthermore, the answer prompting unit may also prompt the user's answer to multiple evaluators, and the evaluation receiving unit may also receive multiple evaluators' input of multiple evaluation scores. In this case, the task evaluation unit 132 may also calculate the average of the multiple evaluation scores as the evaluation result of the task performed by the user.

[0138] The psychological state evaluation device 1 may also include a communication unit that sends the user's answer to the terminal used by the evaluator and receives the evaluator's evaluation result of the user's answer from the terminal used by the evaluator.

[0139] The correlation calculation unit 133 calculates the correlation between the second organism characteristic quantity obtained by the characteristic quantity calculation unit 12 and the evaluation result shown by the task evaluation information obtained by the task evaluation unit 132. The correlation calculation unit 133 calculates the correlation coefficient between the second organism characteristic quantity and the evaluation score, which is the evaluation result. For example, the Pearson product-moment correlation coefficient is used as the correlation coefficient. The Pearson product-moment correlation coefficient is calculated by dividing the covariance of the second organism characteristic quantity and the evaluation score by the value obtained by multiplying the standard deviation of the second organism characteristic quantity by the standard deviation of the evaluation score.

[0140] In addition, the second biological characteristic quantities used in calculating the correlation coefficient include, for example, the activity level of gamma waves, the activity level of theta waves, total hemoglobin concentration, deoxyhemoglobin concentration, the magnitude of facial rotation around the Z-axis (vertical direction), the deviation (standard deviation) of the vertical (up-down) movement of the face and the deviation (standard deviation) of the horizontal (left-right) movement of the face, the deviation (standard deviation) of the horizontal (left-right) movement of the eyes (pupils), the heart rate interval (RRI) time, and skin temperature.

[0141] The selection model storage unit 139 pre-stores a selection model for selecting one cluster model from multiple clustering models. The selection model is learned using, for example, the Ward method or the k-means method. The selection model classifies the input correlation coefficients into one of the multiple clusters and outputs identification information for recognizing the clustering model associated with a given cluster.

[0142] The model selection unit 134 selects a suitable clustering model from multiple clustering models used to determine the user's mental state. For example, the model selection unit 134 selects one clustering model from six clustering models. The model selection unit 134 selects a clustering model from multiple clustering models based on the correlation between a second biometric feature and the evaluation result represented by task evaluation information. The model selection unit 134 selects a clustering model from multiple clustering models by applying the correlation coefficient calculated by the correlation calculation unit 133 to the learned selection model. When the correlation coefficient is input, the selection model outputs identification information for identifying a clustering model. The model selection unit 134 outputs the identification information for identifying the selected clustering model to the mental state determination unit 141.

[0143] The psychological state evaluation processing unit 14 evaluates the user's psychological state based on the first biological feature quantity obtained by the feature quantity calculation unit 12 and a clustering model selected by the model selection processing unit 13.

[0144] The psychological state evaluation and processing unit 14 includes a psychological state determination unit 141 and a clustering model storage unit 142.

[0145] The clustering model storage unit 142 pre-stores multiple learned clustering models for determining the user's mental state. The clustering models are learned using ensemble learning models such as XGBoost (Extreme Gradient Boosting) or random forests. When a first biometric feature is input, the clustering model outputs a determination of the user's mental state. As the clustering model, regression models such as linear regression or SVM (Support Vector Machine) can be used, as well as deep learning models such as neural networks, and other machine learning methods. The mental state is a state of divergent thinking. A divergent thinking state refers to a state in which the user's thoughts are wandering and exploring. Here, a divergent thinking state can include a state in which the user is exercising creativity.

[0146] The psychological state determination unit 141 determines the user's psychological state during a first period based on a first biometric feature obtained by the feature calculation unit 12 and a clustering model selected by the model selection unit 134. The psychological state determination unit 141 reads a clustering model corresponding to the recognition information output from the model selection unit 134 from the clustering model storage unit 142. The psychological state determination unit 141 determines the user's psychological state by applying the first biometric feature to a clustering model. When the first biometric feature is input, a clustering model outputs a determination result for the user's psychological state. The determination result is represented by a determination score indicating the degree of the user's divergent thinking state. For example, the determination score is represented by values ​​from five stages, ranging from 1 point (the lowest level of divergent thinking state) to 5 points (the highest level of divergent thinking state). Furthermore, the determination score is also referred to as a creativity score.

[0147] The psychological state determination unit 141 can convert the determination score, which is represented by a score from 1 to 5, into a determination score represented by a score from 1 to 100%. The psychological state determination unit 141 outputs the determination result of the user's psychological state to the output unit 15.

[0148] The output unit 15, for example, is a display device or a touch panel, which outputs the judgment result of the user's mental state as determined by the mental state judgment unit 141. The judgment result of the user's mental state is a judgment score indicating the degree of the user's divergent thinking. The output unit 15 displays the judgment score indicating the degree of the user's divergent thinking. In addition, the output unit 15 can also send the judgment result of the user's mental state to a terminal communicatively connected to the mental state evaluation device 1. The terminal is, for example, a personal computer, a smartphone, or a tablet computer.

[0149] In Implementation 1, the mental state determination unit 141 can determine mental states other than divergent thinking states. The mental state determination unit 141 can also determine the degree of a user's convergent thinking state based on a first biometric feature obtained by the feature calculation unit 12 and a clustering model selected by the model selection unit 134. A convergent thinking state refers to a state in which the user convergently surveys and thinks. Here, a convergent thinking state can include a focused state. The mental state determination unit 141 can also determine the user's level of alertness or concentration based on the first biometric feature obtained by the feature calculation unit 12 and a clustering model selected by the model selection unit 134.

[0150] Next, the evaluation process of the psychological state evaluation device 1 in this embodiment 1 will be explained.

[0151] Figure 2 This is a flowchart illustrating the evaluation process of the psychological state evaluation device 1 in Embodiment 1.

[0152] First, in step S1, the data acquisition unit 11 acquires the user's biological information sent by the measuring device 2.

[0153] Next, in step S2, the task prompting unit 135 prompts the user with the task to be performed by the user. The task prompting unit 135 displays a task for the user to engage in divergent thinking.

[0154] Next, in step S3, the task receiving unit 136 receives the user's input of the answer to the question prompted by the task prompting unit 135.

[0155] Next, in step S4, the task evaluation unit 132 evaluates the originality of the task performed by the user and obtains the task evaluation result. The task evaluation unit 132 evaluates the originality of the task performed by the user based on the similarity between the benchmark word vector read from the evaluation benchmark storage unit 138 and the word vector representing the user's answer received by the task acceptance unit 136, and obtains an evaluation score as the task evaluation result.

[0156] Next, in step S5, the feature calculation unit 12 calculates a second biological feature associated with the biological information obtained during the second period when the user performs the task.

[0157] Next, in step S6, the correlation calculation unit 133 calculates the correlation coefficient between the second organism characteristic obtained by the characteristic calculation unit 12 and the task evaluation result obtained by the task evaluation unit 132.

[0158] Next, in step S7, the model selection unit 134 selects a clustering model from multiple clustering models by applying the correlation coefficient calculated by the correlation calculation unit 133 to the learned selection model. The model selection unit 134 inputs the correlation coefficient calculated by the correlation calculation unit 133 into the learned selection model and obtains the identification information of a clustering model output from the selection model.

[0159] Next, in step S8, the model selection unit 134 outputs the recognition information of the selected clustering model to the mental state determination unit 141.

[0160] Next, in step S9, the data acquisition unit 11 acquires the user's biological information sent by the measuring device 2.

[0161] Next, in step S10, the feature calculation unit 12 calculates a first biometric feature associated with the biometric information obtained during the first period of determining the user's mental state.

[0162] Next, in step S11, the psychological state determination unit 141 determines the user's psychological state by applying the first biometric feature calculated by the feature calculation unit 12 to a clustering model selected by the model selection unit 134. The psychological state determination unit 141 inputs the first biometric feature calculated by the feature calculation unit 12 into a clustering model selected by the model selection unit 134, and obtains a determination result of the user's psychological state output from a clustering model. The psychological state determination unit 141 determines the degree to which the user's divergent thinking is engaged.

[0163] Next, in step S12, the output unit 15 outputs the determination result of the user's mental state as determined by the mental state determination unit 141. The output unit 15 displays a determination score indicating the degree of the user's divergent thinking.

[0164] In this way, a clustering model suitable for determining a user's psychological state is selected from multiple clustering models used to determine the user's psychological state. The user's psychological state is determined based on the first biometric feature associated with the user's biometric information and the selected clustering model, thus enabling the determination of the user's psychological state with high accuracy.

[0165] Furthermore, by having users perform tasks in advance, a clustering model suitable for determining the user's psychological state can be selected with high precision based on the correlation between the second biometrics associated with the user's biometric information obtained during the second period of the user's task performance and the evaluation results of the task.

[0166] Furthermore, the model selection process represented by steps S1 to S8 and the psychological state determination process represented by steps S9 to S12 can be performed continuously in time or not. After selecting a clustering model suitable for the user, the identification information of the selected clustering model can be stored in the storage unit. In this case, model selection processing can be omitted, and the psychological state determination process can be performed using a clustering model corresponding to the stored identification information. That is, after performing model selection processing first, the psychological state determination process can be repeatedly performed using the first selected clustering model without performing model selection processing.

[0167] Alternatively, the data from the first period can be used to determine the user's psychological state after the first period ends.

[0168] Next, the evaluation processing of the psychological state evaluation device and server in Modification 1 of Embodiment 1 will be described.

[0169] In a variation of Embodiment 1, the mental state evaluation device and the server are communicatively connected via a network, such as the Internet. The mental state evaluation device according to Variation 1 of Embodiment 1 includes a data acquisition unit 11, a task execution unit 131, a task evaluation unit 132, a task prompting unit 135, a task acceptance unit 136, a task database 137, an evaluation benchmark storage unit 138, an output unit 15, and a communication unit. Furthermore, the server according to Variation 1 of Embodiment 1 includes a feature calculation unit 12, a correlation calculation unit 133, a model selection unit 134, a selection model storage unit 139, a mental state determination unit 141, a clustering model storage unit 142, and a communication unit.

[0170] The communication unit of the mental state evaluation device sends the biometric information acquired by the data acquisition unit 11 during the second period of the user's task performance and the task evaluation result evaluated by the task evaluation unit 132 to the server. Additionally, the communication unit of the mental state evaluation device sends the biometric information acquired by the data acquisition unit 11 during the first period of determining the user's mental state to the server. Furthermore, the communication unit of the mental state evaluation device receives the determination result of the user's mental state from the server.

[0171] Additionally, the server's communication unit receives biometric information acquired during the second period and task evaluation results sent by the psychological state evaluation device. Furthermore, the server's communication unit receives biometric information acquired during the first period sent by the psychological state evaluation device. Additionally, the server's communication unit sends the user's psychological state determination result, as determined by the psychological state determination unit 141, to the psychological state evaluation device.

[0172] Figure 3This is a flowchart illustrating the evaluation process of the psychological state evaluation device and server in Modification 1 of Embodiment 1.

[0173] Processing of steps S21 to S24 Figure 2 The processes in steps S1 to S4 are the same, so the explanation is omitted.

[0174] Next, in step S25, the communication unit of the psychological state evaluation device sends the biological information obtained by the data acquisition unit 11 during the second period of the user's task and the evaluation result of the task evaluated by the task evaluation unit 132 to the server.

[0175] Next, in step S31, the server's communication unit receives the biological information and task evaluation results obtained during the second period from the psychological state evaluation device.

[0176] The processing of steps S32 to S35 and step S26 Figure 2 The processes in steps S5 to S9 are the same, so the explanation is omitted.

[0177] Next, in step S27, the communication unit of the psychological state evaluation device sends the biological information obtained during the first period when the user's psychological state is determined by the data acquisition unit 11 to the server.

[0178] Next, in step S36, the server's communication unit receives the biological information obtained during the first period, sent by the psychological state evaluation device.

[0179] The processing of steps S37 and S38 Figure 2 The processes in steps S10 and S11 are the same, so the explanation is omitted.

[0180] Next, in step S39, the server's communication unit sends the determination result of the user's psychological state as determined by the psychological state determination unit 141 to the psychological state evaluation device.

[0181] Next, in step S28, the communication unit of the psychological state evaluation device receives the judgment result of the user's psychological state sent by the server.

[0182] Processing in step S29 and Figure 2 The process of step S12 is the same, so the explanation is omitted.

[0183] Furthermore, in a variation of Embodiment 1, the mental state evaluation device can also calculate a second biometric value associated with biometric information obtained during a second period of user task performance, and send the second biometric value and the task evaluation result to the server. The server can also receive the second biometric value and the task evaluation result, and calculate the correlation coefficient between the second biometric value and the task evaluation result. Alternatively, the mental state evaluation device can also calculate a first biometric value associated with biometric information obtained during a first period of user mental state determination, and send the first biometric value to the server. The server can also receive the first biometric value, apply it to a selected clustering model, and thereby determine the user's mental state.

[0184] Furthermore, in a variation of Implementation 1, the server may also send the calculated correlation coefficient to the mental state evaluation device. The mental state evaluation device may receive the correlation coefficient sent by the server and apply the received correlation coefficient to the selection model, thereby selecting a clustering model from multiple clustering models. Additionally, the mental state evaluation device may also calculate a first biometric feature associated with biometric information obtained during the first period of determining the user's mental state, and apply the calculated first biometric feature to the selected clustering model, thereby determining the user's mental state.

[0185] Next, the experimental results of using one of the multiple clustering models to evaluate divergent thinking states are explained.

[0186] Figure 4 This is a graph showing the creative evaluation results when a clustering model is selected versus when no clustering model is selected.

[0187] In the experiment, 98 participants were given tasks that encouraged divergent thinking. Additionally, the accuracy of the clustering model was evaluated using a leave-one-out cross-validation method for all 98 participants.

[0188] A test dataset is assigned to one of six clusters using a selection model learned by the Ward method, and creativity is evaluated using the clustering model corresponding to the assigned cluster. XGBoost is used to learn the clustering model for evaluating creativity. Figure 4 The figure shows the average rating of creativity when 98 subjects were not classified into 6 clusters and the average rating of creativity when 98 subjects were classified into 6 clusters.

[0189] The RMSE (Root Mean Square Error) is 0.795 when the data is not classified into 6 clusters, and 0.730 when it is classified into 6 clusters. Thus, the accuracy is improved when the data is classified into 6 clusters compared to when it is not classified into 6 clusters.

[0190] Furthermore, the average correlation coefficient per subject was 0.055 when the subjects were not classified into 6 clusters, and 0.404 when they were classified into 6 clusters. Thus, the correlation results were better when the subjects were classified into 6 clusters compared to when they were not classified into 6 clusters.

[0191] In addition, the proportion of correlation coefficients exceeding 0.4 was higher when the data was classified into 6 clusters than when it was not classified into 6 clusters.

[0192] Next, a specific example of the first biological feature input into the clustering model will be explained.

[0193] In this embodiment 1, the first biological feature quantity input into multiple clustering models can also be different according to the multiple clustering models.

[0194] Figure 5 This is a graph showing the relationship between biological features input into two clustering models and Shap (SHapley Additive Explanations) values. The Shap value represents the contribution of the biological feature to the clustering model's decision (prediction).

[0195] exist Figure 5 In the first and second clustering models, the average of the five highest-ranking biological characteristics with high Shap values ​​and the absolute values ​​of the Shap values ​​are shown. Additionally, in... Figure 5 In the degree of divergent thinking, "+" indicates that the more positive the biological characteristic quantity is, the greater the degree of divergent thinking is; "-" indicates that the more negative the biological characteristic quantity is, the greater the degree of divergent thinking is.

[0196] The activity level of brain waves (gamma waves), the magnitude of facial rotation around the Z-axis, skin temperature, total hemoglobin concentration, and deviations in the horizontal movement of the eyes contribute to the determination of the first cluster model. Additionally, the heart rate interval, deviations in the horizontal movement of the face, deoxyhemoglobin concentration, the activity level of brain waves (theta waves), and deviations in the horizontal movement of the face contribute to the determination of the second cluster model.

[0197] The first clustering model determines that the greater the activity of gamma waves (a primary biometric indicator) and / or the greater the facial movements (a primary biometric indicator represented by facial images), the greater the degree of divergent thinking. The second clustering model determines that the shorter the heart rate interval (a primary biometric indicator) and / or the greater the facial movements (a primary biometric indicator represented by facial images), the greater the degree of divergent thinking.

[0198] The mental state determination unit 141 can also change the input first biometric features based on the type of clustering model selected. For example, when a first clustering model is selected, the mental state determination unit 141 can extract the activity level of brain waves (gamma waves), the magnitude of facial rotation around the Z-axis, skin temperature, total hemoglobin concentration, and deviation of horizontal eye movement from multiple first biometric features and input them into the first clustering model. Alternatively, for example, when a second clustering model is selected, the mental state determination unit 141 can extract the heart rate interval, deviation of horizontal facial movement, deoxyhemoglobin concentration, activity level of brain waves (theta waves), and deviation of horizontal facial movement from multiple first biometric features and input them into the second clustering model.

[0199] Furthermore, the first biological feature input to the first clustering model and the second clustering model is not limited to... Figure 5 The five biological characteristics shown can also be used to input other biological characteristics.

[0200] In this way, the biological characteristics that help in the determination will vary depending on the multiple clustering models. By inputting the optimal biological characteristics into each of the multiple clustering models, the user's psychological state can be determined more accurately.

[0201] Figure 6 This is a graph showing the correlation between measured values ​​of divergent thinking states and estimated values ​​of divergent thinking states based on the first clustering model. Figure 7 This is a graph showing the correlation between the measured values ​​of divergent thinking states and the estimated values ​​of divergent thinking states based on the second clustering model.

[0202] The first clustering model is selected as the clustering model suitable for determining the psychological state of the first user, and the second clustering model is selected as the clustering model suitable for determining the psychological state of the second user.

[0203] like Figure 6 The correlation coefficient between the measured value (judgment score) of the divergent thinking state and the estimated value (judgment score) of the divergent thinking state based on the first clustering model is 0.686, indicating a high level of accuracy. Furthermore, as... Figure 7The correlation coefficient between the measured value (judgment score) of the divergent thinking state and the estimated value (judgment score) of the divergent thinking state based on the second clustering model is 0.756, indicating that high accuracy has been obtained.

[0204] Next, the evaluation process of the psychological state evaluation device in Modification 2 of Embodiment 1 will be explained.

[0205] The task of Embodiment 1 described above is a task for enabling the user to engage in divergent thinking, and the psychological state is the divergent thinking state of the user. In contrast, the task of a variation of Embodiment 1, Embodiment 2, includes appreciating content designed to change at least one of the user's comfort and alertness, and the psychological state includes at least one of a comfort state representing the user's comfort and an alertness state representing the user's alertness.

[0206] The structure of the information processing system 101 in the variation 2 of embodiment 1 is the same as that in embodiment 1 described above.

[0207] Task database 137 pre-stores content data representing changes in at least one of the user's comfort and alertness. Furthermore, task database 137 may also store multiple pieces of content data. Additionally, task database 137 stores subjective questionnaires related to at least one of comfort and alertness.

[0208] The task execution unit 131 instructs the user to perform a task. The task is to appreciate content designed to alter at least one of the user's comfort and alertness. The content may be, for example, dynamic images of scenery such as the sky, sea, river, or coastline. After appreciating the content, the user answers a subjective questionnaire related to at least one of comfort and alertness. The task execution unit 131 reads the content data and questionnaire stored in the task database 137 and outputs them to the task prompt unit 135.

[0209] For example, the questionnaire related to comfort includes seven levels: "very uncomfortable", "uncomfortable", "slightly uncomfortable", "neither", "slightly pleasant", "pleasant", and "very pleasant". Users select any one of these levels.

[0210] For example, the questionnaire related to alertness includes seven levels: "very sleepy", "sleepy", "slightly sleepy", "neither", "slightly alert", "alert", and "very alert". Users select any one of these levels.

[0211] The task prompting unit 135 obtains the content data and questionnaire survey output from the task execution unit 131. The task prompting unit 135 prompts the user with content. After prompting the user with content, the task prompting unit 135 prompts the user with a subjective questionnaire survey related to at least one of comfort and alertness.

[0212] The task acceptance unit 136 accepts the user's input of answers to the questionnaire prompted by the task prompting unit 135. After appreciating the content, the user inputs answers to a subjective questionnaire related to at least one of comfort and alertness.

[0213] The task execution unit 131 obtains the user's answer input by the task acceptance unit 136 and outputs the obtained answer to the task evaluation unit 132.

[0214] The task evaluation unit 132 acquires task evaluation information representing the evaluation results of a task performed by a user. The task evaluation information represents the user's evaluation results related to at least one of comfort and alertness. The evaluation results are represented by evaluation scores corresponding to the user's responses regarding comfort and alertness. For example, when comfort is represented using seven levels, the task evaluation unit 132 assigns evaluation scores from 1 to 7 points sequentially, starting with the lowest comfort level. Similarly, when alertness is represented using seven levels, the task evaluation unit 132 assigns evaluation scores from 1 to 7 points sequentially, starting with the lowest alertness level.

[0215] The feature calculation unit 12 acquires a first biometric feature associated with the user's biometric information during a first period. The first period is the period for determining the user's mental state. Furthermore, the feature calculation unit 12 acquires a second biometric feature associated with the user's biometric information during a second period in which the user performs a task. The mental state includes at least one of a comfort state representing the user's comfort level and a wakefulness state representing the user's alertness level. The biometric information includes information associated with at least one of a facial image and a heart rate indicator.

[0216] The feature calculation unit 12 can also calculate, based on the facial image, the average magnitude of the facial rotation around the Z-axis (vertical direction) and the deviation (standard deviation) of the amount of facial rotation around the Z-axis (vertical direction). "The amount of facial rotation around the Z-axis (vertical direction)" can also refer to the rotation angle of the face around the vertical direction. Let σ be the deviation of the amount of facial rotation around the vertical direction. neck(z) The reciprocal of the value representing the deviation of the facial movement around the vertical direction can also be |1 / σ|. neck(z) |

[0217] Additionally, the feature calculation unit 12 can also calculate the deviation (standard deviation) of eye opening during the first or second period based on the facial image. Eye opening is the length between the upper and lower eyelids. If the value representing eye opening is set to l... e Then, the reciprocal of the value indicating that the eyes are open can also be |1 / l e Furthermore, the deviation in eye opening can be the deviation in opening of either the left or right eye, or the deviation in opening of both eyes. Additionally, the feature calculation unit 12 can also calculate the deviation (standard deviation) of eye (pupil) movement (movement amount) during the first or second period based on the facial image. Furthermore, the deviation in eye movement can be the deviation in movement of either the left or right eye, or the deviation in movement of both eyes. Additionally, the feature calculation unit 12 can also calculate the average vertical length of the cornea of ​​the eyes during the first or second period based on the facial image. Furthermore, the vertical length of the cornea of ​​the eyes can be the vertical length of the cornea of ​​either the left or right eye, or the average vertical length of the cornea of ​​both eyes.

[0218] Furthermore, the feature quantity calculation unit 12 can also calculate the HF component of heart rate fluctuations during the first or second period based on the heart rate index. Additionally, the feature quantity calculation unit 12 can also calculate the difference between the maximum and minimum heart rate values ​​during the first or second period based on the heart rate index. Furthermore, the feature quantity calculation unit 12 can also calculate the median heart rate during the first or second period based on the heart rate index. Finally, the feature quantity calculation unit 12 can also calculate the average heart rate during the first or second period based on the heart rate index.

[0219] The correlation calculation unit 133 calculates the correlation between the second organism characteristic quantity obtained by the characteristic quantity calculation unit 12 and the evaluation result shown in the task evaluation information obtained by the task evaluation unit 132. The correlation calculation unit 133 calculates the correlation coefficient between the second organism characteristic quantity and the evaluation score, which is the evaluation result.

[0220] The model selection unit 134 selects a clustering model from multiple clustering models based on the correlation calculated by the correlation calculation unit 133.

[0221] The multiple clustering models include a first clustering model, which determines that the greater the rotation of the face around the vertical direction, which is a first biometric quantity represented by the facial image, and / or the smaller the deviation of the amount of movement of the face around the vertical direction, which is a first biometric quantity represented by the facial image, and / or the smaller the deviation of the eye opening, which is a first biometric quantity represented by the facial image, the greater the comfort of the comfort state.

[0222] In addition, multiple clustering models include a second clustering model, which determines that the greater the vertical rotation of the face as a first biometric quantity represented by the facial image, and / or the smaller the deviation of the vertical rotation of the face as a first biometric quantity represented by the facial image, and / or the greater the HF component of the heart rate fluctuation as a first biometric quantity represented by the heart rate index, the greater the comfort level of the comfort state.

[0223] In addition, multiple clustering models, including the third clustering model, determine the level of alertness as conscious by the greater the difference between the maximum and minimum values ​​of the heart rate (a first biological characteristic), and / or the greater the median value of the heart rate (a first biological characteristic), and / or the greater the deviation of eye movement (a first biological characteristic represented by facial images).

[0224] In addition, multiple clustering models, including the fourth clustering model, determine the level of alertness as a conscious state by the following criteria: the greater the average heart rate (as a first biometric), the greater the difference between the maximum and minimum heart rate (as a first biometric), and the greater the vertical length of the cornea of ​​the eye (as a first biometric) represented by the facial image.

[0225] The psychological state determination unit 141 determines the user's psychological state during a first period based on a first biometric feature obtained by the feature calculation unit 12 and a clustering model selected by the model selection unit 134. The psychological state determination unit 141 determines the user's psychological state by applying the first biometric feature to a clustering model. When the first biometric feature is input, a clustering model outputs a determination result for the user's psychological state. The determination result is represented by a score indicating at least one of the user's comfort level and alertness level. For example, the comfort level score is represented by values ​​from 1 (the lowest comfort level) to 7 (the highest comfort level). Similarly, the alertness level score is represented by values ​​from 1 (the lowest alertness level) to 7 (the highest alertness level).

[0226] The psychological state assessment unit 141 can convert the assessment score, which is represented by 1 to 7 points, from 1 to 100%.

[0227] Furthermore, the comfortable state and the awake state can also be represented by two stages: high or low. When the output value (judgment score) from the clustering model is equal to or greater than the threshold, the mental state determination unit 141 can determine it as a comfortable state with high comfort level and an awake state with high awakeness level. Conversely, when the output value (judgment score) from the clustering model is lower than the threshold, the mental state determination unit 141 can determine it as a comfortable state with low comfort level and an awake state with low awakeness level. When the output value is represented as 1 to 100%, the threshold is, for example, 50%; when the output value is represented by 7 stages, the threshold is, for example, 4 points.

[0228] Furthermore, in a variation 2 of this embodiment 1, after the content is viewed, the user's response to a questionnaire regarding at least one of comfort and alertness is received, but this disclosure is not particularly limited to this. The task database 137 may also store content in advance in association with evaluation results (evaluation scores). In this case, responses to the questionnaire are not required. For example, an evaluation result of "1" can be associated with content that increases comfort, an evaluation result of "-1" can be associated with content that decreases comfort, an evaluation result of "1" can be associated with content that increases alertness, and an evaluation result of "-1" can be associated with content that decreases alertness. The task evaluation unit 132 generates evaluation information representing the evaluation results corresponding to the content viewed by the user. When a questionnaire is not performed, the psychological state evaluation device 1 may not include the task receiving unit 136.

[0229] The evaluation processing of the psychological state evaluation device 1 in the variation 2 of this embodiment 1 and Figure 2 The evaluation process is the same as that of the psychological state evaluation device 1 in Embodiment 1. Furthermore, in a variation 2 of Embodiment 1, the evaluation process can also be performed using the psychological state evaluation device and server in the same manner as in Variation 1 of Embodiment 1.

[0230] The optional model storage unit 139 can store a first cluster model and a second cluster model for determining a comfortable state, and a third cluster model and a fourth cluster model for determining a conscious state. Alternatively, the optional model storage unit 139 can store the first and second cluster models for determining a comfortable state, but not the third and fourth cluster models for determining a conscious state. The optional model storage unit 139 can also store the third and fourth cluster models for determining a conscious state, but not the first and second cluster models for determining a comfortable state. The psychological state determination unit 141 can be configured to (1) perform a comfortable state determination and not perform a conscious state determination, (2) not perform a comfortable state determination and perform a conscious state determination, or (3) perform a comfortable state determination and perform a conscious state determination. That is, the psychological state determination unit 141 can determine at least one of a comfortable state and a conscious state.

[0231] Next, in a variation 2 of this embodiment 1, the experimental results of using one of the multiple clustering models to evaluate the comfort state and the awake state will be explained.

[0232] Figure 8 This is a graph showing the evaluation results of comfort and alertness in Variation 2 of Embodiment 1, where a clustering model was selected and no clustering model was selected.

[0233] In the experiment, eight participants completed a task involving appreciating content and answering subjective questionnaires on comfort and alertness. Additionally, the accuracy of the clustering model was evaluated using a leave-one-out cross-validation method for the eight participants.

[0234] A selection model learned using the Ward method was used to assign one test data point to either of two clusters. Comfort and alertness were then evaluated using the clustering model corresponding to the assigned cluster. The clustering models used to evaluate comfort and alertness were learned using Lasso (Least Absolute Shrinkage and Selection Operator). Figure 8 The figure shows the average ratings of comfort and alertness when the eight subjects were not classified into two clusters, and the average ratings of comfort and alertness when the eight subjects were classified into two clusters.

[0235] The RMSE (Root Mean Square Error) for comfort when not classified into two clusters is 3.496, while the RMSE for comfort when classified into two clusters is 2.920. Thus, the accuracy of comfort assessment is improved when classified into two clusters compared to when not classified into two clusters.

[0236] Furthermore, the RMSE for alertness was 2.181 when the data was not classified into two clusters, and 1.754 when it was classified into two clusters. Thus, the accuracy of alertness determination is improved when the data is classified into two clusters compared to when it is not classified into two clusters.

[0237] Furthermore, the average correlation coefficient for comfort per subject was 0.095 when the subjects were not classified into two clusters, and 0.402 when they were classified into two clusters. Thus, the correlation results for comfort were better when the subjects were classified into two clusters compared to when they were not classified into two clusters.

[0238] Furthermore, the average correlation coefficient for alertness per subject was 0.081 when the subjects were not classified into two clusters, and 0.385 when they were classified into two clusters. Thus, the correlation results for alertness were better when the subjects were classified into two clusters compared to when they were not classified into two clusters.

[0239] In addition, the proportion of cases where the correlation coefficient between comfort and alertness exceeded 0.4 was higher when the data were classified into two clusters than when the data were not classified into two clusters.

[0240] Next, in a variation 2 of this embodiment 1, a specific example of the first biological feature quantity input into the clustering model will be described.

[0241] In a variation of this embodiment 1, the first biological feature input to the multiple clustering models may also vary depending on the multiple clustering models.

[0242] Figure 9 This is a graph showing the relationship between biological features and their weights in two clustering models used to determine comfort levels. The weights represent the contribution of each biological feature to the clustering model's decision (prediction).

[0243] exist Figure 9 In the first and second clustering models, the average of the absolute values ​​of the three highest-weighted biological features and their weights is shown. Additionally, in... Figure 9 In the comfort rating system, "+" indicates that the more positive the biological characteristic quantity, the greater the comfort level; "-" indicates that the more negative the biological characteristic quantity, the greater the comfort level.

[0244] The magnitude of facial rotation around the Z-axis, the deviation in eye opening, and the deviation in the amount of facial rotation around the Z-axis contribute to the determination of the first clustering model. Additionally, the magnitude of facial rotation around the Z-axis, the HF component of heart rate fluctuations, and the deviation in the amount of facial rotation around the Z-axis contribute to the determination of the second clustering model.

[0245] The first clustering model determines that the greater the facial rotation around the Z-axis (represented by the facial image as a first biometric), and / or the smaller the deviation in the amount of facial rotation around the Z-axis (represented by the facial image as a first biometric), and / or the smaller the deviation in eye opening (represented by the facial image as a first biometric), the greater the comfort level of the comfort state. The second clustering model determines that the greater the facial rotation around the Z-axis (represented by the facial image as a first biometric), and / or the smaller the deviation in the amount of facial rotation around the Z-axis (represented by the facial image as a first biometric), and / or the larger the HF component of heart rate fluctuation (represented by the heart rate index as a first biometric), the greater the comfort level of the comfort state.

[0246] In both the first and second clustering models, a tendency was observed that greater facial rotation around the Z-axis, coupled with smaller deviations in the amount of facial rotation around the Z-axis, correlated with increased comfort. This is because increased comfort is more likely to induce sleep, causing the face to turn to the side. Additionally, the first clustering model showed that increased comfort resulted in smaller deviations in eye opening. This is attributed to the effect of sustained eye closure when comfort is increased. In the second clustering model, biometric parameters of the HF (High-Frequency Hip-Hop) component emerged. Higher HF components correlated with a stronger relaxing effect, thus demonstrating a tendency for higher comfort levels to correlate with higher HF components.

[0247] The mental state determination unit 141 can also change the input first biometric feature quantity according to the type of clustering model selected. For example, when the first clustering model is selected, the mental state determination unit 141 can also extract the magnitude of facial rotation around the Z-axis, the deviation of facial rotation around the Z-axis, and the deviation of eye opening from multiple first biometric features and input them into the first clustering model. Alternatively, for example, when the second clustering model is selected, the mental state determination unit 141 can also extract the magnitude of facial rotation around the Z-axis, the deviation of facial rotation around the Z-axis, and the HF component of heart rate fluctuation from multiple first biometric features and input them into the second clustering model.

[0248] Furthermore, the first biological feature input to the first clustering model and the second clustering model is not limited to... Figure 9 The three organism characteristics shown can also be used to input other organism characteristics.

[0249] Figure 10 This is a graph showing the relationship between biological features and weights in two clustering models used to determine alertness. The weights represent the contribution of each biological feature to the clustering model's decision (prediction).

[0250] exist Figure 10In the third and fourth clustering models, the average of the absolute values ​​of the weights of the top three organism features with the highest weights is shown. Additionally, in... Figure 10 In the concept of alertness, "+" indicates that the more positive the biological characteristic quantity, the greater the alertness; "-" indicates that the more negative the biological characteristic quantity, the greater the alertness.

[0251] The difference between the maximum and minimum heart rate, the median heart rate, and the deviation of eye movement help determine the third clustering model. Additionally, the average heart rate, the difference between the maximum and minimum heart rate, and the vertical length of the cornea help determine the fourth clustering model.

[0252] The third clustering model determines that the greater the difference between the maximum and minimum values ​​of heart rate (a primary biometric), and / or the greater the median heart rate (a primary biometric), and / or the greater the deviation in eye movement (a primary biometric) represented by facial images, the greater the level of alertness in the conscious state. The fourth clustering model determines that the greater the average heart rate (a primary biometric), and / or the greater the difference between the maximum and minimum heart rate (a primary biometric), and / or the greater the vertical length of the cornea (a primary biometric) represented by facial images, the greater the level of alertness in the conscious state.

[0253] In the third and fourth cluster models, a tendency was observed that increased heart rate correlated with increased alertness. In the third cluster model, a bias in eye movement was observed, with increased alertness leading to more eye movement. In the fourth cluster model, the vertical length of the cornea was considered, and a tendency was observed that increased alertness resulted in more open eyes.

[0254] The mental state determination unit 141 can also change the input first biological feature quantity according to the type of clustering model selected. For example, when the third clustering model is selected, the mental state determination unit 141 can also extract the difference between the maximum and minimum heart rate, the median heart rate, and the deviation of eye movement from multiple first biological feature quantities and input them into the third clustering model. In addition, for example, when the fourth clustering model is selected, the mental state determination unit 141 can also extract the average heart rate, the difference between the maximum and minimum heart rate, and the vertical length of the cornea of ​​the eye from multiple first biological feature quantities and input them into the fourth clustering model.

[0255] Furthermore, the first biological feature input into the third and fourth clustering models is not limited to... Figure 10 The three organism characteristics shown can also be used to input other organism characteristics.

[0256] Figure 11 This is a graph showing the correlation between measured comfort values ​​and estimated comfort values ​​based on the first clustering model. Figure 12 This is a graph showing the correlation between measured comfort values ​​and estimated comfort values ​​based on the second clustering model.

[0257] The first clustering model is selected as the clustering model suitable for determining the comfort state of the first user, and the second clustering model is selected as the clustering model suitable for determining the comfort state of the second user.

[0258] like Figure 11 As shown, the correlation coefficient between the measured comfort values ​​(comfort score) and the estimated comfort values ​​(comfort score) from the first clustering model is 0.750, indicating high accuracy. Furthermore, as... Figure 12 As shown, the correlation coefficient between the measured comfort value (comfort judgment score) and the estimated comfort value (comfort judgment score) of the second clustering model is 0.641, which indicates that high accuracy has been achieved.

[0259] Figure 13 This is a graph showing the correlation between measured values ​​of alertness and estimated values ​​of alertness based on the third clustering model. Figure 14 This is a graph showing the correlation between measured values ​​of alertness and estimated values ​​of alertness based on the fourth clustering model.

[0260] The third clustering model was selected as the clustering model suitable for determining the conscious state of the first user, and the fourth clustering model was selected as the clustering model suitable for determining the conscious state of the second user.

[0261] like Figure 13 As shown, the correlation coefficient between the measured value of alertness (the alertness judgment score) and the estimated value of alertness (the alertness judgment score) based on the third clustering model is 0.773, indicating high accuracy. Furthermore, as... Figure 14 As shown, the correlation coefficient between the measured value of alertness (the score for judging alertness) and the estimated value of alertness (the score for judging alertness) based on the fourth clustering model is 0.506, indicating that high accuracy has been obtained.

[0262] (Implementation Method 2) In Implementation 1, task evaluation information representing the evaluation result of a task performed by a user is obtained, a second biometric feature associated with the user's biometric information during a second period of task performance is obtained, and a clustering model is selected from multiple clustering models based on the correlation between the second biometric feature and the evaluation result. In contrast, in Implementation 2, response information containing the user's answers to a questionnaire (question) is obtained, and a clustering model is selected from multiple clustering models based on the response information.

[0263] Figure 15 This is a diagram showing the structure of the information processing system 102 of Embodiment 2.

[0264] Figure 15 The information processing system 102 shown includes a psychological state evaluation device 1A and a measurement device 2. Furthermore, in this embodiment 2, structures identical to those in embodiment 1 are labeled with the same reference numerals, and descriptions are omitted.

[0265] The psychological state evaluation device 1A includes, for example, a computer system. This computer system includes a control program, a processing circuit such as a processor or logic circuit that executes the control program, and a recording device such as internal memory or accessible external memory that stores the control program. It should be noted that the psychological state evaluation device 1A can be implemented, for example, through hardware installation based on the processing circuit, or through the execution of a software program based on the processing circuit that is stored in memory or distributed from an external server, or through a combination of these hardware and software installations.

[0266] The psychological state evaluation device 1A includes a data acquisition unit 11, a feature quantity calculation unit 12A, a model selection and processing unit 13A, a psychological state evaluation processing unit 14, and an output unit 15.

[0267] The feature calculation unit 12A acquires a first biometric feature associated with the user's biometric information during the first period. The first period is the period for determining the user's mental state. The first biometric feature in Embodiment 2 is the same as the first biometric feature in Embodiment 1. During the first period for determining the user's mental state, the feature calculation unit 12A calculates the first biometric feature associated with the user's biometric information.

[0268] The model selection and processing unit 13A includes a questionnaire survey database 151, a questionnaire survey processing unit 152, a questionnaire survey prompting unit 153, a questionnaire survey acceptance unit 154, a model selection storage unit 155, and a model selection unit 156.

[0269] The questionnaire database 151 pre-stores questionnaire data that represents questionnaires related to users' personal attributes.

[0270] The questionnaire processing unit 152 prompts users to answer questionnaires related to their personal attributes. The questionnaire processing unit 152 reads the questionnaire data stored in the questionnaire database 151 and outputs the read questionnaire data to the questionnaire prompt unit 153.

[0271] The questionnaire prompt unit 153, such as a display device or touch panel, acquires the questionnaire data output from the questionnaire processing unit 152 and prompts the user to complete the questionnaire (ask questions). The questionnaire prompt unit 153 displays multiple questionnaires and displays answer fields for answering the multiple questionnaires.

[0272] The questionnaire processing unit 154, which may be a keyboard, mouse, touch panel, or microphone, accepts user input for responses to questionnaire (questions) prompted by the questionnaire prompt unit 153. Users input responses to the questionnaire (questions). Responses may, for example, be based on visual analog ratings.

[0273] In addition, questionnaires can also be recorded on paper, with users using writing instruments to record their answers. In this case, the questionnaire processing unit 154 may also record answers from third parties other than the responding user, for any reason.

[0274] The questionnaire processing unit 152 obtains response information, which includes the user's answers to a questionnaire (question) related to the user's personal attributes input by the questionnaire receiving unit 154. The questionnaire processing unit 152 outputs the obtained response information to the model selection unit 156. Furthermore, the user's responses are represented numerically.

[0275] Figure 16 This is a diagram illustrating an example of a questionnaire survey item in Embodiment 2.

[0276] For example, a questionnaire (question) may include questionnaire items (question items) related to demographic attributes, positive emotional characteristics, behavioral characteristics, physical and mental health, and psychological characteristics. In addition, a category of questionnaire items may also contain multiple detailed questionnaire items that ask more detailed questions.

[0277] Questionnaires related to demographic attributes include at least one item related to the frequency of user relocation, the user's emotional attachment to the company, and the degree of standardization of the user's work. Questionnaires related to positive emotional characteristics include at least one item related to user joy and user pride. Questionnaires related to behavioral characteristics include items related to benefit proximity intention (mediation focus - benefit proximity) in moderating focus theory. Questionnaires related to mental and physical health include items related to the degree of user loneliness. Questionnaires related to psychological characteristics include items related to the degree of user self-esteem.

[0278] Furthermore, when a survey includes multiple detailed survey items, the sum of the answers to each of the detailed survey items can be considered as one answer to a single survey item. Alternatively, the average of the multiple answers can be used as one answer to a single survey item. Additionally, when a survey includes multiple detailed survey items, users can answer all of the detailed survey items, or they can answer only a portion of the detailed survey items.

[0279] For the questionnaire about the number of times a user has moved, the user answered how many times they have moved.

[0280] In addition, for the questionnaire about the level of affection users have for the company, users responded on a seven-stage scale. For example, for the question "Do you have affection for the company?", users selected one of the following answers: "No affection at all", "No affection", "Not much affection", "Neither", "Slight affection", "Affectionate", and "Very affectionate".

[0281] In addition, for the questionnaire on the standardization of work, users responded with a seven-stage scale.

[0282] Additionally, for questionnaires related to joy and pride, users responded with a seven-stage answer to determine if they met the criteria. The Japanese version of the DPES (Dispositional Positive Emotion Scales) can also be used in the questionnaires.

[0283] For example, questionnaires related to joy include items such as "My life has always been good," "Many events in my daily life make me happy," and "Good things always happen to me." Additionally, questionnaires related to pride include items such as "I feel good about myself," "People often recognize my worth," and "I feel proud of myself or my achievements." Users respond in seven stages to determine whether they meet the criteria for each of the various questionnaire items.

[0284] In addition, the Japanese version of DPES is described in the following document.

[0285] (Reference) Daichi Sugawara, Koki Arimitsu, Masaru Sugie, “An Attempt to Produce a Japanese Version of DPES”, Journal of Health Psychology Research 2020, Vol.33, No.1, 57-65.

[0286] In addition, for the questionnaire on the level of loneliness, users responded on a seven-stage scale. For example, the questionnaire included multiple items such as "I feel alienated at work" and "I feel isolated by people at work." Users responded on a seven-stage scale to indicate whether they met the criteria for each of the multiple questionnaire items.

[0287] Additionally, for questionnaire items related to the benefit proximity intention in moderating focus theory, users responded in seven stages to determine if they met the criteria. For example, questionnaire items related to benefit proximity intention in moderating focus theory included multiple items such as "I often imagine how I can achieve my goals and hopes," "I generally focus my awareness on achieving good results in life," and "I often think about how to achieve good results." Users responded in seven stages to determine if they met each of the multiple questionnaire items.

[0288] In addition, users answered a questionnaire about their level of self-esteem on a seven-level scale. For example, the self-esteem level questionnaire included multiple items such as "I am generally satisfied with myself," "I feel I have various good qualities," and "I feel pretty good about myself." Users answered whether they met each of the seven levels of the questionnaire.

[0289] The questionnaire processing unit 152 can also sum the values ​​of multiple responses to multiple questionnaire items input by the questionnaire receiving unit 154, and output response information including the sum value. Alternatively, the questionnaire processing unit 152 can also average the values ​​of multiple responses to multiple questionnaire items input by the questionnaire receiving unit 154, and output response information including the average value.

[0290] The selection model storage unit 155 pre-stores a learned selection model for selecting one cluster model from multiple clustering models. For example, XGBoost is used to learn the selection model. The selection model classifies the input response information into one of the multiple clusters and outputs identification information for identifying the clustering model associated with a given cluster.

[0291] The model selection unit 156 selects a clustering model suitable for determining the user's psychological state from multiple clustering models used to determine the user's psychological state. The model selection unit 156 selects a clustering model from multiple clustering models based on the response information.

[0292] The model selection unit 156 selects a clustering model from multiple clustering models by applying the response information obtained from the questionnaire processing unit 152 to the learned selection model. When response information is input, the selection model outputs identification information for identifying a clustering model. The model selection unit 156 outputs the identification information for identifying the selected clustering model to the mental state determination unit 141.

[0293] Next, the evaluation process of the psychological state evaluation device 1A in this embodiment 2 will be explained.

[0294] Figure 17 This is a flowchart illustrating the evaluation process of the psychological state evaluation device 1A in Embodiment 2.

[0295] First, in step S41, the questionnaire prompting unit 153 obtains questionnaire data related to the user's personal attributes, which is read from the questionnaire database 151 by the questionnaire processing unit 152, and prompts the user to complete the questionnaire. The questionnaire prompting unit 153 displays a questionnaire for the user to answer regarding their personal attributes.

[0296] Next, in step S42, the questionnaire receiving unit 154 receives the user's input of the questionnaire responses prompted by the questionnaire prompting unit 153.

[0297] Next, in step S43, the questionnaire processing unit 152 obtains response information, which includes the user's answers to the questionnaire related to the user's personal attributes input by the questionnaire receiving unit 154.

[0298] Next, in step S44, the model selection unit 156 selects a clustering model from multiple clustering models by applying the response information obtained by the questionnaire processing unit 152 to the learned selection model. The model selection unit 156 inputs the response information obtained by the questionnaire processing unit 152 into the learned selection model and obtains the identification information of a clustering model output from the selection model.

[0299] Next, in step S45, the model selection unit 156 outputs the identification information used to identify a selected clustering model to the mental state determination unit 141.

[0300] Processing of steps S46 to S49 Figure 2 The processes in steps S9 to S12 are the same, so the explanation is omitted.

[0301] In this way, by having users answer questions in advance, a clustering model suitable for determining the user's psychological state can be selected with high precision based on the answer information containing the user's answers to the questions.

[0302] Furthermore, the model selection process shown in steps S41-S45 and the psychological state determination process shown in steps S46-S49 can be performed continuously in time series or not. After selecting a clustering model suitable for the user, the identification information of the selected clustering model can be stored in the storage unit. In this case, model selection processing can be omitted, and the psychological state determination process can be performed using a clustering model corresponding to the stored identification information. That is, after performing model selection processing first, the psychological state determination process can be repeatedly performed using the first selected clustering model without performing model selection processing.

[0303] Next, the evaluation processing of the psychological state evaluation device and server in the modified example of this embodiment 2 will be described.

[0304] In a variation of Embodiment 2, the psychological state assessment device and the server are communicatively connected via a network, such as the Internet. The psychological state assessment device in this variation of Embodiment 2 includes a data acquisition unit 11, a questionnaire database 151, a questionnaire processing unit 152, a questionnaire prompting unit 153, a questionnaire receiving unit 154, an output unit 15, and a communication unit. The server in this variation of Embodiment 2 includes a feature calculation unit 12A, a selection model storage unit 155, a model selection unit 156, a psychological state determination unit 141, a clustering model storage unit 142, and a communication unit.

[0305] The communication unit of the mental state evaluation device sends response information to the server. This response information includes the user's answers to a questionnaire related to the user's personal attributes, obtained by the questionnaire processing unit 152. Additionally, the communication unit of the mental state evaluation device sends biometric information obtained during the first period of determining the user's mental state via the data acquisition unit 11 to the server. Furthermore, the communication unit of the mental state evaluation device receives the determination result of the user's mental state from the server.

[0306] In addition, the server's communication unit receives response information sent by the mental state evaluation device. Furthermore, the server's communication unit receives biometric information acquired during the first period, sent by the mental state evaluation device. Additionally, the server's communication unit sends the determination result of the user's mental state, as determined by the mental state determination unit 141, to the mental state evaluation device.

[0307] Figure 18 This is a flowchart illustrating the evaluation process of the psychological state evaluation device and server in a modified example of Embodiment 2.

[0308] Processing of steps S51 to S53 Figure 17 The processes in steps S41 to S43 are the same, so the explanation is omitted.

[0309] Next, in step S54, the communication unit of the psychological state evaluation device sends response information to the server. This response information includes the user's answers to a questionnaire survey related to the user's personal attributes obtained by the questionnaire processing unit 152.

[0310] Next, in step S61, the server's communication unit receives the response information sent by the psychological state evaluation device.

[0311] The processing of steps S62, S63 and S55 Figure 17 The processes in steps S44 to S46 are the same, so the explanation is omitted.

[0312] Next, in step S56, the communication unit of the psychological state evaluation device sends the biological information obtained during the first period when the user's psychological state is determined by the data acquisition unit 11 to the server.

[0313] Next, in step S64, the server's communication unit receives the biological information obtained during the first period, sent by the psychological state evaluation device.

[0314] The processing of steps S65 and S66 Figure 17 The processes in steps S47 and S48 shown are the same, so the explanation is omitted.

[0315] Next, in step S67, the server's communication unit sends the determination result of the user's psychological state as determined by the psychological state determination unit 141 to the psychological state evaluation device.

[0316] Next, in step S57, the communication unit of the psychological state evaluation device receives the judgment result of the user's psychological state sent by the server.

[0317] Processing in step S58 and Figure 17 The process of step S49 shown is the same, so the explanation is omitted.

[0318] Furthermore, in a variation of Implementation 2, the mental state evaluation device may also calculate a first biometric feature associated with biometric information obtained during the first period of determining the user's mental state, and send the first biometric feature to the server. The server may also receive the first biometric feature, apply it to a selected clustering model, and thereby determine the user's mental state.

[0319] Next, the experimental results are presented, evaluating the classification accuracy of clusters derived from user responses to questionnaires related to user personal attributes.

[0320] In the experiment, for example, 96 subjects answered 42 questionnaire items, and the answers from the 96 subjects were input into the selection model to evaluate the classification accuracy of the selection model.

[0321] The responses to multiple questionnaire items are represented numerically. XGBoost is used to perform clustering classification, where the response information is input into a selection model, and the selection model outputs cluster numbers used to identify the clusters.

[0322] First, in order to determine the questionnaire items that are effective for model selection, the Shap value is used to identify important features (responses to the questionnaire items).

[0323] Next, the features were input into the selection model in descending order of Shap values, and the relationship between the number of features and classification accuracy was analyzed.

[0324] The model analysis was performed on 96 subjects using k-split cross-validation (k=5). Additionally, the XGBoost parameters were tuned within the training data.

[0325] Figure 19 This is a graph showing the relationship between the number of feature quantities (responses to questionnaire items) and accuracy in Embodiment 2. Figure 19 In the graph, the horizontal axis represents the number of characteristic quantities (responses to questionnaire items), and the vertical axis represents precision. Additionally, in... Figure 19 In the diagram, double-dotted lines represent accuracy, dashed lines represent recall, single-dotted lines represent precision, and solid lines represent the F-score.

[0326] As the results of the experiment where features were added one by one, when using 8 features, the accuracy was 35.9%, the reproducibility was 32.0%, the precision was 29.6%, and the F-score was 27.1%, all exceeding the baseline of 16.6%. This shows that the highest precision can be obtained when using 8 features.

[0327] Figure 20 This is a graph representing the classification accuracy of the selected model when the responses to eight questionnaire items were input. Figure 20 The diagram shows the classification accuracy of the first cluster using the first clustering model and the second cluster using the second clustering model among multiple clusters.

[0328] The first cluster had a reproduction rate of 63%, an accuracy of 47%, and an F-score of 48%. The second cluster had a reproduction rate of 54%, an accuracy of 49%, and an F-score of 48.

[0329] Thus, it can be seen that the reproduction rate exceeds 50% in the first and second clusters, achieving good accuracy.

[0330] Figure 21 This is a graph showing the top 8 questionnaire survey items with the highest Shap values ​​among multiple survey items.

[0331] like Figure 21 As shown, the highest cluster selection accuracy was obtained when the responses to questionnaires related to the moderating focus theory (moderating focus_benefit proximity), the company-related emotional level, the joy level, the self-esteem level, the job standardization level, the pride level, the loneliness level, and the number of moves were input. That is, the responses to the above eight questionnaires are helpful in selecting the clustering model.

[0332] Figure 22 This is a graph showing the average responses of subjects belonging to the first and second clusters to two questionnaire items. Figure 23 This is a graph showing the average responses of subjects belonging to the first and second clusters to the four questionnaire items.

[0333] like Figure 22 As shown, subjects in the first cluster have lower self-esteem and higher loneliness than those in other clusters. Therefore, it can be concluded that subjects classified in the first cluster have a tendency for negative emotions.

[0334] In addition, such as Figure 23 As shown, in the second cluster, subjects whose moderating focus—gain side, joy, pride, and self-esteem were higher than those in other clusters were classified, thus indicating that subjects classified in the second cluster tended to have positive emotions.

[0335] In this way, based on users' answers to questionnaires about their personal emotional characteristics, users are classified into one of multiple clusters that corresponds to their personal emotional characteristics, and the most suitable clustering model for the user is selected from those clusters.

[0336] Figure 24 This is a diagram showing an example of the evaluation result screen displayed in Embodiment 1 and Embodiment 2.

[0337] Output section 15 display Figure 24 The evaluation results screen shown. The evaluation results screen includes a score prompt area 201 indicating past creativity scores (judgment scores) and the current creativity score (judgment score), and an evaluation message 202 related to the user's divergent thinking state. Figure 24 In the scoring area 201, a mental state assessment is performed every 10 minutes, displaying the creativity score (assessment score) for the past 80 minutes and the current creativity score (assessment score). Mental state assessments can also be performed at specified intervals. These intervals can be, for example, 10 seconds, 1 minute, or 5 minutes, and can also be selected by the user.

[0338] The creativity score (judgment score) represents the degree of divergent thinking as determined by the clustering model. The creativity score (judgment score) is represented, for example, from 0 to 100%. The output unit 15 converts the creativity score (judgment score) into a percentage and displays the result.

[0339] When past creative scores (judgment scores) are displayed together with current creative scores (judgment scores), the mental state determination unit 141 can store the creative scores (judgment scores) in memory (not shown).

[0340] The scoring hint area 201 can display a line graph indicating the time trend of the creativity score value, the current creativity score value, and the past creativity score values.

[0341] Furthermore, the evaluation message 202 is a suggestion based on the tendency and current value of the divergent thinking state at each time. For example, the output unit 15 may also display an evaluation message 202 such as "Exercise creativity and maintain a good state." The output unit 15 may also read from memory and display an evaluation message 202 indicating a high level of divergent thinking state (creativity) when the creativity score (judgment score) is above a threshold. In addition, if the creativity score (judgment score) is below the threshold, the output unit 15 may also read from memory and display an evaluation message 202 indicating a low level of divergent thinking state (creativity).

[0342] Furthermore, in Embodiment 1 and Embodiment 2, past creative scores are displayed along with the current creative score. However, this disclosure is not particularly limited to this, and the current creative score may also be displayed without displaying past creative scores.

[0343] Additionally, the evaluation message 202 may also include a message instructing the user to change their posture when a psychological state cannot be determined without obtaining biological information. For example, if the user's face lowers for a certain period due to increased slump or if the posture is poor and seat pressure cannot be accurately measured, the output unit 15 may also read the message from the memory. Figure 24 The evaluation message 202 shown is displayed.

[0344] (Implementation Method 3) In Embodiment 3, the control device is controlled based on the psychological state determined by the psychological state evaluation device 1 of Embodiment 1 and the psychological state evaluation device 1A of Embodiment 2.

[0345] Figure 25 This is a diagram showing the structure of the information processing system 103 in Embodiment 3.

[0346] Figure 25 The information processing system 103 shown includes a psychological state evaluation device 1, a measuring device 2, and a control device 3. The psychological state evaluation device 1 may be a psychological state evaluation device 1A.

[0347] The control device 3 includes at least one of a lighting device that illuminates the space where the user is located and an air conditioning device that regulates the air in the space where the user is located.

[0348] The psychological state evaluation device 1 and the psychological state evaluation device 1A may also include a device control unit that controls the control device 3 based on the psychological state determined by the psychological state determination unit 141.

[0349] If the divergent thinking state is deemed to be of low level, the device control unit sends a control signal to the control device 3, causing a change in the output content of the control device 3. Conversely, if the divergent thinking state is not deemed to be of low level, the device control unit does not send a control signal and maintains the output content of the control device 3. Furthermore, if the divergent thinking state determination score is less than a threshold, the device control unit determines that the divergent thinking state is of low level. The control device 3 can also be a lighting device. In this case, the control signal may also include a control signal that reduces the color temperature and illuminance of the lighting light.

[0350] The equipment control unit can also obtain the user's alertness state. Control device 3 can also be a lighting device. If it is determined that the user's divergent thinking state is low and the alertness state indicates low alertness, the equipment control unit sends a control signal to increase the illuminance of the lighting light. Conversely, if it is determined that the user's divergent thinking state is low and the alertness state indicates high alertness, the equipment control unit sends a control signal to decrease the illuminance of the lighting light.

[0351] Alternatively, control device 3 can also be an air conditioning device. In this case, the control signal can also include a control signal to increase the room temperature.

[0352] When the comfort level is determined to be low within a comfortable state, the equipment control unit sends a control signal to the control device 3 to change the output content of the control device 3. Conversely, when the comfort level is not determined to be low within a comfortable state, the equipment control unit does not send a control signal and maintains the output content of the control device 3. Furthermore, the equipment control unit determines that the comfort level is low if the comfort score is less than a threshold. The control device 3 can also be a lighting device. In this case, the control signal may include a control signal for reducing the color temperature of the lighting light. Alternatively, the control device 3 can also be an air conditioning device. In this case, the control signal may also include a control signal for lowering the room temperature.

[0353] If the level of alertness is determined to be low, the equipment control unit sends a control signal to the control device 3, causing a change in the output of the control device 3. Conversely, if the level of alertness is not determined to be low, the equipment control unit does not send a control signal and maintains the output of the control device 3. The control device 3 can also be a lighting device. In this case, the control signal may include a control signal to increase the color temperature of the lighting light. Alternatively, the control device 3 can also be an air conditioning device. In this case, the control signal may include a control signal to lower the room temperature.

[0354] Furthermore, in the above embodiments, each component may be constructed using dedicated hardware, or implemented by executing software programs suitable for each component. Each component may also be implemented by a program execution unit such as a CPU or processor reading and executing software programs recorded on recording media such as hard disks or semiconductor memory. Alternatively, the program may be implemented using a separate computer system by recording the program on a recording medium and transferring it, or by transferring the program via a network.

[0355] The functionality of the apparatuses in the embodiments of this disclosure is typically implemented, in part or in whole, as an integrated circuit, i.e., a Large Scale Integration (LSI). They can be implemented as a single chip, or as a single chip including some or all of them. Furthermore, integrated circuit implementation is not limited to LSIs; it can also be implemented using dedicated circuits or general-purpose processors. LSIs can also utilize FPGAs (Field Programmable Gate Arrays) that can be programmed after manufacturing, or reconfigurable processors that can reconfigure the connections and settings of the circuitry within the LSI.

[0356] Alternatively, a CPU or other processor can be used to execute programs to achieve some or all of the functions of the apparatus of the present disclosure.

[0357] Furthermore, all the figures used above are illustrative for the purpose of explaining this disclosure, and this disclosure is not limited to the illustrative figures.

[0358] Furthermore, the order in which the steps shown in the flowchart above are executed is illustrative for the purpose of explaining this disclosure, and other orders may be used to achieve the same effect. Additionally, some of the above steps may be executed simultaneously (in parallel) with other steps.

[0359] Industrial applicability The technology disclosed herein can accurately determine a user's psychological state, and is therefore useful as a technology for determining a user's psychological state.

[0360] Symbol Explanation 1,1A Psychological state assessment device 2. Measuring apparatus 3. Control equipment 11 Data Acquisition Department 12,12A Characteristic Calculation Unit 13,13A Model Selection Processing Unit 14. Psychological Status Assessment and Processing Department 15 Output Section Information processing systems 101, 102, 103 131 Mission Execution Department 132 Task Evaluation Department 133 Correlation Calculation Department 134 Model Selection Department 135 Mission Prompt Department 136 Task Acceptance Department 137 Task Database 138 Evaluation Benchmark Storage Department 139 Select Model Storage Unit 141 Psychological State Assessment Department 142 Clustering Model Storage Section 151 Questionnaire Survey Database 152 Questionnaire Processing Department 153 Questionnaire Survey Prompt Department 154 Questionnaire Survey Handling Department 155 Select Model Storage Unit 156 Model Selection Department

Claims

1. An information processing method, Information processing methods performed by computers include: Choose one clustering model from among the multiple clustering models used to determine the user's psychological state, and select the one that is suitable for determining the user's psychological state. Obtain a first biometric quantity associated with the biometric information of the user during the first period; as well as Based on the first biological characteristic quantity and the clustering model, the user's psychological state during the first period is determined.

2. The information processing method according to claim 1 further includes: Obtain task evaluation information representing the evaluation results of the tasks performed by the user; as well as Obtain a second biometric quantity associated with the user's biometric information during the second period when the user performs the task. The selection of a clustering model includes: selecting one clustering model from the plurality of clustering models based on the correlation between the second biological feature quantity and the evaluation result shown by the task evaluation information.

3. The information processing method according to claim 2, wherein, The task is designed to enable the user to engage in divergent thinking, and the mental state is the divergent thinking state of the user during this process.

4. The information processing method according to claim 3, wherein, The bio-information includes information associated with at least one of brain waves, cerebral blood flow, facial images, heart rate indicators, seat pressure, skin potential, and skin temperature.

5. The information processing method according to claim 3, wherein, The biological information includes brain waves, facial images, or information associated with the brain waves and the facial images. The plurality of clustering models includes a first clustering model, which determines that the degree of divergent thinking is proportional to at least one of the magnitude of the activity of the γ wave, which is a first biometric quantity of the brainwave, and the magnitude of the facial movement, which is a first biometric quantity of the facial image.

6. The information processing method according to claim 3, wherein, The biological information includes heart rate indicators, facial images, or information associated with the heart rate indicators and the facial images. The plurality of clustering models includes a second clustering model, which determines that the degree of divergent thinking is proportional to at least one of the magnitude of the reciprocal of the value of the heart rate index, which is a first biometric quantity representing the heart rate interval, and the magnitude of the facial movement shown in the facial image, which is a first biometric quantity.

7. The information processing method according to claim 1, wherein, The information processing method further includes obtaining answer information containing the user's response to the question. The selection of a clustering model includes choosing one clustering model from the plurality of clustering models based on the response information. The psychological state referred to is the divergent thinking state in which the user engages in divergent thinking.

8. The information processing method according to claim 7, wherein, The questions include questions related to at least one of the following: demographic attributes, positive emotional characteristics, behavioral characteristics, physical and mental health, and psychological characteristics.

9. The information processing method according to claim 8, wherein, Questions related to the demographic attributes include at least one of the following: questions related to the number of times the user has moved; questions related to the user's emotional attachment to the company; and questions related to the standardization of the user's work.

10. The information processing method according to claim 8, wherein, Questions related to the positive emotional characteristics include at least one of questions related to the user's joy and questions related to the user's pride.

11. The information processing method according to claim 8, wherein, Questions related to the aforementioned behavioral characteristics include those related to the intention to approach benefits in the moderating focus theory.

12. The information processing method according to claim 8, characterized in that, The questions related to the user's physical and mental health include those related to the degree of loneliness the user experiences.

13. The information processing method according to claim 8, wherein, Questions related to the psychological characteristics include those related to the user's level of self-esteem.

14. The information processing method according to claim 7, wherein, The questions include items related to the user's level of self-esteem.

15. The information processing method according to claim 2, wherein, The task includes appreciating content designed to alter at least one of the user's comfort and alertness. The psychological state includes at least one of a comfort state, which represents the user's level of comfort, and a lucid state, which represents the user's level of alertness.

16. The information processing method according to claim 15, wherein, The biological information includes information related to at least one of facial images and heart rate indicators.

17. The information processing method according to claim 15, wherein, The psychological state includes the user's comfort state. The biological information includes information associated with facial images. The plurality of clustering models includes a first clustering model, which determines that the comfort level of the comfort state increases proportionally to at least one of the following: the magnitude of the rotation of the face about the vertical direction as a first biometric quantity shown in the facial image; the magnitude of the reciprocal of the value representing the deviation of the movement of the face about the vertical direction as a first biometric quantity shown in the facial image; and the magnitude of the reciprocal of the value representing the opening of the eyes as a first biometric quantity shown in the facial image.

18. The information processing method according to claim 15, wherein, The psychological state includes the user's comfort state. The biological information includes facial images, heart rate indicators, or information associated with the facial images and heart rate indicators. The plurality of clustering models includes a second clustering model, which determines that the comfort level of the comfort state increases proportionally to at least one of the following: the magnitude of the facial rotation about the vertical direction as a first biometric quantity shown in the facial image; the magnitude of the reciprocal of the value of the deviation representing the amount of movement of the facial rotation about the vertical direction as a first biometric quantity shown in the facial image; and the magnitude of the HF component of the heart rate variability as a first biometric quantity of the heart rate index.

19. The information processing method according to claim 15, The psychological state includes the user's conscious state. The biological information includes facial images, heart rate indicators, or information associated with the facial images and heart rate indicators. The plurality of clustering models includes a first clustering model, which determines that at least one of the following increases proportionally: the difference between the level of alertness in the waking state and the maximum and minimum values ​​of the heart rate as a first biological characteristic, the median value of the heart rate as a first biological characteristic, and the magnitude of the deviation of eye movement as a first biological characteristic as shown in the facial image.

20. The information processing method according to claim 15, The psychological state includes the user's conscious state. The biological information includes facial images, heart rate indicators, or information associated with the facial images and heart rate indicators. The plurality of clustering models includes a second clustering model, which determines that the level of alertness in the waking state is proportional to at least one of the following: the magnitude of the average heart rate as a first biological characteristic, the magnitude of the difference between the maximum and minimum values ​​of the heart rate as a first biological characteristic, and the magnitude of the vertical length of the cornea of ​​the eye as a first biological characteristic as shown in the facial image.

21. The information processing method according to claim 15, The task evaluation information represents the user's evaluation results related to the comfort level or the alertness level.

22. An information processing system, comprising: The selection unit selects a clustering model suitable for determining the user's psychological state from multiple clustering models used to determine the user's psychological state. The acquisition unit acquires a first biometric feature associated with the user's biometric information during the first period; and The determination unit determines the user's psychological state during the first period based on the first biological characteristic quantity and the clustering model.

23. An information processing method, Information processing methods performed by computers include: Obtain task evaluation information that represents the evaluation results of tasks performed by users; Obtain a first biometric quantity related to the user's biometric information during a first period when the user performs the task; as well as Based on the correlation between the first biological characteristic quantity and the evaluation result shown by the task evaluation information, a clustering model suitable for determining the user's psychological state is selected from multiple clustering models used to determine the user's psychological state.

24. An information processing method, Information processing methods performed by computers include: Obtain answer information that includes the user's response to the question; as well as Based on the answer information, from among multiple clustering models used to determine the user's divergent thinking state, a clustering model suitable for determining the user's divergent thinking state is selected.

Images