Information processing method, information processing device, and information processing program

Abstract

An information processing method according to the present disclosure comprises the following performed by a computer: acquiring user data relating to a user's sleep; generating, on the basis of the acquired user data, classification data in which the type of the user's sleep is classified as a given type of sleep, and recommended action data relating to a recommended action which is an action recommended for improving the user's sleep; and presenting the generated classification data and recommended action data to the user.

Description

Information processing method, information processing device, and information processing program

[0001] The present disclosure relates to an information processing method, an information processing device, and an information processing program.

[0002] With the recent spread of smartphones, technologies that recommend actions to improve illnesses using mobile applications have become known. Similarly, technologies that present recommended actions to improve sleep, which many people have problems with, have also become known.

[0003] Known technologies for presenting recommended behaviors include a technology that outputs messages encouraging improvement in sleep based on information such as the user's wake-up time, sleep duration, QOL (Quality of Life) assessment indicating an evaluation of the user's daytime quality of life, and the user's daytime activities.

[0004] Furthermore, a technology is known that presents to a user an output automatically determined by a model according to the phase of a sleep improvement program based on CBT-I (Cognitive Behavior Therapy for Insomnia). This technology corrects the criteria for determining the model's output based on accumulated performance data such as past model inputs and outputs.

[0005] JP 2016-122347 A International Publication No. 2019 / 077898

[0006] Technology that outputs messages that promote sleep improvement can present recommended actions to a wide range of people who have sleep problems, while technology that presents automatically determined outputs can present recommended actions that are automatically determined to be appropriate for the user from a collection of outputs.

[0007] However, conventional technologies may not convince users to accept recommended actions. For example, technologies that output messages promoting sleep improvement recommend uniform actions for a wide range of people without classifying users' sleep types. However, because the level of sleep disorder ranges from severe illness to temporary sleep disorder in healthy people, the effective response varies greatly depending on the cause of the sleep disorder.

[0008] Furthermore, the technology that presents automatically determined output assumes treatment with CBT-I, and does not anticipate milder responses or the improvement of mental health through CBT that is not limited to insomnia, which in turn has an effect on sleep.

[0009] As described above, in conventional technologies, the recommended actions may not correspond to the level or causes of the user's own sleep disorder, and the user may think that the recommended actions are not suitable for them.

[0010] Therefore, an object of the present disclosure is to propose an information processing method, an information processing device, and an information processing program that enable a user to accept recommended actions with a sense of satisfaction.

[0011] The information processing method according to the present disclosure includes a computer acquiring user data related to a user's sleep, generating classification data based on the acquired user data, in which the user's sleep type is classified into one of the sleep types, and recommended action data related to recommended actions that are actions recommended for improving the user's sleep, and presenting the generated classification data and recommended action data to the user.

[0012] 1 is a diagram illustrating an example of a schematic configuration of an information processing system according to a first embodiment; FIG. 2 is a diagram illustrating an example of image data displayed on a user terminal; FIG. 3 is a diagram illustrating an example of image data displayed on a user terminal; FIG. 4 is a block diagram illustrating an example of a configuration of a user terminal and a server according to the first embodiment; FIG. 5 is a diagram illustrating an example of a relationship between feature amounts of user data and original data of a profile; FIG. 6 is a diagram illustrating an example of image data displayed on a user terminal; FIG. 7 is a block diagram illustrating an example of a configuration of an information processing system according to the first embodiment; FIG. 8 is a flowchart illustrating an example of a procedure for generating and presenting a profile; FIG. 9 is a flowchart illustrating an example of a procedure for generating and presenting content; FIG. 10 is a block diagram illustrating an example of a configuration of a user terminal and a server according to a second embodiment; FIG. 11 is a diagram illustrating an example of feature amount data and sleep quality data; and FIG. 12 is a block diagram illustrating an example of a configuration of a user terminal and a server according to a third embodiment.

[0013] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In the following embodiments, the same components are denoted by the same reference numerals, and redundant description will be omitted.

[0014] Hereinafter, embodiments of the present disclosure will be described in the following order: 1. First embodiment 1-1. Overview of information processing system according to first embodiment 1-2. Configuration of user terminal and server according to first embodiment 1-3. Configuration of information processing system according to first embodiment 1-4. Information processing procedure according to first embodiment 2. Second embodiment 3. Third embodiment 4. Other embodiments 5. Effects of information processing method according to the present disclosure 6. Hardware configuration 7. Supplementary information

[0015] (1. First Embodiment) (1-1. Overview of Information Processing System According to First Embodiment) An overview of an information processing system 100 according to a first embodiment will be described using Fig. 1. Fig. 1 is a diagram showing an example of a schematic configuration of the information processing system 100 according to the first embodiment.

[0016] 1, the information processing system 100 includes an information acquisition device 1, a server 2, a medical professional terminal 3, a community member terminal 4, a life insurance / health insurance terminal 5, a service provider terminal 6, and an analysis terminal 7. Also shown in Fig. 1 is a user 110 of the information processing system 100. The devices and terminals included in the information processing system 100 are connected to each other so as to be able to communicate with each other via a network 8 such as the Internet.

[0017] The information acquisition device 1 includes a sensor 11 and a user terminal 12. The sensor 11 is a wearable device or the like provided to a user 110 who may have sleep problems.

[0018] The user terminal 12 is a smartphone or the like used by the user 110. The user terminal 12 acquires user data related to the sleep of the user 110 from the sensor 11.

[0019] For example, the user terminal 12 acquires wearable data, which is data derived from a wearable device, as user data from the sensor 11. In addition, the user terminal 12 acquires subjective index data, which is data related to subjective measures of the user 110's sleep state, mental state, personality, sleep preferences, habits, etc., by accepting input from the user 110 to the user terminal 12.

[0020] The user terminal 12 transmits user data to the server 2. At least a part of the sensor 11 may be incorporated in the user terminal 12. There may be a plurality of user terminals 12.

[0021] The server 2 is an information processing device such as a cloud server that presents recommended actions, which are actions recommended for improving sleep, to the user 110. For example, the server 2 receives user data from the user terminal 12, thereby acquiring the user data.

[0022] The server 2 also generates content including recommended action data based on the acquired user data. The "recommended action data" is data related to recommended actions. For example, the recommended action data is original data of content in which recommended actions to be recommended to the user 110 are determined based on the acquired user data.

[0023] The "content" is the recommended action data in the form of output content for the user 110. For example, the content is image data such as an application page that includes recommended action data that describes recommended actions and their explanations.

[0024] Next, the server 2 transmits the content to the user terminal 12. The user terminal 12 displays the content. In this way, the server 2 presents the content to the user 110. The server 2 can also transmit the content to other devices or terminals.

[0025] The medical staff terminal 3 is a terminal used by medical staff 30 such as doctors, nurses, pharmacists, physical therapists, caregivers, etc. The community member terminal 4 is a terminal used by members 40 who belong to the same community as the user 110, such as family members of the user 110 and other patients with the same sleep disorder as the user 110.

[0026] The life insurance / health insurance terminal 5 is a terminal used by insurance companies, health insurance companies, etc. The service provider terminal 6 is a business terminal used by service providers who provide products and services to users 110. The analysis terminal 7 is a terminal used by companies that provide products / services, pharmaceutical companies that conduct clinical development, etc.

[0027] As described above, the server 2 presents recommended actions to the user 110. However, because the level of sleep disorder varies widely, from a serious illness to a temporary condition experienced by healthy individuals, the effective response varies greatly depending on the cause of the disorder. For this reason, simply presenting recommended actions to the user 110 does not allow the user 110 to understand whether the recommended actions correspond to the level or cause of the user's sleep disorder, and the user 110 may not be able to accept the recommended actions with a sense of conviction.

[0028] Therefore, in order to convince the user 110 to accept the recommended actions, the server 2 generates and presents a profile based on the user data in addition to acquiring the user data and generating and presenting the content as described above.

[0029] A "profile" is classification data in which the sleep type of the user 110 is classified into one of the sleep types. An example of a procedure for generating and presenting a profile by the server 2 will be described below. The following describes an example in which the server 2 has acquired the above-mentioned wearable data and subjective index data as user data.

[0030] For example, if the wearable data indicates that the average weekly sleep time is 310 minutes, the server 2 classifies the sleep type of the user 110 into a type called "#slight insomnia" and a type called "#insomnia." Furthermore, if the subjective index data indicates that the user 110 has a habit of waking up in the middle of the night, the server 2 classifies the sleep type of the user 110 into a type called "#awakened midnight."

[0031] The server 2 generates a profile in which the sleep-related types of the user 110 are classified into a type "#slight insomnia", a type "#middle of the night awakening", and a type "#insufficient sleep".

[0032] Next, the server 2 transmits the generated profile to the user terminal 12. The user terminal 12 displays the profile. In this way, the server 2 presents the profile to the user 110.

[0033] An example of a profile and content will be described below with reference to Figures 2 and 3. Figures 2 and 3 are diagrams showing an example of image data displayed on a user terminal.

[0034] 2 includes a profile 501 in which the sleep types of the user 110 are classified into "#Slight Insomnia," "#Awakening in the Middle of the Night," and "#Lack of Sleep." Furthermore, the image data 60, which is the content of FIG. 3, includes recommended behavior data 601 that recommends a sleep schedule program and a visit to a specialized medical institution as recommended behaviors.

[0035] The sleep schedule program is a recommended action that recommends a program based on the sleep schedule method, which is a means of adjusting sleep rhythms, to improve sleep. For example, the sleep schedule program is a recommended action that recommends finding an appropriate sleep rhythm and bedtime and wake-up time that maintains sleep efficiency over several weeks. The consultation with a specialized medical institution is a recommended action that recommends visiting a specialized medical institution online or in person to improve sleep.

[0036] In this way, the server 2 presents to the user 110, in addition to the content, a profile in which the sleep type of the user 110 is classified into one of the sleep types.

[0037] This allows the user 110 to roughly understand their own sleep state and the causes of their sleep disorders from the presented profile, and therefore they can understand that the recommended actions correspond to the level and causes of their sleep disorders. For example, the user 110 can understand that the sleep schedule program, etc. of the recommended action data 601 corresponds to the causes of sleep disorders based on the sleep problems "#Mild insomnia," "#Midnight awakening," and "#Lack of sleep" in the profile 501, rather than to mental illness. This allows the user 110 to be convinced that the recommended actions are suitable for them.

[0038] Therefore, the server 2 can make the user 110 accept the recommended action with a sense of satisfaction.

[0039] (1-2. Configuration of User Terminal and Server According to First Embodiment) Next, an example of the configuration of the user terminal 12 and server 2 that constitute the information processing system 100 according to the first embodiment will be described with reference to Fig. 4. Fig. 4 is a block diagram showing an example of the configuration of the user terminal and server according to the first embodiment.

[0040] (User Terminal) The user terminal 12 includes an acquisition unit 13 and a presentation unit 14. Although not shown in Fig. 4, the user terminal 12 also includes a communication unit 15, a storage unit 16, and a processing unit 17. The communication unit 15, the storage unit 16, and the processing unit 17 will be described later.

[0041] (Acquisition Unit) The acquisition unit 13 functions as a control unit that controls the user terminal 12 and acquires user data. The control unit is a CPU (Central Processing Unit) or MPU (Micro Processing Unit) that executes a program stored in a storage device such as a RAM (Random Access Memory). The control unit may be an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array).

[0042] The acquisition unit 13 can acquire the user data by different means, for example, by acquiring some of the user data from the sensor 11 such as a wearable device and acquiring other user data as input data to the user terminal 12 .

[0043] For example, the acquisition unit 13 acquires wearable data as user data from the sensor 11. The acquisition unit 13 also acquires subjective index data, demographics data, therapy history data, sleep diary data, and behavioral record data as user data input by the user 110 to the user terminal 12 through a digital application.

[0044] As described above, "wearable data" is data derived from a wearable device. The wearable data is, for example, data related to sleep over a week and activity data measured by a wearable device. Specifically, the wearable data is the user's 110 heart rate, sleep stage, body surface temperature, number of steps, calories burned, acceleration of the part of the user terminal 12 worn by the user 110, and behavior classification data that classifies the user's 110 behavior.

[0045] The sleep stages include REM sleep, non-REM sleep, etc. The behavior classification data includes running, swimming, cycling, traveling by train, etc. Other wearable data includes location data indicated by a Global Positioning System (GPS), air pressure, wearing time, etc., which can be obtained from a biometric information logger.

[0046] As described above, the "subjective index data" is data relating to subjective measures of the user 110's sleep state, mental state, personality, sleep preferences, habits, etc. The subjective index data is, for example, the user's 110's answers to questionnaires, surveys, etc. relating to the user's 110's mental state and personality. Specifically, the subjective index data is the user's 110's answers to each item in questionnaires such as the Patient Health Questionnaire (PHQ)-9, Generalized Anxiety Disorder (GAD)-7, and 16 Personalities, as well as the total score of the answers.

[0047] The subjective index data is, for example, the answers of the user 110 to a questionnaire, a survey, or the like regarding the sleep state of the user 110. Specifically, the subjective index data is the answers of the user 110 to each item of a questionnaire such as the Insomnia Severity Index (ISI), the Pittsburgh Sleep Quality Index (PSQI), the Athens Insomnia Scale (AIS), the Epworth Sleepiness Scale (ESS), or the Dysfunctional Beliefs and Attitudes about Sleep (DBAS), the total score of the answers, or the like.

[0048] The subjective index data is data on the habits of the user 110, such as whether the user 110 sleeps alone or with others, whether the user sleeps on a bed or tatami mat, and whether the user wakes up in the middle of the night.

[0049] "Demographics data" is data related to the characteristics and demographics of the user 110. Demographics data includes, for example, the age, sex, occupation, family, medical history, and working status of the user 110.

[0050] The "therapy history data" is data related to the therapy history of the user 110. The therapy history data includes, for example, the progress status of cognitive therapy, cognitive behavioral therapy, and behavioral therapy programs, input data from the user 110 for various input items used in these therapies, and whether or not sleeping pills are taken and how often.

[0051] The "sleep diary data" is data related to a sleep diary used in cognitive behavioral therapy to improve insomnia. The sleep diary data is data that expresses, as subjective numerical values, subjective sleep duration, time spent awake during the night, daytime sleepiness, and quality of awakening after waking, for example.

[0052] The "activity record data" is data relating to a record of specific activities of the user 110, such as when, where, and what the user 110 did. The activity record data is, for example, data indicating that the user 110 exercises by swimming twice a week.

[0053] (Presentation Unit) The presentation unit 14 is a display unit such as a desktop of a smartphone that presents the generated profile and content to the user 110. The presentation unit 14 presents the profile generated by a profile output generation unit 24 (described later) and the content generated by a content output generation unit 26 (described later) to the user 110.

[0054] For example, the presentation unit 14 displays the image data 50 including the profile 501 of FIG. 2 received from the server 2, and then displays the image data 60 including the recommended behavior data 601 of FIG.

[0055] (Server) The server 2 includes a preprocessing unit 21, a learning unit 22, a profiling unit 23, a profile output generation unit 24, a recommended behavior determination unit 25, and a content output generation unit 26. The server 2 also includes a first trained model 281, a second trained model 282, a first database 283, and a second database 284. Although not shown in FIG. 4 , the server 2 also includes a communication unit 27, a storage unit 28, and a processing unit 29. The communication unit 27, the storage unit 28, and the processing unit 29 will be described later.

[0056] (Preprocessing Unit) The preprocessing unit 21 functions as a control unit that controls the server 2 and calculates feature quantities based on the acquired user data. The control unit is a CPU, an MPU, or the like that executes a program stored in a storage device such as a RAM. The control unit may also be an integrated circuit such as an ASIC or an FPGA.

[0057] For example, the preprocessing unit 21 acquires user data by receiving it from the user terminal 12. The preprocessing unit 21 calculates a feature vector based on the acquired user data. The "feature vector" is a vector in which each component of a feature corresponding to each user data is expressed as a value of each dimension.

[0058] Specifically, the preprocessing unit 21 calculates, as the values ​​of the components of the feature vector, values ​​(such as the mean, maximum, minimum, median, variance, and standard deviation) obtained as summary statistics based on the values ​​of each user data itself or a set of values ​​for the same user 110 over a predetermined period. The preprocessing unit 21 can also use various scaling methods, such as min-max normalization and z-score normalization, to calculate the values ​​of the components of the feature vector.

[0059] The preprocessing unit 21 can also calculate the values ​​of the components of the feature vector by comparing the values ​​of the user data of the user 110 with those of other users, such as the relative ranking of the values ​​of the user data of the user 110 in the member 40, which is a group to which the user 110 who meets the criteria for calculating the feature belongs.

[0060] The preprocessing unit 21 can also calculate a feature vector by reducing the dimension of all variables of the feature. The dimension reduction method is not particularly limited as long as it reduces the dimension of the original data by one or more dimensions. Examples of dimension reduction methods include PCA (Principal Component Analysis), t-SNE (t-distributed Stochastic Neighbor Embedding), and UMAP (Uniform Manifold Approximation and Projection).

[0061] An example of calculation of feature quantities by the preprocessing unit 21 will be described below with reference to Fig. 5. Fig. 5 is a diagram showing an example of the relationship between feature quantities of user data and original data of a profile. The original data of a profile will be described later.

[0062] In FIG. 5 , it is assumed that the preprocessing unit 21 receives wearable data and demographics data as user data from the user terminal 12 .

[0063] Specifically, the preprocessing unit 21 acquires, as wearable data, data indicating that the average weekly sleep duration measured by the wearable device is 310 minutes, etc. The preprocessing unit 21 also acquires, as demographic data, data indicating that the user 110 is female, 52 years old, and an office worker (fixed shift) as the occupation.

[0064] For wearable data, the preprocessing unit 21 calculates the value of the wearable data itself as the value of the component of the feature vector. For example, if the average weekly sleep time measured by a wearable device is 310 minutes, the preprocessing unit 21 calculates the value of the component of the feature vector of the item "Weekly Mean Wearable Sleep Time" in the item "Wearable Data" 71 in Fig. 5 as "310."

[0065] In addition, for demographic data, if the data corresponds to a specified item, the preprocessing unit 21 calculates the value of the component of the feature vector as "1," and if the data does not correspond to the specified item, the preprocessing unit 21 calculates the value of the component as "0."

[0066] For example, the gender of the user 110 indicated by the demographics data does not meet the condition of "being male" defined by the item "Male / Female" in the item "Demographics Data" 72 in Fig. 5. Therefore, the preprocessing unit 21 calculates the value of the component of the feature vector of the item "Male / Female" as "0".

[0067] (Learning Unit) Returning to the explanation of Fig. 4, the learning unit 22 functions as a control unit that controls the server 2, and generates a first trained model 281 and a second trained model 282.

[0068] The first trained model 281 is a model that outputs classification data in which sleep types are classified into one of several types in response to input feature data. "Feature data" is data related to sleep features. The feature data may be, for example, data related to the sleep of an individual or group corresponding to user data, or may be features related to the sleep of an individual or group calculated based on data corresponding to user data. The second trained model 282 is a model that outputs recommended behavior data related to recommended behaviors in response to input classification data.

[0069] The learning unit 22 generates a first trained model 281 by training a model based on training data in which, for example, feature data is an input variable and classification data is an output variable (correct answer). The learning unit 22 also generates a second trained model 282 by training a model based on training data in which, for example, classification data is an input variable and recommended action data is an output variable.

[0070] The learning unit 22 can also use the data on the sleep of the user 110 as the data on the sleep of an individual among the teacher data. In this case, the learning unit 22 generates, for example, an individualized first trained model 281 that outputs the classification data of the user 110 as classification data in response to input of the feature data of the user 110 as feature data.

[0071] The learning unit 22 can also use feature data as an input variable instead of classification data to generate the second trained model 282. In this case, the second trained model 282 is a model that outputs recommended action data in response to input feature data.

[0072] (Profiling Unit) The profiling unit 23 functions as a control unit that controls the server 2, and generates original data for a profile based on user data. The "original data for a profile" is data that is the basis for the profile 501 shown in FIG. 2 and the like that is presented to the user 110. For example, the original data for a profile is the original data 80 for a profile shown in FIG. 5.

[0073] When generating the original profile data, the profiling unit 23 calculates a profile vector based on, for example, a feature vector calculated based on user data. A "profile vector" is a vector in which each component of the original profile data is expressed as a value of each dimension. The profile vector is also a vector that indicates the degree to which each item of the original profile data 80 applies to the user 110.

[0074] 5, an example of generation of the profile raw data 80 by the profiling unit 23 will be described. When generating the profile raw data 80, the profiling unit 23 refers to the degree to which one or more items related to the sleep state apply to the user 110.

[0075] Specifically, when generating the original profile data 80, the profiling unit 23 refers to the degree to which each of the following items applies to the user 110: item 81, "Causes of sleep problems," item 82, "Level of insomnia," item 83, "Sleep worries," and item 84, "Demographic data."

[0076] Item 81, "Causes of sleep problems," is composed of subitems such as "insomnia," "depression," "menopausal disorders," "sleep problems specific to shift workers," "lack of sleep due to long working hours," "difficulty falling asleep due to lack of exercise," and "health."

[0077] The item 82 titled "Level of insomnia" is composed of sub-items such as "Severe insomnia," "Insomnia," "Mild insomnia," and "Healthy." The item 83 titled "Sleep problems" is composed of sub-items such as "Slow to fall asleep," "Frequent awakenings during the night," "Waking up earlier than planned," "Insufficient sleep time," and "Large discrepancy between weekday and weekend sleep habits."

[0078] The "demographics data" item 84 is made up of a part of the demographics data of the user data, such as the "gender" item and the "age" item.

[0079] Furthermore, when generating the profile raw data 80, the profiling unit 23 inputs user data to the first trained model 281. In Fig. 5 , when generating the profile vector that constitutes the profile raw data 80, the profiling unit 23 inputs a feature vector composed of values ​​of each component element as the feature 70 calculated based on the user data to the first trained model 281.

[0080] For example, the profiling unit 23 calculates a profile vector based on the product of a pre-selected feature vector and a weight "w" from among the feature vectors, so that the upper limit value of the component values ​​of the profile vector is "1" and the lower limit value is "0".

[0081] Specifically, when calculating the values ​​of the components of the profile vector of the item "insomnia" in the item 81 "causes of sleep problems," the profiling unit 23 refers to the product of "w" and a feature vector made up of values ​​of the components of the wearable data, demographics data, sleep diary data, and other data. The learning unit 22 targets "w" to be updated by learning the first trained model 281.

[0082] The profiling unit 23 can also calculate the profile vector by referring to a manually updated model such as a first database 283 that has been created manually or the like in advance, for values ​​of the components of the profile vector such as items 83 and 84. The first database 283 is a table or the like in which the items of the feature vector and the profile vector are associated in advance in a one-to-one correspondence.

[0083] For example, when calculating the values ​​of components of a profile vector such as the item "Age" in the item "Demographics Data" 84, the profiling unit 23 refers to the first database 283 in which the values ​​of the components of the feature vector of the item "Age" in the item "Demographics Data" 72 in Figure 5 have been adjusted so that 0 years old is "0" and 70 years old or older is "1".

[0084] Furthermore, when calculating the profile vector, the profiling unit 23 can use separate models, such as the first trained model 281 and the first database 283, for each profile item, such as items 81 to 84. In this case, the profiling unit 23 integrates the results of the separate models used.

[0085] (Profile Output Generator) Returning to the description of Fig. 4, the profile output generator 24 functions as a controller that controls the server 2, and generates a profile to be presented to the user 110 based on the generated profile. For example, the profile output generator 24 generates image data 50 including the profile 501 of Fig. 2 based on the calculated profile vector.

[0086] 2 and 5, an example of profile generation by the profile output generation unit 24 will be described. For example, the profile output generation unit 24 classifies the sleep type of the user 110 as "#Slight Insomnia" based on the values ​​of the components of the profile vector of each subitem in the item 81 titled "Causes of Sleep Problems" in Fig. 5 and the values ​​of the components of the profile vector of each subitem in the item 82 titled "Level of Insomnia."

[0087] The profile output generation unit 24 classifies the sleep type of the user 110 as "#awakening in the middle of the night" based on the values ​​of the components of the profile vector of the item "frequent awakening in the middle of the night" in the item "sleep concerns" 83 in Fig. 5. The profile output generation unit 24 also classifies the sleep type of the user 110 as "#insufficient sleep" based on the values ​​of the components of the profile vector of the item "not enough sleep" in the item "sleep concerns" 83 in Fig. 5.

[0088] The profile output generation unit 24 generates image data 50 including the profile 501 shown in Fig. 2 based on the results of classifying the sleep types of the user 110. As shown in Fig. 2, the presentation unit 14 of the user terminal 12 displays the image data 50 including the profile 501.

[0089] 2 , a button 503 or a button 504, which is a UI (User Interface) displayed on the presentation unit 14, receives a selection from the user 110 as to whether or not to accept the profile 501. The profile output generation unit 24 acquires selection data from the user terminal 12, which is data received by the button 503 or the button 504 as to whether or not the user 110 will accept the profile 501.

[0090] If button 503 is selected, that is, if the selection data indicates that profile 501 has been accepted by user 110 , profile output generating unit 24 sets the profile of user 110 to profile 501 .

[0091] When button 504 is selected, that is, when the selection data indicates that profile 501 is not accepted, profile output generating unit 24 sets the profile of user 110 to a neutral profile. A "neutral profile" is, for example, a profile in which the sleep type of user 110 is classified as a general type that applies to a wide range of people.

[0092] 2, button 503 or button 504 accepts a selection of whether to accept all of the multiple sleep types that make up profile 501, namely, "#Slight Insomnia," "#Awakening in the Middle of the Night," and "#Lack of Sleep," at once. However, button 503 or button 504 can also accept a selection of whether to accept each of the multiple sleep types.

[0093] In this case, if at least one of the multiple sleep types constituting the profile 501 is accepted, the profile output generation unit 24 sets the profile of the user 110 to a profile corresponding to the accepted sleep type among the profile 501. Furthermore, if none of the multiple sleep types constituting the profile 501 is accepted, the profile output generation unit 24 sets the profile of the user 110 to a neutral profile.

[0094] Returning to the description of FIG. 4 , when generating a profile, the profile output generation unit 24 can also select a character corresponding to the profile. An example of a character corresponding to the profile will be described below with reference to FIG. 6 . FIG. 6 is a diagram showing an example of image data displayed on a user terminal. For example, when the image data 50 of FIG. 2 is displayed on the presentation unit 14 of the user terminal 12, the image data 90 in FIG. 6 is image data transitioned from the image data 50 when the button 503 receives an input from the user 110 by tapping or the like.

[0095] For example, when the profile 501 is accepted, the profile output generation unit 24 selects at least one of the character A901 and the character B902 from a plurality of characters generated in advance. Specifically, when the profile 501 is accepted, the profile output generation unit 24 refers to a predetermined correspondence relationship between the profile and the character, and selects at least one of the character A901 and the character B902.

[0096] "Character A" is a character that represents the sleep type of user 110. "Character B" is a character that is suitable as a partner for user 110 in improving the sleep of user 110. Character B communicates recommended actions to user 110 by any means, such as text data displayed on the presentation unit 14 of the user terminal 12 or audio data output from the user terminal 12.

[0097] When generating a profile, the profile output generating unit 24 can also generate a profile configuration as the profile. The "profile configuration" is a profile that includes information about user data that is the basis for classifying the sleep type of the user 110 into any of the sleep types among the user data.

[0098] An example of the configuration of a profile will be described below with reference to Fig. 7. Fig. 7 is a diagram showing an example of image data displayed on a user terminal. In a case where the image data 50 in Fig. 2 is displayed on the presentation unit 14 of the user terminal 12, the image data 120 in Fig. 7 is image data transitioned from the image data 50 when the UI 502 "Display original data" receives an input from the user 110 by tapping or the like.

[0099] The profile output generation unit 24 generates a profile configuration 1202 that includes a description 1201 of the "sleep time" that is a feature of the user data that contributed to determining the profile 501 and corresponds to the item "Weekly Mean Wearable Sleep Time" in the item 71 "Wearable Data" in Figure 5.

[0100] Furthermore, the profile output generation unit 24 generates image data 50 including the profile 501 of FIG. 2 based on the results of the sleep type classification of these users 110. If the profile 501 includes "#insufficient sleep," the profile output generation unit 24 can also include average data for people of the same gender and age as reference data in the profile configuration 1202. For example, the profile output generation unit 24 includes reference data such as "This is about one hour shorter than the average for women of the same age" in the description 1201 based on the profile vectors of "gender" and "age" in the "demographics data."

[0101] (Recommended Action Determination Unit) The recommended action determination unit 25 functions as a control unit that controls the server 2, and determines a recommended action to be recommended to the user 110 based on the acquired user data. The recommended action determination unit 25 also generates original data of content called recommended action data related to the determined recommended action.

[0102] Examples of recommended actions are as follows: ・Experience a cognitive behavioral therapy program using an application to improve depressive symptoms and insomnia. ・Visit a specialized medical institution that suits the user 110's symptoms. ・Receive simple online counseling. ・Learn through other content to acquire correct sleep knowledge. ・Set a target time to maintain a regular sleep schedule. ・Use sleep restriction, a sleep improvement method that sets bedtime as sleep time and gets out of bed when not sleepy. ・Create a sleep-friendly environment in terms of light, sound, and temperature, such as prohibiting smartphone use before falling asleep. ・Make sure the user adheres to recommended times for caffeine, baths, tobacco, etc., and prohibited times for caffeine, baths, tobacco, etc., calculated backwards from bedtime. ・Implement a sleep schedule program, a program based on the sleep schedule method. ・Implement a program based on progressive muscle relaxation, a method of relaxing the body by consciously repeating the tension and relaxation of specific muscles. ・Perform light exercise during the day.

[0103] The recommended behavior determination unit 25 can also refer to the generated profile when generating recommended behavior data. For example, when generating recommended behavior data, the recommended behavior determination unit 25 refers to the user's 110 acceptance result for the presented profile. In this case, when generating recommended behavior data, the recommended behavior determination unit 25 refers to the selection data, etc., acquired as the acceptance result.

[0104] If the presented profile is accepted by the user 110, the recommended action determination unit 25 generates recommended action data based on the profile accepted as the profile by the user 110. If the presented profile is not accepted by the user 110, the recommended action determination unit 25 generates, as the recommended action data, recommended actions that are most likely to be presented to the user 110 or specific recommended action data that recommends predetermined recommended actions.

[0105] An example of generation of recommended behavior data by the recommended behavior determination unit 25 will be described below with reference to Fig. 3. The following description will be given on the assumption that the profile 501 in Fig. 2 has been accepted by the user 110. In this case, when generating recommended behavior data, the recommended behavior determination unit 25 inputs the profile 501 into the second trained model 282 to generate recommended behavior data 601 that recommends a sleep schedule program or the like.

[0106] For example, the recommended behavior determination unit 25 recommends the implementation of a sleep schedule program that is centered around the following sleep schedule method and is implemented over several weeks using a digital device alone or through a counselor such as a medical professional 30.

[0107] Specifically, the user 110 first sets a target wake-up time. The user 110 then calculates the bedtime by subtracting the amount of sleep required from the set target wake-up time. The user 110 then adjusts the bedtime while monitoring the sleep efficiency. Over several weeks, the user 110, with the help of a counselor, finds the bedtime and wake-up time that allow them to maintain an appropriate sleep rhythm and sleep efficiency.

[0108] Furthermore, when recommending the implementation of a sleep schedule program, the recommended behavior determination unit 25 can also recommend sleep hygiene education to learn correct knowledge about sleep, the establishment of correct sleep habits, relaxation techniques such as progressive muscle relaxation, etc. The recommended behavior determination unit 25 may generate recommended behavior data recommending the implementation of a sleep schedule program as specific recommended behavior data generated when a neutral profile is set.

[0109] Furthermore, the recommended behavior determination unit 25 can also generate recommended behavior data that optionally recommends visiting a medical institution, in addition to the recommendation to implement a sleep schedule program, which is the main recommended behavior, as the recommended behavior data 601. The recommended behavior determination unit 25 generates the recommended behavior data including, for example, a reservation site, a link to a homepage, contact information, etc., so that the user can visit a specialized medical institution online or in person.

[0110] 3, the recommended action determination unit 25 refers to the second trained model 282 when generating the recommended action data 601. However, the recommended action determination unit 25 can also refer to a manually updated model such as a second database 284 that has been created manually or the like in advance when generating the recommended action data 601. The second database 284 is a table or the like in which profiles and recommended actions are associated in advance.

[0111] When the second trained model 282 or the second database 284 is used to generate recommended behavior data, the recommended behavior determination unit 25 can also select whether to use the second trained model 282 or the second database 284 to generate the recommended behavior data depending on the generated profile.

[0112] Furthermore, when generating recommended behavior data, the recommended behavior determination unit 25 can also differentiate between recommended behaviors when the generated profile is based on a mental illness and recommended behaviors when the generated profile is based on a sleep problem.

[0113] For example, if the generated profile is based on a mental illness, the recommended behavior determination unit 25 generates recommended behavior data for a mental illness that recommends at least one of implementing an application in accordance with medical therapy and counseling.Also, if the generated profile is based on a sleep problem, the recommended behavior determination unit 25 generates recommended behavior data for a sleep problem that recommends implementing a sleep schedule program.

[0114] Specifically, when the generated profile is a mental illness such as a moderate tendency to depression, such as "#feeling uneasy" or "#mental disorder," the recommended behavior determination unit 25 selects to use the second database 284 to generate recommended behavior data. This is because the recommended behavior according to the profile of a mental illness such as a moderate tendency to depression is more in line with the second database 284 than with the second trained model 282.

[0115] The recommended behavior determination unit 25 then generates recommended behavior data that recommends the practice of an application or counseling in accordance with therapies implemented by psychiatrists or psychologists, such as a cognitive behavioral therapy program or a behavioral activation program, because the profile generated based on the answers to questionnaires or the like regarding the mental state of the user 110 suggests that the cause of the sleep-related problems of the user 110 is a mental disorder.

[0116] When the generated profile is a serious illness such as depression or a sleep disorder, such as "#depression tendency" or "#insomnia tendency," the recommended behavior determination unit 25 selects to use the second database 284 to generate recommended behavior data. This is because the recommended behavior according to the profile of a serious illness such as a sleep disorder is more in line with the second database 284 than with the second trained model 282.

[0117] The recommended behavior determination unit 25 then generates recommended behavior data recommending face-to-face counseling. This is because, when the generated profile indicates a serious illness such as depression or a sleep disorder, there are limits to how much improvement in sleep can be achieved through systematic measures such as implementing an application in accordance with the therapy provided by a psychiatrist or psychologist.

[0118] When the generated profile is sleep apnea syndrome, the recommended behavior determination unit 25 selects to use the second database 284 to generate recommended behavior data. This is because the recommended behavior corresponding to the profile of sleep apnea syndrome is more in line with the second database 284 than with the second trained model 282.

[0119] The recommended behavior determination unit 25 then generates recommended behavior data recommending receiving a service that prescribes CPAP (Continuous Positive Airway Pressure) therapy or visiting a medical institution, because the profile based on the wearable data suggests sleep apnea syndrome.

[0120] As in the above example, if the user 110 is further suspected of having any of various diseases, the recommended action determination unit 25 generates recommended action data that recommends receiving a service that can treat the disease or visiting a medical institution.

[0121] When the generated profile has bad sleeping habits such as "#Excessive Caffeine," "#Not Enough Sunlight," "#Lack of Exercise," or "#Difference Between Weekdays and Holidays," the recommended behavior determination unit 25 selects to use the second trained model 282 to generate recommended behavior data. This is because the recommended behavior corresponding to a profile with bad sleeping habits is more in line with the second trained model 282 than with the second database 284.

[0122] The recommended behavior determination unit 25 then generates recommended behavior data that basically recommends a sleep schedule program. However, the recommended behavior determination unit 25 can also adjust part of the content of the sleep schedule program, such as increasing the frequency of presenting advice about bad sleep habits of the user 110 within the sleep schedule program.

[0123] For example, if the profile is "#Excessive caffeine intake," the recommended behavior determination unit 25 generates recommended behavior data including a guideline for the maximum daily caffeine amount and advice to limit intake after the evening.

[0124] The recommended action determination unit 25 can also change the distribution, frequency, and method of communicating the appropriate recommended action data using the second trained model 282.

[0125] The recommended behavior determination unit 25 can also generate recommended behavior data that recommends a combination of a cognitive behavioral therapy program and a behavioral activation program that involve setting behavioral goals.

[0126] When the generated profile is a sleep problem such as "#slight insomnia," "#tendency to insomnia," or "#insomnia," the recommended behavior determination unit 25 selects to use the second trained model 282 to generate recommended behavior data. This is because the recommended behavior according to the sleep problem profile is more in line with the second trained model 282 than with the second database 284.

[0127] The recommended behavior determination unit 25 then generates recommended behavior data that basically recommends a sleep schedule program. However, if working conditions, child-rearing, or other family relationships may create an environment that contributes to sleep disorders, making it difficult to implement a sleep schedule program, the recommended behavior determination unit 25 generates recommended behavior data that has been adjusted to recommend only less burdensome programs, such as sleep education.

[0128] (Content Output Generator) The content output generator 26 functions as a controller that controls the server 2, and generates content that is output content for presenting the generated recommended action data to the user 110 based on the result of the recommended action determination unit 25. For example, the content output generator 26 generates image data 60 including recommended action data 601, as shown in Fig. 3. The content output generator 26 can also include in the content the profile and user data features that are the basis for selecting the recommended action data 601, as well as explanations thereof.

[0129] (1-3. Configuration of Information Processing System According to First Embodiment) An example of the configuration of the information processing system 100 according to the first embodiment will be described with reference to Fig. 8. Fig. 8 is a block diagram showing an example of the configuration of the information processing system according to the first embodiment.

[0130] (User Terminal) As shown in FIG. 8, the user terminal 12 may also include a communication unit 15, a storage unit 16, and a processing unit 17 in addition to the above-mentioned units.

[0131] The communication unit 15 is a network interface card (NIC), a network interface controller, or the like. The communication unit 15 is connected to a network via a wired or wireless connection, and transmits and receives various data to and from other devices via the network. For example, the communication unit 15 receives wearable data from the sensor 11 and receives profiles and content from the server 2. The communication unit 15 also transmits user data to the server 2.

[0132] The storage unit 16 is a semiconductor memory device such as a RAM or a flash memory, or a storage device such as a hard disk or an optical disk, etc. The storage unit 16 stores various data such as an application program 161.

[0133] The application program 161 is a program related to an application for improving, for example, depressive symptoms or insomnia. Specifically, the application program 161 is a program related to an application in accordance with a therapy such as a cognitive behavioral therapy program or a behavioral activation program implemented by a medical professional such as a psychiatrist or psychologist.

[0134] The processing unit 17 functions as the above-mentioned control unit and executes various processes. For example, the processing unit 17 executes the application program 161 to provide the user 110 with an application of a cognitive behavioral therapy program in which a menu showing the contents of the cognitive behavioral therapy is presented on the presentation unit 14.

[0135] (Server) As shown in FIG. 8, the server 2 may also include a communication unit 27, a storage unit 28, and a processing unit 29 in addition to the above-mentioned units.

[0136] The communication unit 27 is a NIC, a network interface controller, or the like. The communication unit 27 is connected to a network via a wired or wireless connection, and transmits and receives various data to and from other devices, etc. via the network. For example, the communication unit 27 receives user data from the user terminal 12. The communication unit 27 transmits a profile and content to the user terminal 12. The communication unit 27 also transmits anonymously processed information, which is information in which personal information anonymized by a processing unit 29 (described later) is associated with the profile and content, to the analysis terminal 7.

[0137] The storage unit 28 is a semiconductor memory element such as a RAM or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 28 stores various data such as a first trained model 281, a second trained model 282, a first database 283, and a second database 284.

[0138] The processing unit 29 functions as the control unit described above, and also generates anonymous processed information by anonymizing profiles and content.

[0139] (Medical Worker Terminal) The medical worker terminal 3 includes a communication unit 31, a user interface unit 32, and a storage unit 33. An example of information stored in the storage unit 33, such as a RAM, is medical information 331. The medical information 331 includes, for example, medical record information of the user 110, and is used by the medical worker 30 to diagnose the user 110, etc.

[0140] The communication unit 31 is a NIC, a network interface controller, etc. The communication unit 31 communicates with other devices, etc. For example, the communication unit 31 receives profiles and content from the server 2.

[0141] The user interface unit 32 functions as a control unit, and also accepts operations of the medical worker terminal 3 by the medical worker 30, presents information to the medical worker 30, etc. The control unit is a CPU, an MPU, or the like that executes a program stored in the storage unit 33. The control unit may also be an integrated circuit such as an ASIC or an FPGA.

[0142] For example, the user interface unit 32 presents result information from the server 2, which then updates the medical information 331, such as by adding medical record information. Intervention by the medical professional 30 is also possible. For example, the medical professional 30 may generate individual content customized to suit the user 110. The individual content is transmitted to the user terminal 12 by the communication unit 31 and presented to the user 110.

[0143] (Community Member Terminal) The community member terminal 4 includes a communication unit 41 and a user interface unit 42. The communication unit 41 is a NIC, a network interface controller, or the like. The communication unit 41 communicates with other devices, etc. For example, the communication unit 41 receives profiles and content from the server 2.

[0144] The user interface unit 42 functions as a control unit, accepts operations of the community member terminal 4 by the members 40, and presents information to the members 40. The control unit is a CPU, an MPU, or the like that executes a program stored in a storage device such as a RAM. The control unit may also be an integrated circuit such as an ASIC or an FPGA.

[0145] For example, the user interface unit 42 presents profiles and content from the server 2. The profiles and content are also shared with the members 40.

[0146] (Life Insurance / Health Insurance Terminal) The life insurance / health insurance terminal 5 includes a communication unit 51, an analysis unit 52, and a storage unit 53. An example of information stored in the storage unit 53, such as a RAM, is customer / employee information 531. The customer / employee information 531 includes information related to the user 110's life insurance, health insurance, etc.

[0147] The communication unit 51 is a NIC, a network interface controller, or the like. The communication unit 51 communicates with the server 2 and receives, for example, profiles and content from the server 2. The communication unit 51 transmits the identified insurance premium and insurance premium / recommendation information recommending the identified insurance premium to the user terminal 12. The insurance premium / recommendation information is presented by the presentation unit 14 of the user terminal 12.

[0148] The analysis unit 52 analyzes the profile and content to determine insurance premiums and rewards. The determination may involve the work or judgment of an employee of the life insurance company or health insurance company. The insurance premium may be reduced or changed to a limited plan.

[0149] (Service Provider Terminal) The service provider terminal 6 includes a communication unit 61 and a user interface unit 62 .

[0150] The communication unit 61 is a NIC, a network interface controller, or the like. The communication unit 61 communicates with other devices, etc. For example, the communication unit 61 transmits product / service information, which is information about products and services, to the server 2. Products for which information is transmitted from the service provider terminal 6 to the server 2 include, for example, health foods and health equipment. Services for which information is transmitted from the service provider terminal 6 to the server 2 include, for example, applications that can be executed on the user terminal 12.

[0151] The user interface unit 62 functions as a control unit, accepts input of product / service information to be associated with profiles and content, and generates product / service information. The control unit is a CPU, MPU, or the like that executes programs stored in a storage device such as RAM. The control unit may also be an integrated circuit such as an ASIC or FPGA.

[0152] The user interface unit 62 receives input of conditions indicating the association between indicators shown in profiles or content and products or services, and generates information including these conditions as product / service information.

[0153] (Analysis Terminal) The analysis terminal 7 includes a communication unit 74, an analysis unit 75, and a user interface unit 76. The communication unit 74 is a NIC, a network interface controller, etc. The communication unit 74 receives anonymously processed information from the server 2.

[0154] The analysis unit 75 functions as a control unit and performs data analysis based on the anonymously processed information. The analysis may involve the work, judgment, etc. of company employees. The control unit is a CPU, MPU, etc. that executes a program stored in a storage device such as RAM. The control unit may also be an integrated circuit such as an ASIC or FPGA.

[0155] The user interface unit 76 presents information related to data analysis. Examples of analysis include analysis of user demographics for products such as health foods and health equipment, and data analysis for clinical development. Anonymously processed information can be used for a variety of services, such as marketing analysis by manufacturers, analysis of the proportion, age, and gender of users with specific symptoms, and symptom monitoring for patients taking specific medications.

[0156] (1-4. Information Processing Procedure According to First Embodiment) (Profile Generation and Presentation Procedure) An example of an information processing procedure by the server 2 will be described with reference to Figs. 9 and 10. First, an example of a profile generation and presentation procedure by the server 2 will be described with reference to Fig. 9. Fig. 9 is a flowchart showing an example of a profile generation and presentation procedure.

[0157] In step S1, the preprocessing unit 21 acquires user data. For example, the preprocessing unit 21 acquires the user data through reception of the user data from the communication unit 15 of the user terminal 12 by the communication unit 27. In step S2, the profiling unit 23 performs profiling based on the user data. For example, the profiling unit 23 generates raw profile data 80 based on the feature amount 70 of the user data in FIG. 5 .

[0158] In step S3, the profile output generation unit 24 generates, based on the user data, a profile that is a candidate for the profile of the user 110. For example, the profile output generation unit 24 generates image data 50 including the profile 501 shown in Fig. 2 based on the original data 80 of the profile generated based on the feature amount 70 of the user data shown in Fig. 5.

[0159] In step S4, the profile output generation unit 24 presents the generated profile to the user 110. For example, the communication unit 27 transmits image data 50 including the profile 501 to the communication unit 15 of the user terminal 12. The presentation unit 14 of the user terminal 12 displays the image data 50. As a result, the profile output generation unit 24 presents the image data 50 to the user 110.

[0160] In step S5, if the user 110 accepts and selects one of the presented profiles (step S5; Yes), the profile output generation unit 24 sets the selected profile as the profile of the user 110. For example, if the button 503 in Fig. 2 is selected, that is, if the selection data indicates that the profile 501 has been accepted by the user 110, the profile output generation unit 24 sets the profile of the user 110 to the profile 501.

[0161] In step S5, if the user 110 selects not to accept any of the presented profiles (step S5; No), the profile output generation unit 24 sets the profile of the user 110 to a neutral profile. For example, if the button 504 in Fig. 2 is selected, that is, if the selection data indicates that the profile 501 is not to be accepted, the profile output generation unit 24 sets the profile of the user 110 to a neutral profile.

[0162] (Content Generation and Presentation Procedure) Next, an example of a content generation and presentation procedure by the server 2 will be described with reference to Fig. 10. Fig. 10 is a flowchart showing an example of a content generation and presentation procedure. Step S11 is the same as step S1, and therefore description thereof will be omitted.

[0163] In step S12, if the profiling process such as setting the profile of the user 110 has been performed (step S12; Yes), the process proceeds to step S13.

[0164] In step S12, if the processing of each step in the flowchart related to profiling in Fig. 9 has not been completed (step S12; No), the process proceeds to step S14. In step S14, the server 2 shifts the processing to each step in the flowchart related to profiling in Fig. 9. The server 2 first completes the processing of each step, and then moves to the processing of determining a recommended action in step S13.

[0165] In step S13, the recommended action determination unit 25 determines, based on the acquired user data, a recommended action to be recommended to the user 110. Then, the recommended action determination unit 25 generates recommended action data related to the determined recommended action.

[0166] For example, the recommended behavior determination unit 25 determines that the recommended behaviors to be recommended to the user 110 are a sleep schedule program and a consultation at a specialized medical institution based on the profile 501 of Fig. 2 generated based on user data. Then, the recommended behavior determination unit 25 generates recommended behavior data 601 of Fig. 3 that recommends these recommended behaviors.

[0167] In step S15, the content output generation unit 26 generates content based on the determination result by the recommended behavior determination unit 25. For example, the content output generation unit 26 generates, as content, image data 60 including a sleep schedule program determined as a recommended behavior to be recommended to the user 110 and the recommended behavior data 601 of Fig. 3 recommending a consultation at a specialized medical institution.

[0168] In step S16, the content output generation unit 26 presents the generated content to the user 110. For example, the communication unit 27 transmits image data 60 including the recommended behavior data 601 to the communication unit 15 of the user terminal 12. The presentation unit 14 of the user terminal 12 displays the image data 60. As a result, the content output generation unit 26 presents the image data 60 to the user 110.

[0169] (2. Second Embodiment) The server 2 can also predict the quality of sleep of the user 110. An example of the configuration of the server 2A that constitutes the information processing system 100A according to the second embodiment will be described below with reference to Fig. 11. Fig. 11 is a block diagram showing an example of the configuration of a user terminal and a server according to the second embodiment.

[0170] (Server) The server 2A includes a learning unit 22A, a profiling unit 23A, and a recommended behavior determination unit 25A instead of the learning unit 22, the profiling unit 23, and the recommended behavior determination unit 25. The server 2A also includes a third trained model 285. The third trained model 285 is a model that outputs sleep quality data in response to input feature amount data.

[0171] (Learning Unit) The learning unit 22A further trains the third trained model 285. "Sleep quality data" is data related to sleep quality. The sleep quality data may be data based on feature data of a group or an individual corresponding to the user data. For example, the sleep quality data is user data related to sleep quality, such as a subjective sleep efficiency value among the subjective index data or a subjective sleep index.

[0172] For example, the learning unit 22A generates a third trained model 285 by training a model based on training data in which feature data is an input variable and sleep quality data is an output variable (correct answer).

[0173] An example of the feature amount data and sleep quality data will be described below with reference to Fig. 12. Fig. 12 is a diagram showing an example of the feature amount data and sleep quality data. The learning unit 22A can use therapy history data 293, wearable data 294, sleep diary data 295, subjective index data 296, behavior record data 297, demographics data 298, etc. to calculate the feature amount data 291 and the sleep quality data 292.

[0174] For example, the learning unit 22A can use, as feature amount data, feature amounts calculated based on the therapy history data 293, the wearable data 294, the sleep diary data 295, the subjective index data 296, and the behavioral record data 297. Furthermore, the learning unit 22A can use, as sleep quality data 292, a predicted value of sleep quality calculated based on the wearable data 294, the sleep diary data 295, the subjective index data 296, and the behavioral record data 297.

[0175] Specifically, the learning unit 22A generates a third trained model 285 using the feature calculated from the wearable data 294 as the feature data as the input variable and the predicted value of tonight's sleep quality calculated from the subjective index data 296 as the sleep quality data as the output variable.

[0176] (Profiling Unit) Based on the user data, the profiling unit 23A further generates a predicted value of the sleep quality of the user 110. For example, the profiling unit 23A generates a predicted value of the user 110's sleep tonight as of daytime by inputting features calculated from the acquired wearable data 294 into the third trained model 285.

[0177] (Recommended Action Determination Unit) When the predicted value of the user 110's sleep quality is lower than the ideal value, the recommended action determination unit 25A generates, in addition to the recommended action data, recommended effective action data that recommends effective actions for raising the predicted value (such as the predicted value of the user 110's sleep quality tonight as of daytime) to the ideal value. The "ideal value" is a predetermined ideal value for the user 110's sleep quality.

[0178] (3. Third Embodiment) When presenting content, the server 2 can also adjust the wording of the text related to the recommended behavior data according to the generated profile. An example of the configuration of a server 2B constituting an information processing system 100B according to the third embodiment will be described below with reference to Fig. 13. Fig. 13 is a block diagram showing an example of the configuration of a user terminal and a server according to the third embodiment.

[0179] (Server) The server 2B includes a content output generation unit 26B instead of the content output generation unit 26. The server 2B also includes a third database 286. The third database 286 is, for example, a table in which profiles with a high tendency toward specific diseases or personality traits are associated with specific prohibited actions.

[0180] (Content Output Generation Unit) When presenting content, the content output generation unit 26B adjusts the wording of the text related to the recommended behavior data according to the generated profile. For example, depending on the type of disease indicated in the profile, the content output generation unit 26B adjusts the wording of the content according to the profile so that more sympathetic wording is used.

[0181] Specifically, the content output generation unit 26B refers to the third database 286 and adjusts the wording of the content so as not to convey specific taboos or prohibited matters that should not be conveyed to the user 110.

[0182] (4. Other Embodiments) The processes according to the embodiments can be implemented in various different forms other than the above-described embodiments.

[0183] Of the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically using known methods. In addition, the information including the processing procedures, specific names, various data, and parameters shown in the above documents and drawings can be changed as desired unless otherwise specified. The various information shown in each drawing is not limited to the information shown in the drawings.

[0184] The components of each device shown in the figure are conceptual functional components and do not necessarily have to be physically configured as shown. In other words, the specific form of distribution and integration of each device is not limited to that shown in the figure, and all or part of the configuration can be functionally or physically distributed or integrated in any unit depending on various loads and usage conditions. For example, in the configuration of the information processing device that performs the above-mentioned information processing, the server 2 and the user terminal 12 can be integrated.

[0185] The above-described embodiments and modifications can be combined as appropriate within the scope of not causing any contradiction in the processing content.

[0186] The effects described in this specification are merely examples and are not limiting, and other effects may also be present.

[0187] (5. Effects of the Information Processing Method According to the Present Disclosure) As described above, in the information processing method according to the present disclosure, a computer (in this embodiment, the server 2) acquires user data related to the user's sleep, generates classification data in which the user's sleep type is classified into one of the sleep types based on the acquired user data, and recommended action data related to recommended actions that are actions recommended for improving the user's sleep, and presents the generated classification data and recommended action data to the user.

[0188] In this way, by presenting the user with the classification data in addition to the recommended behavior data, the user can roughly understand their own sleep state and the causes of their sleep disorders from the presented classification data, and therefore understand that the recommended behavior corresponds to the level and causes of their sleep disorders. For example, the user can understand that the sleep schedule program in the recommended behavior data corresponds to the causes of their sleep disorders based on the sleep problems listed in the classification data, rather than mental illness.

[0189] Therefore, the user can be convinced that the recommended action suits him / her. Therefore, according to the information processing method, the user can accept the recommended action with a sense of conviction.

[0190] The information processing method refers to the generated classification data when generating the recommended action data. This makes the recommended actions in the recommended action data more suited to the profile, so that the information processing method allows the user to accept the recommended actions with a greater sense of satisfaction.

[0191] The information processing method refers to the user's acceptance of the presented classification data when generating the recommended action data. This makes the recommended action data responsive to the user's acceptance, so the information processing method allows the user to accept the recommended action with a greater sense of satisfaction.

[0192] The information processing method further acquires the selection data accepted by the UI when the user selects whether to accept the presented classification data, and references the acquired selection data as the acceptance result when generating the recommended action data. Thus, according to the information processing method, the recommended action data can be made to correspond to the user's acceptance result simply by the UI accepting the selection of whether to accept or not.

[0193] In the information processing method, if the presented classification data is accepted by the user, recommended action data is generated based on the accepted classification data accepted by the user as classification data, and if the presented classification data is not accepted by the user, specific recommended action data that recommends a recommended action that is presented to the user with the highest probability or a predetermined recommended action is generated as the recommended action data.

[0194] As a result, even if the classification data is not accepted by the user, the recommended action that is most likely to be presented to the user or a predetermined recommended action can be recommended, and therefore, according to the information processing method, the recommended action can be accepted with a sense of conviction by a wide range of users.

[0195] When generating recommended behavior data, the information processing method differentiates the recommended behavior when the generated classification data is based on a mental illness from the recommended behavior when the generated classification data is based on a sleep problem. This allows recommended behaviors tailored to each user with different sleep problems to be presented, so the information processing method can present effective recommended behaviors even when effective responses vary greatly depending on the cause of the sleep disorder.

[0196] In the information processing method, when the generated classification data is based on a mental illness, recommended behavior data for a mental illness is generated, which recommends at least one of implementing an application in accordance with medical therapy and counseling, and when the generated classification data is based on a sleep problem, recommended behavior data for a sleep problem is generated, which recommends implementing a sleep schedule program, which is a program based on a sleep scheduling method, which is a means of adjusting sleep rhythms.

[0197] This generates recommended behavior data regarding recommended behaviors that are effective when the cause of sleep disorders is mental illness or sleep problems, and therefore, this information processing method can be accepted with a sense of satisfaction by users who suffer from sleep disorders caused by these factors.

[0198] The information processing method adjusts the wording of the recommended behavior data according to the generated classification data when presenting the recommended behavior data. According to the information processing method, even if there are taboo matters that should not be communicated to the user, the wording of the wording can be adjusted to be more sympathetic to the user without conveying the taboo matters, thereby enabling communication with the user aimed at improving the user's health condition.

[0199] The information processing method generates classified data by referring to the degree to which one or more items related to sleep state apply to the user. This allows the generated classified data to refer to the degree to which each item, such as the level and cause of the user's sleep disorder, applies to the user, making it possible to generate classified data that is more in line with the recommended behavior data.

[0200] When generating the classification data, the information processing method refers to the degree of applicability of at least one of the following items to the user: factors of the user's sleep problems, level of insomnia, sleep worries, and demographic data related to the user's own characteristics and demographics. This generates classification data that references the degree of applicability of each item that categorizes the user's level and factors of sleep disorders in detail, making it possible to generate classification data that is more in line with the recommended behavior data.

[0201] When generating the classified data, the information processing method generates a configuration of the classified data including information on the user data that is the basis for classifying the user's sleep type into one of the sleep types. This identifies the level and causes of the user's sleep disorder, and the information processing device can provide the user with factors, opportunities, etc. that will help the user achieve better sleep.

[0202] The information processing method selects a character corresponding to the classification data when generating the classification data. This anthropomorphizes the classification data, and the information processing method makes it possible for the user to feel familiar with and interested in the classification data and sleep.

[0203] When generating the classification data, the information processing method selects at least one of a character that represents the user's sleep type and a character that is suitable as a partner for the user in improving the user's sleep.

[0204] In this way, when a character representing the user's sleep type is selected, the user becomes aware of their sleep type, and the information processing method allows the user to confront the level and causes of their sleep disorder. Furthermore, when a character suitable as a partner for the user is selected, the user can work with the character to improve their sleep, and the information processing method encourages the user to improve their sleep.

[0205] In the information processing method, when generating classification data, the acquired user data is input into a first trained model (in the embodiment, the first trained model 281) that outputs classification data in which sleep types are classified into one of several types depending on the input of feature data related to sleep features.

[0206] Specific items such as "insomnia" in the category "causes of sleep problems" that make up the classification data are more in line with the first trained model than with the database (in this embodiment, the first database 283), and therefore, according to the information processing method, specific items can be generated with high accuracy.

[0207] In the information processing method, when generating the classification data, the acquired user data is input to a first trained model that outputs the user's classification data as classification data in response to the input of the user's feature amount data as feature amount data. In this way, the classification data is generated based on the first trained model personalized for the user, and the information processing method can generate classification data in which the user's sleep type is classified with higher accuracy.

[0208] In the information processing method, when generating recommended behavior data, the generated classification data is input into a second trained model (in the embodiment, the second trained model 282) that outputs recommended behavior data regarding recommended behaviors, which are behaviors recommended for improving sleep, in response to input of classification data in which the sleep type is classified into one of the types.

[0209] Since specific classification data such as "having bad sleep habits" or "sleep problems" conforms more to the second trained model than to the database (in this embodiment, the second database 284), the information processing method can generate specific classification data with high accuracy.

[0210] When a second trained model or a database in which classification data and recommended behavior data are associated is used to generate recommended behavior data, the information processing method selects whether to use the second trained model or the database (in this embodiment, the second database 284) to generate the recommended behavior data depending on the generated classification data.

[0211] Depending on the classification data, the recommended action data generated based on the classification data will differ in whether it conforms to the second learned model or the database, and therefore, according to the information processing method, appropriate means can be used to generate recommended action data that is more in line with the classification data.

[0212] The information processing method inputs the acquired user data into a third trained model (in this embodiment, the third trained model 285) that outputs sleep quality data related to sleep quality in response to input feature data related to sleep features, further generates a predicted value of the user's sleep quality, and if the generated predicted value is lower than the ideal value, generates recommended effective behavior data, separate from the recommended behavior data, that recommends effective behaviors for raising the predicted value to the ideal value.

[0213] This generates a predicted value of the user's sleep quality for the current day, etc., based on the user's daytime behavior, etc., indicated by wearable data, etc., acquired by a wearable device, etc., capable of acquiring the user's physiological and activity data throughout the day. Therefore, according to the information processing method, it is possible to inform the user whether the user's daytime behavior is in a state suitable for sleep.

[0214] Furthermore, if the generated predicted value is lower than the ideal value, effective actions for raising the predicted value to the ideal value are recommended to the user, in addition to the recommended action data. Therefore, according to the information processing method, even if the user's daytime actions are not in a state suitable for sleep, recommended actions can be recommended to the user to make the user's daytime actions suitable for sleep.

[0215] (6. Hardware Configuration) Information devices such as the server 2 according to each of the above-described embodiments are realized by a computer 1000 configured as shown in Fig. 14. Fig. 14 is a hardware configuration diagram showing an example of a computer that realizes the functions of a server according to the embodiments. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM 1300, a HDD (Hard Disk Drive) 1400, a communication interface 1500, and an input / output interface 1600. The components of the computer 1000 are connected by a bus 1050.

[0216] The CPU 1100 operates and controls each unit based on programs stored in the ROM 1300 or the HDD 1400. The CPU 1100 loads the programs stored in the ROM 1300 or the HDD 1400 into the RAM 1200 and executes processing corresponding to the various programs.

[0217] The ROM 1300 stores boot programs such as a Basic Input Output System (BIOS) that is executed by the CPU 1100 when the computer 1000 starts up, and programs that depend on the hardware of the computer 1000 .

[0218] HDD 1400 is a computer-readable recording medium that non-temporarily records programs executed by CPU 1100 and data used by such programs. Specifically, HDD 1400 is a recording medium that records an information processing program according to the present disclosure, which is an example of program data 1450.

[0219] The communication interface 1500 is an interface for connecting the computer 1000 to an external network 1550 (such as the Internet). The CPU 1100 receives data from other devices and transmits data generated by the CPU 1100 to other devices via the communication interface 1500.

[0220] The input / output interface 1600 is an interface for connecting the input / output device 1650 and the computer 1000. The CPU 1100 receives data from input devices such as a keyboard and a mouse via the input / output interface 1600. The CPU 1100 transmits data to output devices such as a display, a speaker, and a printer via the input / output interface 1600. The input / output interface 1600 can also function as a media interface for reading a program recorded on a predetermined recording medium.

[0221] The media may be optical recording media such as DVDs (Digital Versatile Discs) and PDs (Phase Change Rewritable Discs), magneto-optical recording media such as MOs (Magneto-Optical disks), tape media, magnetic recording media, or semiconductor memories.

[0222] When the computer 1000 functions as the server 2 according to the embodiment, the CPU 1100 of the computer 1000 executes an information processing program loaded onto the RAM 1200 to realize the functions of the control units such as the preprocessing unit 21 to the processing unit 29 in Figure 8. The information processing program according to the present disclosure and data in the storage unit 28 are stored in the HDD 1400.

[0223] The CPU 1100 reads and executes the program data 1450 from the HDD 1400. However, as another example, the CPU 1100 can also obtain these programs from other devices via an external network 1550.

[0224] (7. Supplementary Information) The present technology can also be configured as follows. (1) An information processing method including: a computer acquiring user data related to a user's sleep; generating, based on the acquired user data, classification data in which the user's sleep type is classified into any one of the sleep types, and recommended action data related to recommended actions that are actions recommended to improve the user's sleep; and presenting the generated classification data and the recommended action data to the user. (2) The information processing method according to (1), wherein the generated classification data is referenced when generating the recommended action data. (3) The information processing method according to (2), wherein a result of the user's acceptance of the presented classification data is referenced when generating the recommended action data. (4) The information processing method according to (3), wherein selection data accepted by a UI for the user's selection of whether to accept the presented classification data is further acquired; and wherein the acquired selection data is referenced as the acceptance result when generating the recommended action data. (5) The information processing method according to (3) or (4), wherein, if the presented classification data is accepted by the user, the recommended action data is generated based on accepted classification data accepted by the user as the classification data, and if the presented classification data is not accepted by the user, the recommended action data is generated as the recommended action data, which is the recommended action that is most likely to be presented to the user or specific recommended action data that recommends a predetermined recommended action. (6) The information processing method according to any one of (2) to (5), wherein, when generating the recommended action data, the recommended action when the generated classification data is based on a mental disorder is differentiated from the recommended action when the generated classification data is based on a sleep problem.(7) The information processing method according to (6), wherein, if the generated classification data is based on the mental disorder, recommended behavior data for a mental disorder is generated, recommending at least one of implementation of an application in accordance with medical therapy and counseling when generating the recommended behavior data, and if the generated classification data is based on the sleep problem, recommended behavior data for a sleep problem is generated, recommending implementation of a sleep schedule program, which is a program based on a sleep scheduling method that is a means of adjusting sleep rhythms when generating the recommended behavior data. (8) The information processing method according to any one of (2) to (7), wherein, when presenting the recommended behavior data, wording of text related to the recommended behavior data is adjusted according to the generated classification data. (9) The information processing method according to any one of (1) to (8), wherein, when generating the classification data, the degree to which one or more items related to sleep states apply to the user is referred to. (10) The information processing method according to (9), wherein, when generating the classification data, the degree to which the user applies to at least one of the following items is referenced: factors of the user's sleep problems, level of insomnia, sleep worries, and demographic data related to the user's own characteristics and demographics. (11) The information processing method according to any one of (1) to (10), wherein, when generating the classification data, a classification data configuration is generated that includes information on user data that is the basis for classifying the user's sleep type into any of the sleep types. (12) The information processing method according to any one of (1) to (11), wherein, when generating the classification data, a character according to the profile is selected. (13) The information processing method according to (12), wherein, when generating the classification data, at least one of a character representing the user's sleep type and a character suitable as a partner for the user in implementing measures to improve the user's sleep is selected.(14) The information processing method according to any one of (1) to (13), wherein, when generating the classification data, the acquired user data is input to a first trained model that outputs classification data in which a sleep type is classified into one of the types in response to input of feature amount data related to sleep features. (15) The information processing method according to (14), wherein, when generating the classification data, the acquired user data is input to the first trained model that outputs classification data of the user as the classification data in response to input of feature amount data of the user as the feature amount data. (16) The information processing method according to any one of (2) to (8), wherein, when generating the recommended behavior data, the generated classification data is input to a second trained model that outputs recommended behavior data that is a behavior recommended for improving the sleep in response to input of classification data in which a sleep type is classified into one of the types. (17) The information processing method according to any one of (1) to (17), further comprising: inputting the acquired user data into a third trained model that outputs sleep quality data related to sleep quality in response to input of feature data related to sleep features, and generating, when the generated predicted value is lower than an ideal value, recommended effective behavior data that recommends effective behaviors for raising the predicted value to the ideal value, separate from the recommended behavior data.(19) An information processing device comprising: an acquisition unit that acquires user data related to a user's sleep; a generation unit that generates, based on the acquired user data, classification data in which the user's sleep type is classified into any of the sleep types and recommended action data on recommended actions that are actions that are recommended for improving the user's sleep; and a presentation unit that presents the generated classification data and the recommended action data to the user. (20) An information processing program that causes a computer to function as an information processing device comprising: an acquisition unit that acquires user data related to a user's sleep; a generation unit that generates, based on the acquired user data, classification data in which the user's sleep type is classified into any of the sleep types and recommended action data on recommended actions that are actions that are recommended for improving the user's sleep; and a presentation unit that presents the generated classification data and the recommended action data to the user.

[0225] 2 Server 12 User terminal 13 Acquisition unit 14 Presentation unit 21 Preprocessing unit 22 Learning unit 23 Profiling unit 24 Profile output generation unit 25 Recommended action determination unit 26 Content output generation unit 281 First trained model 282 Second trained model 283 First database 284 Second database

Claims

1. An information processing method comprising: a computer acquiring user data relating to a user's sleep; generating, based on the acquired user data, classification data in which the user's sleep type is classified into any one of the sleep types, and recommended action data relating to recommended actions that are actions recommended for improving the user's sleep; and presenting the generated classification data and recommended action data to the user.

2. The information processing method according to claim 1, wherein the generated classification data is referenced when the recommended action data is generated.

3. The information processing method according to claim 2, wherein when generating the recommended action data, a result of the user's acceptance of the presented classification data is referred to.

4. The information processing method of claim 3, wherein the user's choice of whether to accept the presented classification data further acquires selected data accepted by a UI (User Interface), and when generating the recommended behavior data, the acquired selected data is referenced as the acceptance result.

5. The information processing method of claim 3, wherein, if the presented classification data is accepted by the user, the recommended action data is generated based on the accepted classification data accepted by the user as the classification data, and if the presented classification data is not accepted by the user, the recommended action data is generated as the recommended action data, which is the recommended action that is most likely to be presented to the user, or specific recommended action data that recommends a predetermined recommended action.

6. The information processing method according to claim 2, wherein, when generating the recommended behavior data, the recommended behavior when the generated classification data is based on a mental illness is differentiated from the recommended behavior when the generated classification data is based on a sleep problem.

7. The information processing method of claim 6, wherein, when the generated classification data is based on the mental disorder, recommended behavior data for the mental disorder is generated, which recommends at least one of implementing an application in accordance with medical therapy and counseling, and when the generated classification data is based on the sleep problem, recommended behavior data for the sleep problem is generated, which recommends implementing a sleep schedule program, which is a program based on a sleep scheduling method that is a means of adjusting sleep rhythms.

8. The information processing method according to claim 2, further comprising: adjusting a textual expression relating to the recommended action data in accordance with the generated classification data when presenting the recommended action data.

9. The information processing method according to claim 1, wherein when generating the classification data, the degree to which one or more items relating to sleep states apply to the user is referenced.

10. An information processing method as described in claim 9, wherein when generating the classification data, the degree to which the user fits at least one of the following items is referenced: factors of the user's sleep problems, level of insomnia, sleep worries, and demographic data regarding the user's own characteristics and demographics.

11. An information processing method as described in claim 1, wherein when generating the classification data, a classification data configuration is generated that includes information about user data that is the basis for classifying the user's sleep type into one of the sleep types among the user data.

12. The information processing method according to claim 1, wherein when generating the classification data, a character is selected according to the classification data.

13. The information processing method of claim 12, wherein when generating the classification data, at least one of a character representing the user's sleep type and a character suitable as a partner for the user in improving the user's sleep is selected.

14. The information processing method of claim 1, wherein, when generating the classification data, the acquired user data is input to a first trained model that outputs classification data in which the sleep type is classified into one of the types depending on the input of feature data related to sleep features.

15. The information processing method of claim 14, wherein, when generating the classification data, the acquired user data is input to the first trained model, which outputs the user's classification data as the classification data in response to input of the user's feature data as the feature data.

16. The information processing method of claim 2, wherein when generating the recommended behavior data, the generated classification data is input to a second trained model that outputs recommended behavior data regarding recommended behaviors that are behaviors recommended for improving sleep in response to input of classification data in which sleep-related types are classified into one of the types.

17. The information processing method of claim 16, wherein, when the second trained model or a database in which the classification data and the recommended action data are associated is used to generate the recommended action data, the information processing method selects whether to use the second trained model or the database to generate the recommended action data depending on the generated classification data.

18. An information processing method as described in claim 1, further comprising inputting the acquired user data into a third trained model that outputs sleep quality data related to sleep quality in response to input of feature data related to sleep features, and further generating a predicted value of the user's sleep quality, and if the generated predicted value is lower than an ideal value, generating recommended effective action data, separate from the recommended action data, that recommends effective actions to raise the predicted value to the ideal value.

19. An information processing device comprising: an acquisition unit that acquires user data related to a user's sleep; a generation unit that generates, based on the acquired user data, classification data in which the user's sleep type is classified into any one of the sleep types, and recommended action data related to recommended actions that are actions recommended for improving the user's sleep; and a presentation unit that presents the generated classification data and recommended action data to the user.

20. An information processing program for causing a computer to function as an information processing device comprising: an acquisition unit that acquires user data related to a user's sleep; a generation unit that generates, based on the acquired user data, classification data in which the user's sleep type is classified into any one of the sleep types, and recommended action data related to recommended actions that are actions recommended for improving the user's sleep; and a presentation unit that presents the generated classification data and recommended action data to the user.

Images