Device and method

Abstract

The purpose of the present disclosure is to provide a device and a method for easily deriving the sleep quality of a user.　In a server 100 of the present disclosure, a log acquisition unit 103 acquires log information of a portable terminal 200 (user terminal) which is used by a user. On the basis of the log information, an estimation unit 108 estimates a user situation (sleep time, terminal use situation immediately before bedtime, sleep environment, circadian rhythm, etc.) that affects sleeping. Then, the estimation unit 108 estimates (corresponding to deriving) the sleep quality of the user on the basis of the user situation.

Description

Apparatus and method

[0001] The present invention relates to an apparatus and method for processing data in a terminal device.

[0002] Patent document 1 describes a prediction device that includes an acquisition unit that acquires activity data related to physical activity, a fragmentation degree calculation unit that calculates a fragmentation degree that represents the degree of activity fluctuation based on the activity data, a prediction unit that predicts the sleep quality corresponding to the calculated fragmentation degree using a prediction model that represents the correspondence between the fragmentation degree and sleep quality, and an output unit that outputs the predicted sleep quality.

[0003] JP 2024-30128 A

[0004] The invention described in Patent Document 1 has the problem that the user needs to wear an activity meter, which is cumbersome. Furthermore, activity meters are special measuring devices and are not commonly used.

[0005] Therefore, in order to solve the above problem, an object of the present disclosure is to provide an apparatus and method for easily estimating a user's sleep quality.

[0006] The device disclosed herein includes a log information acquisition unit that acquires log information of a user terminal used by a user, a situation derivation unit that estimates a user situation that affects sleep based on the log information, and a quality estimation unit that estimates the quality of the user's sleep based on the user situation.

[0007] According to the present invention, the quality of a user's sleep can be easily estimated.

[0008] FIG. 1 is a diagram illustrating an overview of the learning process of the sleep quality estimation model 107 of the present disclosure. FIG. 2 is a diagram illustrating a system configuration including a server 100 equipped with the sleep quality estimation model 107 of the present disclosure. FIG. 3 is a diagram illustrating the functional configuration of the server 100 of the present disclosure. FIG. 4 is a diagram illustrating the functional configuration of a learning device 350 of the sleep time estimation model 106 of the present disclosure. FIG. 5 is a diagram illustrating the functional configuration of the learning device 300. FIG. 6 is a flowchart illustrating the generation process of the sleep time estimation model 106 of the learning device 350. FIG. 7 is a flowchart illustrating the generation process of the sleep quality estimation model 107 of the learning device 300. FIG. 8 is a flowchart illustrating the sleep quality estimation process in the server 100. FIG. 9 is a diagram illustrating a method for estimating sleep time feature quantities related to sleep time. FIG. 10 is a diagram illustrating the generation of behavioral feature quantities indicating the usage status of a mobile device of a user immediately before going to bed from log information in collected data D1. FIG. 11 is a diagram illustrating the process of having the sleep quality estimation model 107 estimate sleep quality based on the generated behavioral feature quantities. A diagram illustrating the learning process of a usage status estimation model. FIG. 13 is a diagram showing a process for estimating sleep quality using a sleep quality estimation model by utilizing estimated sleep environment information. FIG. 14 is a diagram showing a learning process for the sleep environment estimation model. FIG. 15 is an explanatory diagram for midnight awakening. FIG. 16 is an explanatory diagram for a lighting environment. FIG. 17 is an explanatory diagram for a noise environment. FIG. 18 is an explanatory diagram for autonomic nervous activity. FIG. 19 is an explanatory diagram for a thermal environment. FIG. 20 is an explanatory diagram for circadian rhythm estimation. FIG. 21 is a diagram showing a method for acquiring feature quantities used in circadian rhythms. FIG. 22 is a diagram showing an example of the hardware configuration of a server 100, a learning device 300, and a learning device 350 according to an embodiment of the present disclosure.

[0009] The present disclosure will be described with reference to the accompanying drawings. Whenever possible, the same parts are designated by the same reference numerals and redundant description will be omitted.

[0010] 1 is a diagram showing an overview of the learning process of the sleep quality estimation model 107 of the present disclosure. As shown in the diagram, the sleep quality estimation model 107 is trained based on feature quantities that explain factors that affect sleep quality, or estimation results of factors that affect sleep quality, obtained from collected data D1 that can be automatically acquired from a mobile device used by a user, and training data obtained from collected data D2.

[0011] The automatically obtainable collected data D1 is log information, such as a device log, location information, application log, and contract information, but may also include other data (e.g., weather information). These data are not necessarily required. The collected data D1 is an example for training the sleep quality estimation model 107 of the present disclosure. In the present disclosure, the collected data D1 is data that can be automatically collected from a mobile device, but may also include other data input by the user from the mobile device.

[0012] From the collected data D1, feature quantities and the like that explain factors that affect the quality of sleep are derived.

[0013] Examples of features that explain factors that affect sleep quality include: (1) sleep duration (wake-up time, bedtime), (2) using a mobile phone before going to bed, (3) sleep environment, (4) circadian rhythm, (5) daytime sleepiness, (6) busyness of daily life (work, housework, childcare, nursing care, etc.), (7) health problems (cough, pain), and (8) demographic data (attribute information such as gender, etc.). Of the above, (1) to (4) are factors (behavioral features) that particularly affect sleep quality, and at least one of them is used to train the sleep quality estimation model 107, and estimation is performed using these factors (behavioral features).

[0014] The training data may be collected data prepared in advance, such as a questionnaire. Alternatively, heart rate data obtained by a wearable device may be used as the training data.

[0015] Using the sleep quality estimation model 107 trained in this way, the quality of sleep of the user can be estimated from the collected data D1 of the mobile terminal 200 used by the user.

[0016] 2 is a diagram showing a system configuration including a server 100 equipped with a sleep quality estimation model 107 of the present disclosure. As shown in the figure, this system includes the server 100, a mobile terminal 200, a learning device 300, and a learning device 350. Although the figure shows one mobile terminal 200, there are actually multiple mobile terminals 200.

[0017] The learning device 300 learns the sleep quality estimation model 107 from collected data D1 and teacher data (collected data D2) that can be automatically acquired for learning from multiple mobile devices (which may include the mobile device 200). The sleep quality estimation model 107 is stored in the server 100.

[0018] The learning device 350 learns the sleep time estimation model 106 from collected data D1 that can be automatically acquired for learning from multiple mobile devices and collected data D2 that is teacher data. This sleep time estimation model 106 is stored in the server 100.

[0019] Thereafter, the server 100 acquires automatically obtainable collected data D1 from the mobile terminal 200 of the user to be estimated, and inputs the acquired collected data D1 into the sleep quality estimation model 107 to estimate the quality of sleep.

[0020] 3 is a diagram showing the functional configuration of the server 100 of the present disclosure. The server 100 is a device that estimates sleep quality from log information using a sleep quality estimation model 107, etc. As shown in the figure, the server 100 includes a user input unit 101, a demographic information storage DB 102, a log acquisition unit 103, a log storage DB 104, a weather information DB 105, a sleep time estimation model 106, a sleep quality estimation model 107, an estimation unit 108, and an output unit 109.

[0021] The user input unit 101 is a part that receives contract information (including attribute information) from the mobile terminal 200, which is part of the collected data D1.

[0022] The demographic information storage DB 102 is a section that stores demographic data. This demographic data includes contract information (including user attribute information). The attribute information includes the user's age, gender, family structure, family ages, and place of residence (address and location). The contract information indicates the details of the contract between the mobile terminal 200 and the telecommunications carrier (contract plan: fees, amount of packets that can be communicated, etc.). This information is input by the user via the user input unit 101.

[0023] The log acquisition unit 103 acquires log information from the collected data D1 automatically acquired by the mobile device 200. The log information includes a device log, location information, and an application log. The device log is log information acquired by the mobile device 200, such as acceleration (including a step count log), device tilt, gyro, ambient illuminance (illuminance sensor), screen on / off count, unlock count, application launch count, earphone connection count (or time), remaining battery level, microlog (number of inputs or time), Wi-Fi connection count, whether a charger is attached, screen brightness, and battery temperature. The application log indicates the application used, the number of launches, and the launch time. The location information is information obtained by GPS, but may also be location information obtained using base station information of the mobile device 200. This location information may be coordinate information or information indicating the distance or proximity from home.

[0024] The log storage DB 104 is a part that stores the log information acquired by the log acquisition unit 103 .

[0025] The weather information DB 105 is a section for storing weather information. The weather information is information that associates regions, temperatures, humidity, solar radiation, etc. This information is information that is acquired from an external weather information server.

[0026] The model storage unit stores a sleep time estimation model 106 and a sleep quality estimation model 107. The sleep time estimation model 106 is a learning model that estimates at least one of bedtime, wake-up time, and sleep duration from a device log, an application log, and location information (distance from home). The sleep quality estimation model 107 is a learning model that inputs examples of factors that affect sleep duration or other sleep quality and estimates sleep quality based on the input.

[0027] The estimation unit 108 calculates various features from log information and other information (contract information, weather information, etc.), inputs these features into the sleep time estimation model 106 and the sleep quality estimation model 107 stored in the model storage unit, and obtains the output to estimate the quality of sleep.

[0028] The output unit 109 is a part that outputs information indicating the quality of sleep estimated by the estimation unit 108 to the mobile terminal 200 .

[0029] 4 is a diagram showing the functional configuration of a learning device 350 for the sleep time estimation model 106 of the present disclosure. As shown in the figure, the learning device 350 includes a correct answer data storage DB 351, a log storage DB 352, a log acquisition unit 353, a feature calculation unit 354, and a sleep time model construction unit 355.

[0030] The correct data storage DB 351 stores correct data related to sleep input by the user. The correct data is data obtained from a questionnaire including bedtime, wake-up time, and sleep duration. Note that the correct data may be not only the questionnaire data but also sleep duration obtained by a wearable device or a mobile device.

[0031] The log storage DB 352 is a part that stores log information (terminal log, location information, application log) from the collected data D1. The log information is associated with time information when the information was generated.

[0032] The log acquisition unit 353 is a part that acquires the supervised data and log information stored in the supervised data storage DB 351 and the log storage DB 352 .

[0033] The feature calculation unit 354 is a part that calculates feature amounts from log information. In the present disclosure, feature amounts are information that indicate log information per unit time. For example, the feature calculation unit 354 performs a binary classification of sleep / wake in 10-minute increments and calculates the result as feature amounts. The feature calculation unit 354 may also perform the binary classification of sleep / wake per unit time (10 minutes in this example) using a machine learning model based on training data. This machine learning model is trained using log information as an explanatory variable and bedtime, wake-up time, and sleep duration as objective variables. The bedtime, wake-up time, sleep duration, etc. are training data obtained on the terminal as described above.

[0034] The sleep time model construction unit 355 is a part that learns the sleep time estimation model 106 based on the feature amounts. The sleep time model construction unit 355 learns by using the feature amounts obtained from the log information (device log, location information, and application log) as explanatory variables and the sleep time feature amounts that are the correct answer data (bedtime, wake-up time, and sleep duration) as objective variables.

[0035] Next, the learning device 300 will be described. Fig. 5 is a diagram showing the functional configuration of the learning device 300. As shown in the figure, the learning device 300 includes a correct answer data storage DB 301, a log storage DB 302, a weather information DB 303, a demographic information storage DB 304, a log acquisition unit 305, a sleep time feature calculation unit 306, a sleep time feature storage unit 307, a user situation calculation unit 308, a user situation feature storage unit 309, and a sleep quality model construction unit 310.

[0036] The correct data storage DB 301 is a part that stores correct data about sleep quality. This correct data is data based on the collected data D2, and is objective or subjective information that indicates the user's sleep quality, such as a user questionnaire or heart rate data.

[0037] Log storage DB 302 is a part that stores log information (terminal logs, location information, application logs) of multiple mobile terminals (which may include mobile terminal 200).

[0038] The weather information DB 303 is a part that stores weather information for each region and time.

[0039] The demographic information storage DB 304 is a section that stores demographic data (contract information, attribute information, etc.).

[0040] The log storage DB 302, the weather information DB 303, and the demographic information storage DB 304 may be the log storage DB 104, the weather information DB 105, and the demographic information storage DB 102 in FIG.

[0041] The log acquisition unit 305 is a part that acquires collected data D1 including various data such as log information from the log storage DB 302, the weather information DB 303, and the demographic information storage DB 304.

[0042] The sleep time feature amount calculation unit 306 is a unit that calculates sleep time feature amounts, and is functionally the same as the feature amount calculation unit 354. That is, the sleep time feature amount calculation unit 306 calculates feature amounts indicating sleep / wakefulness per unit time based on the log information, inputs the calculated feature amounts to the sleep time estimation model 106, and acquires the results. The sleep time feature amount calculation unit 306 is a unit that calculates the bedtime, wake-up time, and sleep duration as sleep time feature amounts.

[0043] The sleep duration feature amount storage unit 307 is a part that stores the bedtime, wake-up time, and sleep duration calculated as sleep duration feature amounts.

[0044] The user status calculation unit 308 is a part that calculates a feature quantity indicating a user status that affects sleep based on the bedtime, wake-up time, and sleep duration, as well as the collected data D1 acquired by the log acquisition unit 305. In the present disclosure, the feature quantity indicating a user status that affects sleep refers to at least one of the usage status of the mobile device before going to bed, the sleeping environment, and the circadian rhythm.

[0045] The user situation feature amount storage unit 309 is a part that stores feature amounts that indicate the user situation that affects sleep.

[0046] The sleep quality model construction unit 310 is a part that learns the sleep quality estimation model 107 based on the features indicating the user situation that affect sleep stored in the user situation feature storage unit 309 and the collected data D2 stored in the correct data storage DB 301.

[0047] Next, the process of generating the sleep time estimation model 106 in the learning device 350 configured as described above will be described. FIG. 6 is a flowchart showing the process of generating the sleep time estimation model 106 in the learning device 350. As shown in the figure, the log acquisition unit 353 acquires log information (device log, location information, application log) and correct answer data (bedtime, wake-up time, sleep duration) (S101). The feature calculation unit 354 calculates feature amounts (sleep / wake) per unit time based on the log information (S102). The sleep time model construction unit 355 constructs (learns) the sleep time estimation model 106 that estimates sleep time using the feature amounts and correct answer data (S103). The constructed sleep time estimation model 106 is stored in the server 100.

[0048] Next, the process of generating the sleep quality estimation model 107 will be described. Fig. 7 is a flowchart showing the process of generating the sleep quality estimation model 107 of the learning device 300. The log acquisition unit 305 acquires log information and correct answer data (S201). The sleep time feature amount calculation unit 306 calculates the feature amount (sleep / wake) per unit time based on the log information (S202).

[0049] Then, the sleep time feature amount calculation unit 306 inputs the feature amount per unit time into the sleep time estimation model 106 and executes estimation of the sleep time etc. using the sleep time estimation model 106 (S203). The sleep time feature amount calculation unit 306 acquires (calculates) the wake-up time, bedtime and sleep time, which are the estimation results, as feature amounts related to the sleep time (S204).

[0050] Next, the user status calculation unit 308 calculates the device usage status feature amount before going to bed based on the log information (S205). The user status calculation unit 308 also calculates the sleeping environment feature amount based on the log information and weather information (S206). The user status calculation unit 308 also calculates the circadian rhythm feature amount based on the location information in the log information and the weather information (S207).

[0051] Then, the sleep quality model construction unit 310 constructs (learns) a sleep quality estimation model 107 that estimates sleep quality using these features (device usage status feature, sleep environment feature, circadian rhythm feature) and the correct answer data (S208). The constructed sleep quality estimation model 107 is stored in the server 100.

[0052] Next, the sleep quality estimation process will be described. Fig. 8 is a flowchart showing the sleep quality estimation process in the server 100. The user input unit 101 and the log acquisition unit 103 read demographic information, log information, etc. stored in the demographic information storage DB 102 and the log storage DB 104 (S301). Here, the demographic information and log information have been acquired and stored in advance in the demographic information storage DB 102 and the log storage DB 104.

[0053] The estimation unit 108 estimates the sleep time using the sleep time estimation model 106. That is, the estimation unit 108 converts the log information acquired from the log storage DB 104 into features and inputs them into the sleep time estimation model 106, and obtains the wake-up time, bedtime, and sleep time from the sleep time estimation model 106 (S302).

[0054] The estimation unit 108 also calculates usage status features of the mobile device 200 before going to bed from the information stored in the demographic information storage DB 102, the log storage DB 104, and the weather information DB 105 (S303). The estimation unit 108 also calculates sleep environment features from the information stored in the demographic information storage DB 102, the log storage DB 104, and the weather information DB 105 (S304). The estimation unit 108 also calculates circadian rhythm features from the information stored in the log storage DB 104 and the weather information DB 105 (S305).

[0055] The estimation unit 108 inputs the thus calculated sleep time, usage status feature, sleep environment feature, and circadian rhythm feature into the sleep quality estimation model 107, thereby estimating the quality of sleep (S306).

[0056] The output unit 109 outputs the estimated result to the mobile terminal 200 (S307).

[0057] Next, a method for calculating each feature amount in the present disclosure will be described. FIG. 9 is a diagram illustrating a method for estimating sleep time feature amounts related to sleep time. FIG. 9( a) shows that feature amounts generated from collected data D1, which is log information, are input into a sleep time estimation model 106 to estimate sleep time, such as bedtime, wake-up time, and sleep time. In the present disclosure, this sleep time is input into another estimation model (e.g., a sleep quality estimation model 107). In the present disclosure, the collected data D1 includes at least one of acceleration, device tilt, gyro, screen on / off, charging status, earphone connection, illuminance, number of steps, number of unlocks (which may be an unlocked state), number of app launches (which may be an launched state), Wi-Fi connection (status or number of times), and location information (distance from home) per unit time. However, other information may also be included, or the collected data D1 may include only log information other than the collected data listed here.

[0058] As described above, this sleep time estimation model 106 is trained in the learning device 350 using features based on the collected data D1, which is log information, as explanatory variables and the bedtime, wake-up time, and sleep duration specified by the user as objective variables.

[0059] FIG. 9B illustrates an overview of how the feature calculation unit 354 and the sleep time feature calculation unit 306 generate sleep time-related features. As shown in the figure, the collected data D1 is used to binary classify sleep and wakefulness per unit time based on at least one of the following: acceleration, device tilt, gyro, screen on / off, charging status, earphone connection, illuminance, number of steps, number of unlocks (which may be the unlocked state), number of app launches (which may be the launched state), Wi-Fi connection (status or number of times), and location information (distance from home). This classification is defined in advance based on rules. For example, if acceleration is detected or the screen is on, it may be determined to be sleep. Alternatively, a binary classification of sleep / wake per unit time (e.g., 10 minutes) may be performed using a machine learning model based on training data. This machine learning model is trained using log information as an explanatory variable and bedtime, wake-up time, and sleep duration as target variables. The bedtime, wake-up time, sleep duration, etc. are training data obtained on the device as described above.

[0060] In the present disclosure, the feature amount calculation unit 354 and the sleep time feature amount calculation unit 306 classify the state St into sleep or wakefulness in 10-minute increments. The feature amount calculation unit 354 and the sleep time feature amount calculation unit 306 calculate the binary-classified value as a feature amount related to sleep time. Based on this feature amount, a model for estimating the bedtime d, wake-up time d, etc. can be learned, and estimation using the model is also possible.

[0061] During learning, the sleep time model construction unit 355 uses these feature amounts as explanatory variables and learns using the sleep time, wake-up time, and bedtime as objective variables to construct the sleep time estimation model 106. Furthermore, when estimating sleep time using the sleep time estimation model 106, the feature amounts can be input into the sleep time estimation model 106 to determine the feature amounts related to sleep time.

[0062] 10 to 12 are diagrams for explaining how to derive the usage status of a mobile terminal (mobile terminal 200 at the time of estimation) before going to bed and how to estimate the quality of sleep based on the usage status.

[0063] 10 is a diagram showing that the feature calculation unit 354 and the sleep time feature calculation unit 306 have generated behavioral features indicating the user's mobile device usage status immediately before going to bed from the log information in the collected data D1. As shown in the figure, behavioral features such as the launch time or number of launches for each app category (games, communication tools, entertainment, etc.) immediately before going to bed are calculated from the app usage history. Here, the log information in the collected data D1 includes the app usage history, screen on / off status, charger connection status, and earphone connection status. However, at least one of these may be used, and the log information is not limited to these, and other information may also be used.

[0064] From the screen on / off history, a behavioral feature, that is, the screen on time immediately before going to bed, is obtained.

[0065] From the history of attaching the charger and connecting the earphones, the percentage of time the earphones / charger are connected (per unit of time just before going to bed) and the number of times they are connected can be calculated.

[0066] In this figure, "just before going to bed" is, for example, a time period based on the bedtime estimated by the sleep time estimation model 106, and indicates, for example, a period of one hour from bedtime.

[0067] 11 is a diagram showing a process of estimating sleep quality based on the generated behavioral features in the sleep quality estimation model 107. As described above, in addition to the usage status, features of other factors that affect sleep quality may also be input to the sleep quality estimation model 107, and the estimation process may be performed.

[0068] Alternatively, a mobile device usage estimation model for before going to bed may be used to utilize the estimation results. In this case, similarly to the above, the feature quantities of other factors that affect the quality of sleep may also be input to the sleep quality estimation model 107, and the estimation process may be performed.

[0069] The usage of mobile devices immediately before going to bed has a significant impact on the quality of sleep, and passively acquired log information is used to indirectly predict whether sleep is being disrupted due to excessive use of mobile devices before going to bed.

[0070] 12 is a diagram showing the learning process of the usage status estimation model. As shown in the figure, the usage status estimation model is learned based on the generated behavioral features and training data. This training signal is information based on a questionnaire or the like, for example, information indicating whether or not a mobile device was being operated based on the questionnaire. Incidentally, whether or not a mobile device was being operated may also be indicated based on heart rate data. Generally, when the heart rate is fast, the user is awake and exhibiting some kind of behavior, and in the present disclosure, this is considered to be operating a mobile device.

[0071] The use status estimation model is not essential, and the generated behavioral feature may be input directly to the sleep quality estimation model 107 to estimate the quality of the sleep, without estimating the use status using the use status estimation model. In this case, the sleep quality estimation model 107 needs to be trained using the behavioral feature.

[0072] 13 and 14 are diagrams illustrating the process of estimating a sleeping environment. As shown in the figures, the sleeping environment is estimated by calculating or extracting behavioral features related to five indices related to the sleeping environment using log information that can be passively acquired by the mobile device and meteorological information (public information). This calculation and extraction process of behavioral features is performed by the user state calculation unit 308 during learning, and by the estimation unit 108 during estimation.

[0073] 13 is a diagram showing a process for estimating sleep quality using a sleep quality estimation model based on estimated sleep environment information. As shown in the figure, the sleep environment indicates at least one of the following: determination of midnight awakening, lighting environment, noise environment, autonomic nervous activity, and temperature and humidity environment. In the present disclosure, behavioral features for each of these sleep environments are calculated and input into the sleep quality estimation model 107 to estimate sleep quality. Alternatively, a sleep environment is estimated using a separately trained sleep environment estimation model, and the sleep quality estimation model 107 is used to estimate sleep quality.

[0074] The midnight awakening determination is a behavioral feature that is a binary classification of sleep or wakefulness per unit time. The lighting environment is a behavioral feature based on a comparison between an ideal lighting environment and the illuminance environment before and after sleep. The noise environment is a behavioral feature based on the determination result of the noise state before and after sleep. The autonomic nervous activity is a behavioral feature based on the determination result of the degree of autonomic nervous activity enhancement before going to bed. The temperature and humidity environment is a behavioral feature based on a comparison between an ideal thermal environment and the temperature and humidity before and after sleep.

[0075] 14 is a diagram illustrating a learning process of the sleep environment estimation model. In the present disclosure, the sleep environment estimation model is learned using at least one of the determination of nocturnal awakening, lighting environment, noise environment, autonomic nervous activity, and temperature and humidity environment as explanatory variables, and training data obtained through a questionnaire as a response variable.

[0076] The sleep environment estimation model is not essential, and the generated behavioral features may be input directly to the sleep quality estimation model 107 to estimate the quality of the sleep environment without estimating the sleep environment using the sleep environment estimation model. In this case, the sleep quality estimation model 107 needs to have been trained using the behavioral features.

[0077] Next, detailed processing of the determination of midnight awakening regarding the sleep environment, the lighting environment, the noise environment, the autonomic nerve activity, and the temperature and humidity environment will be described.

[0078] 15(a) is a diagram schematically illustrating a process for generating behavioral features related to mid-awakening from collected data D1 of a mobile device. As shown in the figure, collected data D1 includes an acceleration / step count log, microlog, illuminance sensor log, and screen on / off count, and the duration of mid-awakening and the number of mid-awakenings are obtained from this collected data D1. In this disclosure, features indicating the duration of mid-awakening and the number of mid-awakenings are generated from statistics obtained from the results of determining whether the user is in a sleep state or a wakeful state per unit time.

[0079] FIG. 15(b) is a diagram illustrating how sleep / wake status is determined by unit time. As shown in the diagram, sleep / wake status is determined in 10-minute increments (unit times) from bedtime to wakeup time. In the present disclosure, as shown in FIG. 15(c), a mid-wake is determined between bedtime and wakeup time. This determination is made based on collected data D1. For example, it is assumed that the user operates a mobile device upon waking up. In this case, the mobile device can record acceleration logs, etc. In this manner, the log information can be used to determine the duration of mid-wake and the number of wakeups. Note that the number and duration of mid-wake events between bedtime and wakeup time are determined. The time is the time when the device log is detected. The number of times is the number of times when the device log is detected at a predetermined interval (e.g., at least 10 minutes or more). FIG. 16(a) is a diagram schematically illustrating the process of generating behavioral features related to the lighting environment from the illuminance sensor log of the mobile device's collected data D1. As shown in the figure, the illuminance average / standard deviation for each epoch, the degree of change in illuminance before going to bed and just before going to bed, the degree of change in illuminance when waking up and just before waking up, the number of times the illuminance exceeded the specified value for each epoch, and the score for each epoch can be calculated from the illuminance sensor log collected on the mobile device.

[0080] That is, each sleep epoch is defined based on the estimated bedtime, wake-up time, and mid-sleep awakening time, and the illumination environment for each epoch is indexed. In the present disclosure, an epoch refers to a time unit for making sleep-related judgments.

[0081] FIG. 16(b) shows the specified illuminance for each epoch. In the present disclosure, an epoch refers to a specific time period, such as a predetermined time period before going to bed, a predetermined time period just before going to bed, a time period during sleep, a time period during awakening, or a time period just before waking up. The specified illuminance is a predetermined illuminance, which is a reference illuminance (lx: lux) for each epoch. Note that in the figure, it is written in parentheses that "0.3-11 is optimal." This indicates the optimal illuminance during sleep and awakening. Based on the specified illuminance for each epoch, the number of times the illuminance exceeds the specified value for each epoch is calculated.

[0082] The degree of change in illuminance indicates the difference in illuminance (lux), and is calculated by subtracting the illuminance just before going to sleep from the illuminance just before going to sleep, but other methods of indicating the degree of change may also be used. The same applies to the degree of change in illuminance between waking up and just before waking up. The epoch score is the sum of these values. In this way, by comparing the ideal lighting environment with the illuminance before and after sleep, the sleep environment can be estimated. The epoch score does not necessarily have to be the sum of scores within one day; for example, it can be calculated by taking the average over multiple days for each epoch.

[0083] 17(a) is a diagram showing a schematic diagram of the process of generating behavioral features related to a noise environment from data D1 collected by a mobile device. As shown in the figure, the average and standard deviation of noise during sleep, the maximum and minimum noise during sleep, the number of times the noise exceeded the specified value for each epoch, and the score for each epoch can be calculated from the micrologs collected by the mobile device.

[0084] That is, each sleep epoch is defined based on the estimated bedtime, wake-up time, and mid-sleep awakening time, and the noise environment for each epoch is indexed.

[0085] FIG. 17(b) shows the prescribed noise for each epoch. In the present disclosure, an epoch refers to a specific time period, such as a predetermined time period before going to bed, a predetermined time period just before going to bed, a time period during sleep, a time period after waking up, or a time period just before waking up. The prescribed noise is a predetermined noise, which is a reference noise (dBA: decibel) for each epoch. In the present disclosure, the prescribed noise is defined as a noise during sleep, for example, 50 dBA. In this manner, the sleep environment can be estimated by determining the noise conditions before and after sleep.

[0086] 18(a) is a diagram illustrating the process of generating behavioral features related to autonomic nervous activity from the illuminance sensor log, app usage history, screen on / off count, and screen brightness value among the collected data D1 of a mobile device. As shown in the figure, the illuminance sensor log, app usage history, screen on / off count, and screen brightness value collected on the mobile device are used to calculate the percentage of smartphone on time one hour before bedtime, app usage time one hour before bedtime, average illuminance / standard deviation one hour before bedtime, and average screen brightness value / standard deviation one hour before bedtime. In other words, the degree of autonomic nervous activity is indexed from the smartphone usage status and illuminance before bedtime.

[0087] 18(b) is a schematic diagram showing that the following data were acquired during a predetermined time (here, one hour) immediately before sleep: the time the smartphone was on one hour before bedtime, the type and time of the app used one hour before bedtime, the ambient illuminance one hour before bedtime, and the brightness of the light received from the smartphone one hour before bedtime. These behavioral features are information that influences the user's autonomic nervous activity, and can be used to determine the degree of enhancement of autonomic nervous activity before bedtime.

[0088] 19(a) is a diagram showing a schematic diagram of the process of generating behavioral features related to the thermal environment from meteorological information and battery temperature in the collected data D1 of the mobile device. As shown in the figure, the temperature average / standard deviation for each epoch, the humidity average / standard deviation for each epoch, the deviation (difference) between the actual temperature and humidity and the optimal temperature and humidity, and the battery temperature average / standard deviation are calculated from the meteorological information during sleep and the temperature and humidity information obtained from the mobile device 200 based on the estimated bedtime, wake-up time, and mid-wake time. The sleep environment is indexed based on the estimated bedtime, wake-up time, and mid-wake time.

[0089] Fig. 19(b) is a diagram showing the optimum temperature and humidity during sleep. These temperatures and humidity are stored in the log storage DB 104. Fig. 19(c) is a diagram showing bedtimes and wake-up times. In the present disclosure, it is assumed that the person is asleep during these times and has not woken up during sleep.

[0090] In this way, the sleeping environment can be estimated by comparing the ideal thermal environment (temperature and humidity) with the thermal environment (temperature and humidity) before and after sleep.

[0091] 20 is an explanatory diagram of circadian rhythm estimation by the estimation unit 108 and the user state calculation unit 308. Circadian rhythm is a rhythm exhibited by the human body with a cycle of approximately 24 hours. This rhythm affects sleep and its quality, and is controlled by exposure to light.

[0092] FIG. 20( a) is a diagram illustrating the generation of behavioral features related to circadian rhythms based on light exposure from wake-up time to bedtime. As shown in the figure, the period from wake-up time to bedtime includes daytime (daytime) when the sun is up and nighttime after the sun has set. Light exposure during the day includes the amount of solar radiation at the location and the intensity of lighting from lighting fixtures, etc. Light exposure at night includes the intensity of lighting from lighting fixtures, etc. This information is acquired from log information and weather information on the mobile device. Note that wake-up time and bedtime are estimated information as described above. Daytime and nighttime are based on predetermined times depending on the season or time of year.

[0093] From this information, behavioral features related to circadian rhythms are generated. In the present disclosure, the behavioral features are the amount of solar radiation at the location visited up to noon (when going out), the amount of solar radiation at the location visited from noon to nighttime (when going out), the intensity of lighting exposed during the day, the intensity of lighting exposed at night, the intensity of lighting exposed just before going to bed, and the intensity of lighting exposed during the day / the intensity of lighting exposed at night. The process of acquiring these behavioral features is performed by the estimation unit 108 during estimation, and by the user state calculation unit 308 during learning.

[0094] FIG. 20(b) is an explanatory diagram illustrating the application of a circadian rhythm estimation model to the above behavioral features to output circadian rhythm estimation results. The circadian rhythm estimation model inputs behavioral features and outputs the estimation results. This circadian rhythm estimation model is a rule-based judgment model that generates behavioral features related to each user's circadian rhythm, such as solar radiation and daytime activity status, from multiple perspectives that are thought to be related to circadian rhythm. For example, in FIG. 20(a), scores are assigned for each perspective (item), such as solar radiation at the location and lighting intensity. For example, solar radiation can be estimated from log information (location) and weather information, and a score is assigned based on this estimate. Generally, ideal light levels are known for lighting environments such as natural light in the morning, indoor lighting during the day, indoor lighting at night, and lighting before bedtime, so scoring can be performed based on these values. For example, if the ideal value for daytime indoor lighting is 500 to 1,000 lux, a gradient score can be calculated, with 0 lux being a score of 0 and 500 lux being a score of 100.

[0095] FIG. 21( a) is a diagram showing a method for obtaining solar radiation estimation data used in the circadian rhythm. As shown in the figure, location information and Activity Recognition API are obtained from the collected data D1 of the mobile device 200, and weather information is also obtained, and the amount of solar radiation is determined from this information. The location information determines whether the user is out and about and the destination, and the Activity Recognition API determines the means of transportation used at the destination. The weather information is used to estimate the amount of solar radiation at the destination. The Activity Recognition API is a function (API) that uses various sensors in the mobile device 200 to determine the user's activity (walking, running, cycling, etc.).

[0096] 21B is an explanatory diagram of estimating lighting intensity. As shown in the figure, behavioral features related to lighting intensity are calculated based on the illuminance sensor log, screen brightness value, and screen on / off status from the collected data D1.

[0097] Next, the effects of the disclosed server 100 will be described. In the server 100 of the present disclosure, the log acquisition unit 103 acquires log information from the mobile terminal 200 (user terminal) used by the user. The estimation unit 108 estimates the user's conditions that affect sleep (such as sleep duration, terminal usage conditions immediately before going to bed, sleep environment, and circadian rhythm) based on the log information. The estimation unit 108 then estimates (corresponding to deriving) the user's sleep quality based on the user's conditions.

[0098] According to this configuration, the quality of a user's sleep can be estimated based on the user's situation based on the log information of the mobile device 200, and the user can estimate the quality of their sleep without wearing any special device. In particular, by using the log information of the mobile device 200, the quality of the user's sleep can be estimated based on the user's daily activities.

[0099] In the present disclosure, the user situation includes at least one of the user's sleep situation (sleep duration), the user's immediate actions immediately before going to bed, the sleep environment in which the user is sleeping, and the user's daytime actions during the day, which affect the quality of the user's sleep.

[0100] The estimation unit 108 derives the user status using a learning model trained with features based on the log information as explanatory variables and the user status as a target variable. For example, if the user status is a sleep status (sleep duration), the estimation unit 108 derives information on sleep duration (sleep duration, wake-up time, bedtime), which is one of the user status, using the sleep duration estimation model 106 as a learning model. Similarly, the estimation unit 108 can derive the mobile phone usage status and the sleep environment using a mobile phone usage status estimation model and a sleep environment estimation model as learning models.

[0101] According to this configuration, the user's status can be easily estimated by using a learning model trained by machine learning. During learning, feature amounts are obtained using log information of mobile devices other than mobile device 200 (which may include mobile device 200), and are used to train sleep time estimation model 106 and sleep quality estimation model 107.

[0102] In the present disclosure, the sleep duration (corresponding to the sleep status) includes at least one of the bedtime, wake-up time, and sleep duration in a predetermined period of time in the past. Such information on the sleep status significantly affects the quality of the user's sleep.

[0103] In addition, in the present disclosure, behavior immediately before going to bed indicates the operation status of the mobile terminal 200 immediately before going to bed. This operation status indicates the operation status (startup time, number of startups) for applications by category. The operation status also indicates the connection status of the mobile terminal 200 with external devices. Operation of the mobile terminal 200 immediately before going to bed significantly affects the quality of sleep of the user. In particular, a user who operates an application is exposed to light emitted from the display screen of the mobile terminal 200, which significantly affects the quality of sleep. Furthermore, depending on the category of the application, the quality of sleep is affected. For example, the impact is significant when the user is concentrating on looking at the display screen, such as when playing a game.

[0104] In the present disclosure, the sleep environment includes at least one of information on whether or not a user is asleep per unit time, lighting environment, noise environment, the operation status of the user terminal, and temperature and humidity environment. For example, the sleep information refers to waking up during sleep. Waking up during sleep is considered to indicate poor quality of sleep. The lighting environment, noise environment, operation status, and temperature and humidity environment also affect the quality of sleep.

[0105] In the present disclosure, the operating conditions as a sleeping environment and the mobile terminal usage conditions before going to bed are generally the same, so only one of them may be adopted.

[0106] In the present disclosure, daytime behavior indicates a situation in which a user is exposed to light. The situation in which the user is exposed to light is at least one of the amount of sunlight or the intensity of lighting. This is for estimating a circadian rhythm. In the present disclosure, the amount of sunlight or the intensity of lighting in daytime behavior depends on the location of the user. In addition, a feature that takes into account the light exposed at night, i.e., the intensity of lighting, may be derived to estimate a circadian rhythm.

[0107] The above-mentioned example factors (5) to (8) (see below) that affect sleep quality can also be used. For example factors (5) to (7), an estimation model for estimating daytime sleepiness, busyness in daily life, and health status can be trained using operation logs, etc., and used. Alternatively, estimation can be performed based on a rule base. These example factors (5) to (8), including demographic data, can be further used to train the sleep quality estimation model 107 and estimate quality. (5) Daytime sleepiness (6) Busyness in daily life (work, housework, childcare, nursing care, etc.) (7) Health problems (cough, pain) (8) Demographic data (attribute information such as gender, etc.). The device and method disclosed herein have the following configuration.

[0108] [1] A device comprising: a log information acquisition unit that acquires log information of a user terminal used by a user; a situation derivation unit that estimates a user situation that affects sleep based on the log information; and a quality estimation unit that estimates the quality of the user's sleep based on the user situation.

[0109] [2] The device according to [1], wherein the situation derivation unit derives the user situation using a learning model trained with features based on log information as explanatory variables and the user situation as a target variable.

[0110] [3] The device according to [1] or [2], wherein the user situation includes at least one of the user's sleep situation, the user's most recent behavior just before going to bed, the sleep environment in which the user is sleeping, and the user's daytime behavior during the day.

[0111] [4] The device according to [3], wherein the sleep status includes at least one of a bedtime, a wake-up time, and a sleep duration in a predetermined past period.

[0112] [5] The device according to [3] or [4], wherein the immediately preceding behavior indicates an operation status on the user terminal immediately before going to bed.

[0113] [6] The device according to [5], wherein the operation status indicates an operation status for an application for each category.

[0114] [7] The device according to any one of [3] to [6], wherein the sleep environment includes at least one of information on whether or not the user is sleeping per unit time, a lighting environment, a noise environment, an operation status of a user terminal, and a temperature and humidity environment.

[0115] [8] The device according to any one of [3] to [7], wherein the daytime behavior indicates a situation in which the user is exposed to light.

[0116] [9] The device according to [8], wherein the light exposure condition is a condition based on at least one of the amount of solar radiation or the intensity of lighting.

[0117]

[10] A method comprising: a log information acquisition step of acquiring operation log information of a user terminal used by a user; a situation estimation step of estimating a user situation that affects sleep based on the operation log information; and a quality derivation step of deriving the quality of the user's sleep based on the user situation.

[0118] The block diagrams used to explain the above embodiments show functional blocks. These functional blocks (components) are realized by any combination of hardware and / or software. Furthermore, the method for realizing each functional block is not particularly limited. That is, each functional block may be realized using a single device that is physically or logically coupled, or may be realized using two or more physically or logically separated devices that are connected directly or indirectly (e.g., via wire, wirelessly, etc.) and these multiple devices. The functional block may also be realized by combining the single device or multiple devices with software.

[0119] Functions include, but are not limited to, judgment, determination, assessment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs transmission is called a transmitting unit or transmitter. As mentioned above, there are no particular limitations on how these functions are implemented.

[0120] For example, the server 100, learning device 300, and learning device 350 according to an embodiment of the present disclosure may function as a computer that performs processing of the sleep quality estimation method or estimation model learning method of the present disclosure. Fig. 22 is a diagram illustrating an example of the hardware configuration of the server 100, learning device 300, and learning device 350 according to an embodiment of the present disclosure. The server 100, learning device 300, and learning device 350 described above may be physically configured as a computer device including a processor 1001, memory 1002, storage 1003, communication device 1004, input device 1005, output device 1006, bus 1007, etc.

[0121] In the following description, the term "device" can be interpreted as a circuit, device, unit, etc. The hardware configurations of server 100, learning device 300, and learning device 350 may be configured to include one or more of the devices shown in the figures, or may be configured to exclude some of the devices.

[0122] Each function of the server 100, learning device 300, and learning device 350 is realized by loading specified software (programs) onto hardware such as the processor 1001 and memory 1002, causing the processor 1001 to perform calculations, control communication via the communication device 1004, and control at least one of reading and writing data in the memory 1002 and storage 1003.

[0123] The processor 1001 controls the entire computer by running, for example, an operating system. The processor 1001 may be configured by a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. For example, the above-mentioned estimation unit 108, feature amount calculation unit 354, etc. may be realized by the processor 1001.

[0124] The processor 1001 also reads programs (program codes), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 into the memory 1002 and executes various processes in accordance with these programs. The programs used are those that cause a computer to execute at least some of the operations described in the above-described embodiments. For example, the estimation unit 108 may be implemented by a control program stored in the memory 1002 and running on the processor 1001, and similar implementations may be made for other functional blocks. While the above-described various processes have been described as being executed by one processor 1001, they may also be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented by one or more chips. The programs may also be transmitted from a network via a telecommunications line.

[0125] The memory 1002 is a computer-readable recording medium and may be configured by, for example, at least one of a read-only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a random access memory (RAM), etc. The memory 1002 may also be called a register, a cache, a main memory (primary storage device), etc. The memory 1002 can store executable programs (program codes), software modules, etc. for implementing the sleep quality estimation method or the estimation model training method according to an embodiment of the present disclosure.

[0126] Storage 1003 is a computer-readable recording medium, and may be composed of at least one of, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray (registered trademark) disk), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy (registered trademark) disk, a magnetic strip, etc. Storage 1003 may also be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, a server, or other appropriate medium including at least one of memory 1002 and storage 1003.

[0127] The communication device 1004 is hardware (transmission / reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also referred to as, for example, a network device, a network controller, a network card, or a communication module. The communication device 1004 may be configured to include a high-frequency switch, a duplexer, a filter, a frequency synthesizer, etc. to realize at least one of frequency division duplex (FDD) and time division duplex (TDD). For example, the above-mentioned user input unit 101, log acquisition unit 103, output unit 109, etc. may be realized by the communication device 1004. The communication device 1004 may be implemented with a transmitter and a receiver that are physically or logically separated from each other.

[0128] The input device 1005 is an input device (e.g., a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that receives input from the outside. The output device 1006 is an output device (e.g., a display, a speaker, an LED lamp, etc.) that outputs to the outside. Note that the input device 1005 and the output device 1006 may be integrated into one device (e.g., a touch panel).

[0129] Furthermore, each device, such as the processor 1001 and the memory 1002, is connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or may be configured using different buses between each device.

[0130] Server 100, learning device 300, and learning device 350 may also be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a field-programmable gate array (FPGA), and some or all of the functional blocks may be realized by the hardware. For example, processor 1001 may be implemented using at least one of these pieces of hardware.

[0131] The notification of information is not limited to the aspects / embodiments described in the present disclosure and may be performed using other methods. For example, the notification of information may be performed by physical layer signaling (e.g., Downlink Control Information (DCI) and Uplink Control Information (UCI)), higher layer signaling (e.g., Radio Resource Control (RRC) signaling, Medium Access Control (MAC) signaling, broadcast information (Master Information Block (MIB) and System Information Block (SIB))), other signals, or a combination thereof. Furthermore, the RRC signaling may be referred to as an RRC message, and may be, for example, an RRC Connection Setup message, an RRC Connection Reconfiguration message, or the like.

[0132] The order of the procedures, sequences, flowcharts, etc. of each aspect / embodiment described in this disclosure may be changed unless it is consistent. For example, the methods described in this disclosure present elements of various steps using an example order, and are not limited to the particular order presented.

[0133] Input and output information may be stored in a specific location (for example, memory) or may be managed using a management table. Input and output information may be overwritten, updated, or added to. Output information may be deleted. Input information may be sent to another device.

[0134] The determination may be made based on a value represented by one bit (0 or 1), a Boolean value (true or false), or a numerical comparison (e.g., comparison with a predetermined value).

[0135] The aspects / embodiments described in this disclosure may be used alone, in combination, or switched depending on the implementation. Notification of predetermined information (e.g., notification that "X is true") is not limited to explicit notification, but may be implicit (e.g., not notifying the predetermined information).

[0136] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0137] Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

[0138] Software, instructions, information, etc. may also be transmitted or received over a transmission medium. For example, if software is transmitted from a website, server, or other remote source using wired technologies (such as coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL)), and / or wireless technologies (such as infrared, microwave), then these wired and / or wireless technologies are included within the definition of transmission media.

[0139] The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0140] Note that terms described in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings. For example, at least one of a channel and a symbol may be a signal (signaling). Furthermore, a signal may be a message. Furthermore, a component carrier (CC) may be called a carrier frequency, a cell, a frequency carrier, etc.

[0141] Furthermore, the information, parameters, etc. described in the present disclosure may be expressed using absolute values, relative values ​​from a predetermined value, or other corresponding information. For example, a radio resource may be indicated by an index.

[0142] The names used for the above-described parameters are not intended to be limiting in any way. Furthermore, the mathematical expressions using these parameters may differ from those explicitly disclosed in this disclosure. The various channels (e.g., PUCCH, PDCCH, etc.) and information elements may be identified by any suitable names, and therefore the various names assigned to these various channels and information elements are not intended to be limiting in any way.

[0143] In this disclosure, the terms "Mobile Station (MS)," "user terminal," "User Equipment (UE)," "terminal," and the like may be used interchangeably.

[0144] A mobile station may also be referred to by those skilled in the art as a subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, or some other suitable terminology.

[0145] As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Determining" and "determining" may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiring (e.g., searching in a table, database, or other data structure), ascertaining, and the like. "Determining" and "determining" may also include receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, accessing (e.g., accessing data in memory), and the like. Furthermore, "judgment" and "decision" can include regarding resolving, selecting, choosing, establishing, comparing, etc. as having been "judged" or "decided." In other words, "judgment" and "decision" can include regarding some action as having been "judged" or "decided." Furthermore, "judgment (decision)" can be interpreted as "assuming," "expecting," "considering," etc.

[0146] The terms "connected," "coupled," or any variation thereof, refer to any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are "connected" or "coupled" to each other. The coupling or connection between elements may be physical, logical, or a combination thereof. For example, "connected" may be read as "access." As used in this disclosure, two elements may be considered to be "connected" or "coupled" to each other using one or more wires, cables, and / or printed electrical connections, as well as electromagnetic energy having wavelengths in the radio frequency range, microwave range, and optical (both visible and invisible) range, as some non-limiting and non-exhaustive examples.

[0147] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0148] As used in this disclosure, any reference to an element using a designation such as "first," "second," etc. does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient method of distinguishing between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed or that the first element must in some way precede the second element.

[0149] When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, similar to the term "comprising." Furthermore, when the term "or" is used in this disclosure, it is not intended to be an exclusive or.

[0150] In this disclosure, where articles are added by translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0151] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may also be interpreted in the same way as "different."

[0152] 100...server, 200...mobile terminal, 300...learning device, 350...learning device, 101...user input unit, 102...mole information storage DB, 103...log acquisition unit, 104...log storage DB, 105...weather information DB, 106...sleep time estimation model, 107...quality estimation model, 108...estimation unit, 109...output unit, 301...correct answer data storage DB, 302...log storage DB, 303... Weather information DB, 304... mole information storage DB, 305... log acquisition unit, 306... sleep time feature calculation unit, 307... sleep time feature memory unit, 308... user situation calculation unit, 309... user situation feature memory unit, 310... quality model construction unit, 351... correct answer data storage DB, 352... log storage DB, 353... log acquisition unit, 354... feature calculation unit, 355... sleep time model construction unit.

Claims

1. A device comprising: a log information acquisition unit that acquires log information of a user terminal used by a user; a situation derivation unit that estimates a user situation that affects sleep based on the log information; and a quality estimation unit that estimates the quality of the user's sleep based on the user situation.

2. The device according to claim 1, wherein the situation derivation unit derives the user situation using a learning model trained with features based on log information as explanatory variables and the user situation as a target variable.

3. The device according to claim 1, wherein the user situation includes at least one of the user's sleep situation, the user's immediate actions immediately before going to bed, the sleep environment in which the user is sleeping, and the user's daytime actions during the day.

4. The device according to claim 3, wherein the sleep status includes at least one of bedtime, wake-up time, and sleep duration during a predetermined past period.

5. The device according to claim 3, wherein the immediately preceding behavior indicates an operation status on the user terminal immediately before going to bed.

6. The device according to claim 5, wherein the operation status indicates an operation status for an application for each category.

7. The device according to claim 3, wherein the sleep environment includes at least one of information on whether or not the user is sleeping per unit time, lighting environment, noise environment, user terminal operation status, and temperature and humidity environment.

8. The device of claim 3, wherein the daytime activity indicates a situation in which the user is exposed to light.

9. The device according to claim 8, wherein the light exposure condition is based on at least one of the amount of solar radiation or the intensity of lighting.

10. A method comprising: a log information acquisition step of acquiring operation log information of a user terminal used by a user; a situation estimation step of estimating a user situation that affects sleep based on the operation log information; and a quality derivation step of deriving the quality of the user's sleep based on the user situation.

Images