Notification information generating device, notification information generating system, notification information generating method and program

Abstract

To provide a notification information generation apparatus, a notification information generation system, a notification information generation method and a program that can improve an observance situation of a medication schedule of a medication treatment patient.SOLUTION: A cloud server 1 comprises: a device information acquisition unit 111 that acquires device information indicating a state of a device 9 used by a user; a biological information acquisition unit 112 that acquires biological information of the user; a device information classification unit 113 that identifies a first class to which a state of the device 9 indicated by the device information belongs; a biological information classification unit 114 that identifies a second class to which a body condition of the user indicated by the biological information belongs; an estimation unit 117 that estimates an action of the user from the first class and the second class using a first learned model; a medication time calculation unit 118 that calculates medication time for the user on the basis of an action of the user; and a medication time notification unit 119 that generates notification information which notifies the user of the calculated medication time.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present disclosure relates to a notification information generating device, a notification information generating system, a notification information generating method, and a program. [Background technology]

[0002] A medication support system has been proposed that acquires blood pressure measurement information of a patient obtained by a blood pressure monitor worn on the patient's wrist and medication information showing the patient's medication status, and generates medication treatment performance information based on the acquired blood pressure measurement information, medication information, and medication information including the type and amount of medication prescribed for the patient (see, for example, Patent Document 1). Here, when the patient manually inputs the medication status into a portable information terminal carried by the patient, the medication information showing the medication status is sent from the portable information terminal to a medication support server. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-151993 Summary of the Invention [Problem to be solved by the invention]

[0004] Incidentally, in the medication support system described in Patent Document 1, there is a demand for preventing patients from forgetting to take their medication and improving their compliance with medication schedules by notifying them of the time to take their medication before the time actually arrives.

[0005] The present disclosure has been made in consideration of the above-mentioned reasons, and aims to provide a notification information generating device, a notification information generating system, a notification information generating method, and a program that can improve patients' compliance with medication schedules during medication therapy. [Means for solving the problem]

[0006] In order to achieve the above object, a notification information generating device according to the present disclosure includes: a device information acquisition unit that acquires device information indicating the status of at least one device used by a user; a biometric information acquisition unit that acquires biometric information indicating at least one type of biometric indicator that is preset for the user; a device information classifying unit that identifies a first class to which the state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one biometric indicator indicated by the biometric information belongs when the at least one biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication timing notification unit that generates notification information for notifying the user of the calculated medication timing; a location information acquisition unit that acquires location information of the user; a presence / absence determination unit that determines whether the user is present in a building in which the at least one device is installed based on the location information; Equipped with picture, When the presence / absence determination unit determines that the user is absent from the building, the estimation unit estimates the user's behavior from the second class using a second trained model for estimating the user's behavior from only the second class. . [Effects of the Invention]

[0007] According to the present disclosure, the estimation unit uses the trained model to estimate the user's behavior from a first class to which at least one device state indicated by the device information belongs and a second class to which at least one type of biometric indicator of the user indicated by the biometric information belongs.Then, the medication timing notification unit generates notification information that notifies the user of the medication timing calculated based on the user's behavior estimated by the estimation unit.This makes it possible to calculate the medication timing in advance from the device information and the biometric information and notify the user, thereby reducing the user's forgetting to take their medication and improving the user's compliance with their medication schedule. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a schematic configuration diagram of a notification information generating system according to an embodiment of the present disclosure; [Figure 2] FIG. 1 is a block diagram showing a hardware configuration of a notification information generating system according to an embodiment. [Figure 3] FIG. 1 is a block diagram showing a functional configuration of a terminal device according to an embodiment; [Figure 4] FIG. 1 is a block diagram showing a functional configuration of a cloud server according to an embodiment. [Figure 5] FIG. 10 is a diagram showing an example of information stored in a classification information storage unit according to an embodiment; [Figure 6] FIG. 10 is a diagram showing an example of information stored in a user behavior storage unit according to an embodiment; [Figure 7] FIG. 1 is a sequence diagram illustrating an operation of the notification information generating system according to an embodiment. [Figure 8] FIG. 1 is a sequence diagram illustrating an operation of the notification information generating system according to an embodiment. [Figure 9] 1 is a flowchart showing an example of the flow of a notification process executed by a notification information generating device according to an embodiment; [Figure 10] 1 is a flowchart showing an example of the flow of a notification process executed by a notification information generating device according to an embodiment; [Figure 11] FIG. 10 is a block diagram showing the functional configuration of a cloud server according to a modified example. [Figure 12] (A) is a diagram showing the transition of the user's sleep time, (B) is a diagram showing the user's bedtime, and (C) is a diagram showing the user's bathing interval. [Figure 13] 10 is a flowchart showing an example of the flow of a notification process executed by a notification information generating device according to a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0009] A notification information generating system according to each embodiment of the present disclosure will be described below with reference to the drawings. The notification information generating device according to the present embodiment includes a device information acquiring unit that acquires device information indicating the status of at least one device used by a user, a biometric information acquiring unit that acquires biometric information indicating at least one preset biometric indicator of the user, a device information classifying unit, and a biometric information classifying unit. The device information classifying unit identifies a first class to which the at least one device status indicated by the device information belongs when the status of the at least one device is classified into a plurality of preset classes. The biometric information classifying unit identifies a second class to which the at least one biometric indicator of the user belongs when the at least one biometric indicator of the user is classified into a plurality of preset classes. The notification information generating device further includes an estimation unit that estimates the user's behavior from the first class and the second class using a trained model for estimating the user's behavior from the first class and the second class, a medication timing calculation unit that calculates the timing of medication for the user based on the user's behavior, and a medication timing notification unit that generates notification information to notify the user of the calculated medication timing.

[0010] As shown in Fig. 1, the notification information generating system according to this embodiment includes a plurality of devices 9 (two in Fig. 1) installed in a building H, a cloud server 1, a smartwatch 2, a blood glucose sensor 21, a location management server 3, and a terminal device 5. The building H also includes a router 82 connected to a local network NW2 constructed within the building H, and a data circuit-terminating device 81 connected to the router 82 and a wide area network NW1. The wide area network NW1 is, for example, the Internet. The local network NW2 is, for example, a wired local area network (LAN) or a wireless LAN. The data circuit-terminating device 81 is, for example, an ONU (Optical Network Unit), a modem, a gateway, or the like.

[0011] The devices 9 are so-called cooking appliances such as refrigerators, rice cookers, microwave ovens, etc., and are capable of communicating with the cloud server 1 via the local network NW2 and the wide area network NW1. Each time the devices 9 receive device information request information periodically transmitted from the cloud server 1, requesting the devices 9 to transmit device information indicating operation information of the devices 9, the devices 9 generate device information indicating their own status and transmit the device information to the cloud server 1. Here, the device information includes, for example, information indicating whether the devices are operating or stopped.

[0012] The smartwatch 2 is worn, for example, on the user's wrist and functions as a biosensor that measures multiple types of pre-set biometric indicators of the user. Specifically, the smartwatch 2 measures the user's body temperature, blood pressure, and pulse. The smartwatch 2 is also capable of wireless communication with a blood glucose sensor 21, which is a biosensor worn on part of the user's body, and acquires blood glucose level information indicating the user's blood glucose level measured by the blood glucose sensor 21, from the blood glucose sensor 21. Each time the smartwatch 2 acquires biometric information request information periodically transmitted from the cloud server 1 requesting the smartwatch 2 to transmit biometric information, the smartwatch 2 generates and transmits to the cloud server 1 biometric information including body temperature information, blood pressure information, and pulse information indicating the measured biometric indicators, body temperature, blood pressure, and pulse, as well as blood glucose level information, which is a biometric indicator acquired from the blood glucose sensor 21.

[0013] The location management server 3 manages location information indicating the location of the terminal device 5 transmitted from the terminal device 5 in association with terminal device identification information that identifies the terminal device 5 that transmitted the location information. When a preset location information acquisition time arrives, the location management server 3 transmits to the terminal device 5 location information request information that requests the terminal device 5 to transmit the location information of the terminal device 5, thereby acquiring the location information transmitted from the terminal device 5.

[0014] The terminal device 5 is, for example, a smartphone, and as shown in FIG. 2 , includes a CPU (Central Processing Unit) 501, a main memory 502, an auxiliary memory 503, a display 504, an input unit 505, a wide-area communication unit 506, a GNSS (Global Navigation Satellite System) receiving unit 508, and a bus 509 connecting these components to one another. The main memory 502 has a volatile memory such as a RAM (Random Access Memory) and is used as a work area for the CPU 501. The auxiliary memory 503 is a non-volatile memory such as a semiconductor flash memory and stores programs for the CPU 501 to execute various processes. The display 504 is a display device such as a liquid crystal display or an organic EL (Electro-Luminescence) display. The input unit 505 is, for example, a transparent touchpad placed on top of the display 504. The wide area communication unit 506 is connected to the wide area network NW1, and transmits information transferred from the CPU 501 to the cloud server 1 via the wide area network NW1, and transfers information received from the cloud server 1 via the wide area network NW1 to the CPU 501. The GNSS receiving unit 508 receives GNSS signals emitted from artificial satellites, and generates reference position information indicating the position of the artificial satellite, time information, etc. based on the GNSS signals and transfers them to the CPU 501.

[0015] The CPU 501 reads out the program stored in the auxiliary storage unit 503 into the main storage unit 502 and executes it, thereby functioning as a reception unit 511, a display control unit 512, a positioning unit 513, a location information notification unit 514, a notification acquisition unit 515, a correction request generation unit 516, and a correction request transmission unit 517, as shown in Fig. 3. The main storage unit 502 shown in Fig. 2 also has a location storage unit 521 that temporarily stores location information indicating the location of the terminal device 5, as shown in Fig. 3.

[0016] When the receiving unit 511 receives an operation performed by the user on the input unit 505, it notifies the display control unit 512 of information indicating the content of the received operation. Furthermore, when the receiving unit 511 receives an operation performed by the user on the input unit 505 to start positioning of the terminal device 5, it notifies the positioning unit 513 of positioning start command information. The display control unit 512 generates an operation screen image based on the information notified from the receiving unit 511 and displays it on the display unit 504.

[0017] The positioning unit 513 estimates the position of the terminal device 5 based on the reference position information, time information, etc. input from the GNSS receiving unit 508. Then, the positioning unit 513 updates the position information stored in the position memory unit 521 with position information indicating the estimated position of the terminal device 5. When the position information notification unit 514 acquires position information request information from the position management server 3, it transmits the position information stored in the position memory unit 521 to the position management server 3. The notification acquisition unit 515 acquires notification information for notifying the timing of medication transmitted from the cloud server 1, and notifies the display control unit 512 of the acquired notification information. Here, when the display control unit 512 is notified of the notification information from the notification acquisition unit 515, it generates an operation screen image including the notified notification information and displays it on the display unit 504.

[0018] When the user determines that the medication timing indicated by the notification information acquired from the cloud server 1 is incorrect, the correction request generation unit 516 generates correction request information for the user to feedback the correct medication timing to the cloud server 1 and request that the user correct the determination of the medication timing. Here, it is assumed that the user refers to the medication timing displayed on the display unit 504 and determines that it is incorrect, and then performs an operation to input the correct medication timing via the input unit 505. In this case, the reception unit 511 notifies the correction request generation unit 516 of information indicating the correct medication timing input by the user. Then, the correction request generation unit 516 generates correction request information including the information indicating the correct medication timing notified by the reception unit 511, and notifies the correction request transmission unit 517 of the generated correction request information. The correction request transmission unit 517 transmits the correction request information notified by the correction request generation unit 516 to the cloud server 1.

[0019] Returning to FIG. 2, the cloud server 1 includes a CPU 101, a main memory unit 102, an auxiliary memory unit 103, a wide area communication unit 106, a clock unit 107, and a bus 109 interconnecting these units. The CPU 101 is, for example, a multi-core processor. The main memory unit 102 is made up of volatile memory and is used as a work area for the CPU 101. The auxiliary memory unit 103 is made up of large-capacity non-volatile memory and stores programs for realizing various functions of the cloud server 1. The wide area communication unit 106 is connected to the wide area network NW1. The clock unit 107 is, for example, a real-time clock.

[0020] 4, the CPU 101 reads out the programs stored in the auxiliary storage unit 103 into the main storage unit 102 and executes them, thereby functioning as a device information acquisition unit 111, a biometric information acquisition unit 112, a device information classification unit 113, a biometric information classification unit 114, a position information acquisition unit 115, a presence / absence determination unit 116, an estimation unit 117, a medication timing calculation unit 118, a medication timing notification unit 119, a correction request acquisition unit 120, a behavior identification unit 121, and a model generation unit 122. The auxiliary storage unit 103 also has a device information storage unit 131, a biometric information storage unit 132, a classification information storage unit 133, a model storage unit 134, a user behavior storage unit 135, a medication method storage unit 136, and a medication timing storage unit 137.

[0021] The device information storage unit 131 stores device information indicating the operating state of each device 9 in association with time information indicating the time when the device information was acquired from the device 9. The biological information storage unit 132 stores information indicating multiple types of preset biological indicators of a user undergoing drug treatment in association with time information indicating the time when the biological information was acquired. Specifically, the biological information storage unit 132 stores information indicating the user's body temperature, blood pressure, pulse rate, and blood glucose level in association with time information indicating the time when the corresponding biological information was acquired.

[0022] As shown in the upper part of FIG. 5, for example, the classification information storage unit 133 stores the state transitions of each of the multiple devices 9 and the corresponding time information in association with class information indicating the first class to which the combination of these devices belongs. Here, the class information indicates the first class to which each combination belongs, identified by clustering, into multiple pre-set classes, combinations of the state transitions of each of the multiple devices 9 indicated by all previously acquired device information and the times indicated by the time information. In the example shown in FIG. 5, the device information is classified into two classes: class C21, which indicates a low probability that the user has completed meal preparation, and class C22, which indicates a high probability that the user has completed meal preparation. For example, if, at 12:00, the microwave oven (device 9B) has already transitioned from a cooking state to a non-cooking state and the rice cooker (device 9A) has transitioned from a cooking state to a non-cooking state, this indicates that the device belongs to class C22. 5, the classification information storage unit 133 also stores multiple types (four types in FIG. 5) of biometric indicators indicated by each piece of biometric information and the corresponding time information in association with class information indicating a second class to which the combination belongs. Here, the class information indicates the second class to which each combination belongs, which is determined by clustering combinations of multiple types of biometric indicators indicated by all previously acquired biometric information and times indicated by the time information into multiple pre-defined classes. In the example shown in FIG. 5, the biometric information is classified into three classes: class C11, which indicates a low possibility that the user is eating; class C12, which indicates a medium possibility that the user is eating; and class C13, which indicates a high possibility that the user is eating. For example, if the user's body temperature is 36.4°C, blood pressure is 101 mmHg, pulse rate is 74, and blood glucose level is 100 mg / dL at 10:00, the class C11 indicates that the user belongs to class C11.

[0023] Returning to FIG. 4 , the model storage unit 134 stores a first trained model for estimating user behavior from a first class to which a combination of the state transitions of each of the multiple devices 9 and the time indicated by the time information belongs, and a second class to which a combination of the multiple types of biometric indicators and the time indicated by the time information belongs. The first trained model is, for example, a forward propagation neural network. In this case, the model storage unit 134 stores information indicating the structure of the neural network and information indicating weighting coefficients in the neural network. Here, the information indicating the structure of the neural network includes information indicating the number of nodes and layers of the neural network, as well as weighting coefficients and activation functions corresponding to each node. The model storage unit 134 also stores a second trained model for estimating user behavior from only the second class to which a combination of the multiple types of biometric indicators and the time indicated by the time information belongs. The second trained model is also, for example, a forward propagation neural network. Each neural network has an input layer, a hidden layer, and an output layer, and a numerical value indicating the first class and the second class is input to the input layer. In the output layer, processing using, for example, a softmax function is performed, and an expected value for each of multiple types of user behaviors set in advance is output based on the output from the final hidden layer. Then, the estimation unit 117, which will be described later, identifies the behavior with the highest expected value among the multiple types of user behaviors output from the output layer as the user behavior.

[0024] The user behavior storage unit 135 stores information indicating a user's behavior in association with time information indicating the time when the behavior was performed, as shown in Fig. 6, for example. Returning to Fig. 4, the medication method storage unit 136 stores medication method information including information on the time from when the user's preset behavior was performed to when the medication should be taken, for each medication taken by the user. The medication timing storage unit 137 stores medication timing information indicating the medication timing calculated based on the user's behavior and the medication method information.

[0025] The device information acquisition unit 111 acquires device information indicating the status of multiple devices used by the user. Here, the device information acquisition unit 111 acquires the device information transmitted from the device 9 by transmitting the above-mentioned device information request information to the device 9 each time a preset device information acquisition time arrives. Then, the device information acquisition unit 111 stores the acquired device information in the device information storage unit 131 in association with time information indicating the time when the device information was acquired, which is measured by the clock unit 107.

[0026] The biometric information acquisition unit 112 acquires biometric information indicating multiple types of biometric indicators that are preset for the user. Examples of biometric indicators include the user's body temperature, blood pressure, pulse rate, and blood glucose level. The biometric information acquisition unit 112 acquires the biometric information transmitted from the smartwatch 2 by transmitting the above-mentioned biometric information request information to the smartwatch 2 each time a preset biometric information acquisition time arrives. The biometric information acquisition unit 112 then extracts information indicating the body temperature, blood pressure, pulse rate, and blood glucose level from the acquired biometric information, and stores the extracted information in the biometric information storage unit 132 in association with time information that indicates the time the biometric information, measured by the timing unit 107, was acquired.

[0027] When the device information classification unit 113 determines that a user is present in building H based on the presence / absence information (described below) notified by the presence / absence determination unit 116, the device information classification unit 113 identifies a first class to which the states of the multiple devices 9 indicated by the most recently acquired device information belong, when the states of each of the multiple devices 9 indicated by the device information acquired within a predetermined period including the current time stored in the device information storage unit 131 are classified into a plurality of predetermined classes. The device information classification unit 113 performs clustering using a so-called unsupervised data classification method such as Ward's method or K-means algorithm on all combinations of state transitions of each of the multiple devices 9 indicated by the device information acquired within the aforementioned period and times indicated by the time information, thereby classifying the multiple devices 9, including combinations of state transitions and times indicated by the time information, into a plurality of classes. The device information classification unit 113 then identifies a first class to which the combination of state transitions of the multiple devices 9 indicated by the most recently acquired device information and times indicated by the time information belongs, from the classification results, and stores class information indicating the identified first class in association with the corresponding combination of states of the multiple devices 9 in the classification information storage unit 133. For example, when many of the multiple cooking appliances 9 are currently cooking, the appliance information classification unit 113 identifies class C21 as a class that is unlikely to indicate that meal preparation is complete. On the other hand, when few of the multiple cooking appliances 9 are currently cooking, the appliance information classification unit 113 identifies class C22 as a class that is likely to indicate that meal preparation is complete.

[0028] The biometric information classifying unit 114 identifies a second class to which a plurality of biometric indicators of a user indicated by most recently acquired biometric information belong when a plurality of biometric indicators of a user indicated by biometric information acquired within a predetermined period including the current time stored in the biometric information storage unit 132 are classified into a plurality of predetermined classes. The biometric information classifying unit 114 performs clustering using a so-called unsupervised data classification method such as Ward's method or K-means algorithm on all combinations of a plurality of types of biometric indicators of a user indicated by the biometric information acquired within the aforementioned period and times indicated by the time information, thereby classifying the combinations of a plurality of types of biometric indicators of a user indicated by the most recently acquired biometric information and times indicated by the time information into a plurality of classes. The biometric information classifying unit 114 then identifies a second class to which the combination of a plurality of types of biometric indicators of a user indicated by the most recently acquired biometric information and times indicated by the time information belongs from the classification result, and stores class information indicating the identified second class in association with the corresponding combination of a plurality of biometric indicators in the classification information storage unit 133. For example, when the body temperature, blood pressure, pulse rate, and blood sugar level are relatively low during a time slot that is unlikely to be lunchtime (for example, between 10:00 AM and 11:00 AM), the biometric information classification unit 114 identifies class C11 in Fig. 5 described above, i.e., the second class that is unlikely to indicate that the person is eating. On the other hand, when the body temperature, blood pressure, pulse rate, and blood sugar level are relatively high during lunchtime, the biometric information classification unit 114 identifies class C13 in Fig. 5 described above, i.e., the second class that is likely to indicate that the person is eating.

[0029] When the location information acquisition unit 115 acquires the location information of the terminal device 5 transmitted from the location management server 3, it notifies the presence / absence determination unit 116 of the acquired location information. The presence / absence determination unit 116 determines whether or not the user is present in the building H, based on the location information notified from the location information acquisition unit 115. Then, the presence / absence determination unit 116 generates presence / absence information indicating whether or not the user is present in the building H, based on the determination result, and notifies the estimation unit 117 of the presence / absence information.

[0030] Whenever a predetermined medication timing calculation time arrives, the estimation unit 117 determines that the user is present in building H based on the presence / absence information notified from the presence / absence determination unit 116, and acquires class information indicating the first class identified by the device information classification unit 113 and class information indicating the second class identified by the biological information classification unit 114, which are stored in the classification information storage unit 133. Then, the estimation unit 117 refers to the model information stored in the model storage unit 134, and estimates the user's behavior from the first class and the second class indicated by the acquired class information, using the above-mentioned first trained model. Furthermore, when the estimation unit 117 determines that the user is not present in building H based on the presence / absence information notified from the presence / absence determination unit 116, it acquires only the class information indicating the second class identified by the biological information classification unit 114, which is stored in the classification information storage unit 133. Then, the estimation unit 117 refers to the model information stored in the model storage unit 134, and uses the second trained model to estimate the user's behavior from only the second class indicated by the acquired class information. The estimation unit 117 stores the information indicating the estimated user's behavior in the user behavior storage unit 135 in association with time information.

[0031] The medication timing calculation unit 118 calculates the timing of medication for the user based on information indicating the user's behavior stored in the user behavior storage unit 135 and medication method information stored in the medication method storage unit 136. The medication timing calculation unit 118 stores information indicating the calculated medication timing in the medication timing storage unit 137. The medication timing notification unit 119 generates notification information that notifies the user of the medication timing indicated by the information stored in the medication timing storage unit 137, and transmits the generated notification information to the terminal device 5 carried by the user.

[0032] When the correction request acquisition unit 120 acquires the correction request information from the terminal device 5, it extracts medication timing information included in the acquired correction request information and notifies the behavior identification unit 121 of the extracted medication timing information. The behavior identification unit 121 identifies a preset reference behavior of the user that serves as a basis for calculating the medication timing and the time when the reference behavior was performed, based on the medication timing information calculated by the medication timing calculation unit 118 and the medication method information stored in the medication method storage unit 136. The behavior identification unit 121 then notifies the model generation unit 122 of information indicating the identified reference behavior and the time when the reference behavior was performed. For example, suppose that the medication timing indicated by the medication method information is "3 hours" after the user performed the reference behavior of "finishing the meal." Here, suppose that the user actually finished eating at "7:00 PM," but the estimation unit 117 estimates that the meal end time was "6:00 PM." In this case, the medication time calculation unit 118 calculates the medication time to be "21:00," and notification information indicating the medication time of "21:00" is transmitted to the terminal device 5. Here, it is assumed that the user determines that the notified medication time is incorrect and performs an operation to correct the medication time to "22:00" via the terminal device 5. In this case, correction request information including information indicating the correct medication time of "22:00" is transmitted from the terminal device 5, and this is acquired by the correction request acquisition unit 120. Then, the behavior identification unit 121 identifies "19:00," three hours before the medication time of "22:00," as the time when the reference behavior "finish eating" was performed, and notifies the model generation unit 122 of the identified reference behavior of "finish eating" and information indicating the time when the reference behavior was performed.

[0033] The model generation unit 122 generates a first trained model using a first class to which a combination of a past state of the device 9 and a time belongs, a second class to which a combination of multiple types of biometric indicators and a time of the past user belongs, and reference behavior of the user notified by the behavior identification unit 121, and the time at which the reference behavior was performed. Specifically, the model generation unit 122 first calculates evaluation values ​​for multiple types of pre-set user behaviors from the first class to which a combination of a past state of the device 9 and a time belongs and the second class to which a combination of multiple types of biometric indicators and a time of the past user belongs, using a neural network corresponding to the first trained model stored in the model storage unit 134. Next, the model generation unit 122 sets an evaluation value corresponding to a combination of the user's reference behavior and the time at which the reference behavior was performed, notified by the behavior identification unit 121, to be higher than evaluation values ​​for other combinations of behaviors and the time at which the behavior was performed. The model generation unit 122 then calculates the error between the evaluation value calculated using the neural network and the set evaluation value, and determines the weighting coefficients of the neural network by error backpropagation based on the calculated error. The model generation unit 122 updates the model information indicating the first trained model stored in the model storage unit 134 with model information indicating the determined weighting coefficients. The model generation unit 122 also generates a second trained model in the same manner as when generating the first trained model, using a second class to which combinations of multiple types of biometric indicators and time of the past user stored in the classification information storage unit 133 belong, and the user's reference behavior and the time when the reference behavior was performed, notified by the behavior identification unit 121. The model generation unit 122 then updates the model information indicating the second trained model stored in the model storage unit 134 with model information indicating the generated second trained model.

[0034] Next, the operation of the notification information generating system according to this embodiment will be described with reference to Figs. 7 and 8. First, when a preset device information acquisition time arrives, the above-mentioned device information request information is transmitted from the cloud server 1 to the device 9 (step S1). Meanwhile, upon acquiring the device information request information, the device 9 generates device information indicating the state of the device 9 (step S2). Next, the generated device information is transmitted from the device 9 to the cloud server 1 (step S3). Here, upon acquiring the device information, the cloud server 1 stores the acquired device information in the device information storage unit 131.

[0035] Next, when the preset time for acquiring biological information arrives, the above-mentioned biological information request information is transmitted from the cloud server 1 to the smartwatch 2 (step S4). Meanwhile, upon acquiring the biological information request information, the smartwatch 2 generates biological information including body temperature information, blood pressure information, and pulse information indicating the measured body temperature, blood pressure, and pulse, and blood glucose level information acquired from the blood glucose sensor 21 (step S5). The generated biological information is then transmitted from the smartwatch 2 to the cloud server 1 (step S6). Here, upon acquiring the biological information, the cloud server 1 extracts information indicating the body temperature, blood pressure, pulse, and blood glucose level from the acquired biological information, and stores the extracted various pieces of information in the biological information storage unit 132.

[0036] Next, when a preset time for obtaining location information arrives, location information request information requesting the terminal device 5 to transmit location information of the terminal device 5 is transmitted from the location management server 3 to the terminal device 5 (step S7). Meanwhile, when the terminal device 5 obtains the location information request information, it generates location information of the terminal device 5 (step S8). Subsequently, the generated location information is transmitted from the terminal device 5 to the location management server 3 (step S9). Thereafter, when a preset time for transmitting location information arrives, the location information of the terminal device 5 is transmitted from the location management server 3 to the cloud server 1 (step S10).

[0037] Then, suppose that a predetermined medication timing calculation time arrives, and it is determined that the user is at home based on the location information of the terminal device 5 (step S11). In this case, the cloud server 1 identifies a first class to which a combination of state transitions of the multiple devices 9 indicated by the device information and the time at which the device information was acquired belongs, and identifies a second class to which a combination of multiple biometric indicators of the user indicated by the biometric information and the time at which the biometric information was acquired belongs (step S12). Next, the cloud server 1 estimates the user's behavior using the first trained model (step S13). Next, the cloud server 1 calculates the user's medication timing based on the information indicating the estimated user's behavior and the medication method information stored in the medication method storage unit 136, and stores information indicating the calculated medication timing in the medication timing storage unit 137 (step S14). 8, when the preset notification time arrives, the cloud server 1 generates notification information for notifying the user of the calculated administration time (step S15), and the generated notification information is transmitted from the cloud server 1 to the terminal device 5 carried by the user (step S16). Meanwhile, upon acquiring the notification information, the terminal device 5 notifies the user of the notification information by displaying an operation screen image including the acquired notification information on the display unit 504 (step S17).

[0038] Furthermore, suppose that the aforementioned medication timing calculation time arrives, and it is determined that the user is absent based on the location information of the terminal device 5 (step S18). In this case, the cloud server 1 identifies only the second class to which the combination of the user's multiple biometric indicators indicated by the biometric information and the time the biometric information was acquired belongs (step S19). Next, the cloud server 1 estimates the user's behavior using the aforementioned second trained model (step S20). Subsequently, the processing from step S14 onwards is executed.

[0039] Furthermore, suppose that the user determines that the medication timing displayed on the display unit 504 of the terminal device 5 is incorrect and performs a correction information input operation to input the correct medication timing via the input unit 505. In this case, the terminal device 5 accepts the correction information (step S21) and generates the above-mentioned correction request information (step S22). Thereafter, the generated correction request information is transmitted from the terminal device 5 to the cloud server 1 (step S23). Meanwhile, upon receiving the correction request information, the cloud server 1 identifies the above-mentioned user's preset reference behavior and the time when the reference behavior was performed based on the calculated medication timing information and the medication method information stored in the medication method storage unit 136 (step S24). Next, the cloud server 1 generates a first trained model or a second trained model using a first class to which a combination of the past state of the device 9 and a time belongs, a second class to which a combination of multiple types of biometric indicators of the user in the past and a time belongs, and the above-mentioned user's reference behavior and the time when the reference behavior was performed (step S25). Next, the cloud server 1 updates the model information stored in the model storage unit 134 with model information indicating the generated first trained model or second trained model (step S26).

[0040] Next, the notification process executed by the cloud server 1 according to this embodiment will be described with reference to FIGS. 9 and 10. This notification process is started, for example, when a program for executing the notification process is started in the cloud server 1. First, the device information acquisition unit 111 determines whether a preset device information acquisition time has arrived based on time information notified from the clock unit 107 (step S101). If the device information acquisition unit 111 determines that the device information acquisition time has not yet arrived (step S101: No), the process of step S104, which will be described later, is executed. On the other hand, if the device information acquisition unit 111 determines that the device information acquisition time has arrived (step S101: Yes), the device information acquisition unit 111 transmits the above-mentioned device information request information to the devices 9 (step S102) to acquire device information from each device 9, and stores the acquired device information in the device information storage unit 131 in association with time information indicating the time when the device information was acquired (step S103).

[0041] Next, the biometric information acquisition unit 112 determines whether a preset biometric information acquisition time has arrived based on the time information notified from the clock unit 107 (step S104). If the biometric information acquisition unit 112 determines that the biometric information acquisition time has not yet arrived (step S104: No), the processing of step S107, which will be described later, is executed. On the other hand, if the biometric information acquisition unit 112 determines that the biometric information acquisition time has arrived (step S104: Yes), the biometric information acquisition unit 112 transmits the above-mentioned biometric information request information to the smartwatch 2 (step S105), acquires biometric information from the smartwatch 2, extracts information indicating body temperature, blood pressure, pulse rate, and blood sugar level from the acquired biometric information, and stores the extracted information in the biometric information storage unit 132 in association with time information indicating the time the biometric information was acquired (step S106).

[0042] Next, the location information acquisition unit 115 determines whether or not location information transmitted from the location management server 3 has been acquired (step S107). If the location information acquisition unit 115 determines that location information has not yet been acquired (step S107: No), the processing of step S109, which will be described later, is executed. On the other hand, if the location information acquisition unit 115 determines that location information has been acquired (step S107: Yes), it notifies the presence / absence determination unit 116 of the acquired location information (step S108). Thereafter, the presence / absence determination unit 116 determines whether or not a preset medication timing calculation time has arrived (step S109). Here, if the presence / absence determination unit 116 determines that the medication timing calculation time has not yet arrived (step S109: No), the processing of step S116, which will be described later, is executed. On the other hand, when the presence / absence determination unit 116 determines that the time for calculating the medication timing has arrived (step S109: Yes), it determines whether the user is present in the building H based on the location information notified from the location information acquisition unit 115 (step S110). Here, when the presence / absence determination unit 116 determines that the user is present in the building H (step S110: Yes), the device information classification unit 113 identifies a first class to which a combination of state transitions of the plurality of devices 9 indicated by the device information stored in the device information storage unit 131 and a time indicated by the time information belongs. In addition, the biological information classification unit 114 identifies a second class to which a plurality of biological indicators of the user indicated by the biological information stored in the biological information storage unit 132 belong (step S111). Here, the device information classification unit 113 stores class information indicating the identified first class in the classification information storage unit 133 in association with the combination of states of the corresponding plurality of devices 9. Furthermore, the biometric information classifying unit 114 stores the class information indicating the identified second class in the classification information storage unit 133 in association with the combination of the corresponding plurality of biometric indices.

[0043] Next, the estimation unit 117 acquires class information indicating the first class identified by the device information classification unit 113 and class information indicating the second class identified by the biometric information classification unit 114, which are stored in the classification information storage unit 133, and estimates the user's behavior from the first class and the second class indicated by the acquired class information, using the above-mentioned first trained model indicated by the model information stored in the model storage unit 134 (step S112). Here, the estimation unit 117 stores the information indicating the estimated user's behavior in the user behavior storage unit 135 in association with time information. Subsequently, the processing of step S115, which will be described later, is executed.

[0044] Furthermore, it is assumed that the presence / absence determination unit 116 determines in step S110 that the user is not present in building H (step S110: No). In this case, only the biometric information classification unit 114 identifies a second class to which the multiple biometric indicators of the user indicated by the biometric information stored in the biometric information storage unit 132 belong (step S113). Thereafter, the estimation unit 117 acquires only class information indicating the second class identified by the biometric information classification unit 114, stored in the classification information storage unit 133. Then, the estimation unit 117 estimates the user's behavior from only the second class indicated by the acquired class information, using the above-mentioned second trained model indicated by the model information stored in the model storage unit 134 (step S114). Here, the estimation unit 117 associates information indicating the estimated user's behavior with time information and stores it in the user behavior storage unit 135.

[0045] Next, the medication timing calculation unit 118 calculates the medication timing for the user based on the information indicating the user's behavior stored in the user behavior storage unit 135 and the medication method information stored in the medication method storage unit 136 (step S115). Here, the medication timing calculation unit 118 stores information indicating the calculated medication timing in the medication timing storage unit 137. Next, the medication timing notification unit 119 determines whether a preset notification timing has arrived, as shown in FIG. 10 (step S116). Here, if the medication timing notification unit 119 determines that the notification timing has not arrived yet (step S116: No), the processing of step S118 described below is executed. On the other hand, when the medication timing notification unit 119 determines that the notification time has arrived (step S116: Yes), it generates notification information to notify the user of the medication time indicated by the information stored in the medication timing memory unit 137, and transmits the generated notification information to the terminal device 5 carried by the user (step S117).

[0046] Thereafter, the correction request acquisition unit 120 determines whether or not the correction request information has been acquired from the terminal device 5 (step S118). If the correction request acquisition unit 120 determines that the correction request information has not been acquired (step S118: No), the process of step S101 is executed again. On the other hand, if the correction request acquisition unit 120 determines that the correction request information has been acquired (step S118: Yes), the correction request acquisition unit 120 extracts medication timing information included in the acquired correction request information and notifies the behavior identification unit 121 of the extracted medication timing information (step S119). Next, the behavior identification unit 121 identifies a preset reference behavior of the user that serves as a reference for calculating the medication timing and the time when the reference behavior was performed, based on the medication timing information included in the correction request information and the medication method information stored in the medication method storage unit 136 (step S120). Here, the behavior identification unit 121 notifies the model generation unit 122 of the identified reference behavior and information indicating the time when the reference behavior was performed.

[0047] Next, the model generation unit 122 generates a first trained model or a second trained model using a first class to which a combination of a past state of the device 9 and a time belongs, a second class to which a combination of multiple types of past biometric indicators of the user and a time belongs, and the reference behavior of the user and the time when the reference behavior was performed, which are notified by the behavior identification unit 121, stored in the classification information storage unit 133 (step S121). After that, the model generation unit 122 updates the model information indicating the second trained model stored in the model storage unit 134 with model information indicating the generated first trained model or second trained model (step S122). Next, the processing of step S101 is executed again.

[0048] As described above, according to the notification information generating device of this embodiment, the estimation unit 117 uses the first learned model or the second learned model described above to estimate the user's behavior from a first class to which the states of multiple devices 9 indicated by the device information belong and a second class to which multiple types of biometric indicators of the user indicated by the biometric information belong. Then, the medication timing notification unit 119 generates notification information that notifies the user of the medication timing calculated based on the user's behavior estimated by the estimation unit 117. This makes it possible to calculate the medication timing in advance from the device information and the biometric information and notify the user, thereby reducing the user's forgetting to take their medication and improving the user's compliance with their medication schedule.

[0049] Furthermore, the presence / absence determination unit 116 according to this embodiment determines whether the user is present in the building H in which the device 9 is installed, based on the location information of the terminal device 5 carried by the user. Then, when the presence / absence determination unit 116 determines that the user is absent from the building H, the estimation unit 117 uses the second trained model described above to estimate the user's behavior from only the second class to which the combination of the user's multiple types of biometric indicators and time belongs. This improves the accuracy of the estimation of the user's behavior because the second trained model does not use the first class corresponding to the state of the device 9 when the user is not present in the building H and there is no possibility that the user will use the device 9. Therefore, the accuracy of the timing of the user's medication calculated based on the user's behavior can be improved.

[0050] Furthermore, in the cloud server 1 according to this embodiment, when the correction request acquisition unit 120 acquires, from the terminal device 5, correction request information requesting a correction of the medication timing indicated in the notification information, the behavior identification unit 121 identifies the user's behavior corresponding to the corrected medication timing indicated by the correction request information. Then, the model generation unit 122 generates a first trained model or a second trained model based on the user's behavior identified by the behavior identification unit 121. Then, the model generation unit 122 updates the model information stored in the model storage unit 134 with model information indicating the generated first trained model or second trained model. This gradually improves the accuracy of the estimation of the user's behavior using the first trained model or the second trained model, thereby increasing the accuracy of the user's medication timing calculated based on the user's behavior.

[0051] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments. For example, the cloud server may determine whether an abnormality has occurred in a user based on the user's behavior, and if it determines that an abnormality has occurred in the user, generate notification information to notify the user that an abnormality has occurred. The hardware configuration of the cloud server 2001 according to this modification is the same as the hardware configuration of the cloud server 1 according to the embodiment shown in FIG. 2. Hereinafter, the hardware configuration according to this embodiment will be described using the same reference numerals as those used in the description of the embodiment. 11, the CPU 101 of the cloud server 2001 according to this modification reads out the programs stored in the auxiliary storage unit 103 into the main storage unit 102 and executes them, thereby functioning as a device information acquisition unit 111, a biological information acquisition unit 112, a device information classification unit 113, a biological information classification unit 114, a location information acquisition unit 115, a presence / absence determination unit 116, an estimation unit 2117, a medication timing calculation unit 118, a medication timing notification unit 119, a correction request acquisition unit 120, a behavior identification unit 121, a model generation unit 122, a sleep time calculation unit 2123, a bathing interval calculation unit 2124, an abnormality determination unit 2125, and an abnormality notification unit 2126. In FIG. 11, the same components as those in the embodiment are denoted by the same reference numerals as those in FIG. 4. In addition, the auxiliary memory unit 103 has an equipment information memory unit 131, a biometric information memory unit 132, a classification information memory unit 133, a model memory unit 2134, a user behavior memory unit 135, a medication method memory unit 136, a medication timing memory unit 137, and an abnormality determination criteria memory unit 2138.

[0052] In the notification information generation system according to this modification, the devices 9 include not only so-called cooking appliances such as refrigerators, rice cookers, and microwave ovens, but also home appliances such as air conditioners, water heaters, and lighting fixtures. Similarly to the embodiment, the devices 9 can communicate with the cloud server 1 via the local network NW2 and the wide area network NW1. The cloud server 2001 can also communicate with a terminal device 6, via the wide area network NW1, that is owned by an administrator who manages users other than the user receiving medication treatment. The terminal device 6 is, for example, a smartphone, and upon receiving notification information transmitted from the cloud server 2001 notifying the user that an abnormality has occurred, the terminal device 6 displays a notification screen including the received notification information on a display unit.

[0053] The model storage unit 2134 stores model information indicating a first trained model for estimating a user's behavior from a first class to which a combination of the state transitions of each of the multiple devices 9 and the times indicated by the time information belongs, and a second class to which a combination of the multiple types of biometric indicators and the times indicated by the date and time information belongs. The model storage unit 2134 also stores model information indicating a second trained model for estimating a user's behavior from only the second class to which a combination of the multiple types of biometric indicators and the times indicated by the date and time information belongs. Here, the first trained model and the second trained model are models for estimating not only whether the user has finished eating, but also whether the user has started and finished bathing, gone to bed, and woken up. The user behavior storage unit 2135 stores information indicating multiple types of user behavior in association with time information indicating the time when the behavior was performed. Here, the user behavior storage unit 2135 stores information indicating behaviors such as "before eating" and "during eating," as well as "wake up," "start bathing," and "go to bed," in association with time information indicating the time when the behavior was performed. The abnormality determination criteria storage unit 2137 stores information indicating abnormality determination criteria corresponding to the user's bedtime, sleeping hours, and bathing intervals.

[0054] The estimation unit 2117 estimates the user's behavior using the first trained model or the second trained model, as described in the embodiment, each time a preset medication timing calculation time arrives. Here, the estimation unit 2117 estimates not only that the user has finished eating, but also whether or not the user has performed behaviors such as starting and finishing a bath, and going to bed and waking up. Furthermore, the estimation unit 2117 stores information indicating the estimated user's behavior in the user behavior storage unit 2135 in association with time information.

[0055] The sleep time calculation unit 2123 calculates the user's sleep time based on date and time information associated with the user's actions of waking up and going to bed, which is stored in the user behavior storage unit 2135. Then, the sleep time calculation unit 2123 notifies the abnormality determination unit 2125 of information indicating the calculated sleep time. The bathing interval calculation unit 2124 calculates the user's bathing interval based on date and time information associated with the user's action of starting a bath, which is stored in the user behavior storage unit 2135. Then, the bathing interval calculation unit 2124 notifies the abnormality determination unit 2125 of information indicating the calculated bathing interval.

[0056] The abnormality determination unit 2125 acquires date and time information stored in the user behavior storage unit 2135 associated with the user's bedtime behavior. The abnormality determination unit 2125 then determines whether the user's bedtime indicated by the acquired date and time information, the sleep time indicated by the information notified from the sleep time calculation unit 2123, and the bathing interval indicated by the information notified from the bathing interval calculation unit 2124 satisfy the abnormality determination criteria corresponding to the bedtime, sleep time, and bathing interval, respectively, stored in the abnormality determination criteria storage unit 2137. When the abnormality determination unit 2125 determines that any one of the bedtime, sleep time, and bathing interval satisfies the abnormality determination criteria, the abnormality notification unit 2126 generates abnormality notification information notifying the user that an abnormality has occurred and transmits the information to the terminal device 6 possessed by the administrator who manages the user. Upon acquiring the abnormality notification information from the cloud server 2001, the terminal device 6 displays the acquired abnormality notification information on its display unit to notify the administrator of the user's abnormality.

[0057] The abnormality determination unit 2125 monitors the changes in the sleeping hours, bedtime, and bathing intervals, for example, as shown in each of Figures 12(A) to 12(C). If the abnormality determination unit 2125 determines that the sleeping hours, bedtime, or bathing intervals exceed the abnormality determination criteria shown by the dashed lines in Figures 12(A) to 12(C), the abnormality notification unit 2126 generates the abnormality notification information described above and transmits it to the terminal device 6.

[0058] Next, the notification process executed by the cloud server 2001 according to this embodiment will be described with reference to Fig. 13. In Fig. 13, the same processes as those in the first embodiment are denoted by the same reference numerals as those in Figs. 9 and 10. First, a series of processes from steps S101 to S109 is executed, and the presence / absence determination unit 116 determines that the time for calculating the medication timing has arrived (step S109: Yes), and determines that the user is present in building H based on the location information notified from the location information acquisition unit 115 (step S110: Yes). In this case, the device information classification unit 113 identifies the first class described above, and the biological information classification unit 114 identifies the second class described above (step S111). Next, the estimation unit 2117 acquires class information indicating the first class identified by the device information classification unit 113 and class information indicating the second class identified by the biological information classification unit 114, which are stored in the classification information storage unit 133, and estimates the user's behavior from the first class and the second class indicated by the acquired class information, using the above-mentioned first trained model indicated by the model information stored in the model storage unit 134 (step S2101). Here, the estimation unit 117 estimates not only that the user has finished eating, but also whether the user has started and finished bathing, gone to bed, and woken up.

[0059] Next, the sleep time calculation unit 2123 calculates the user's sleep time based on date and time information stored in the user behavior storage unit 2135 and associated with the time when the user performed the behavior of waking up and going to bed (step S2102). Here, the sleep time calculation unit 2123 notifies the abnormality determination unit 2125 of information indicating the calculated sleep time. Thereafter, the bathing interval calculation unit 2124 calculates the user's bathing interval based on date and time information stored in the user behavior storage unit 2135 and associated with the time when the user performed the behavior of starting a bath (step S2103). Here, the bathing interval calculation unit 2124 notifies the abnormality determination unit 2125 of information indicating the calculated bathing interval.

[0060] Next, the abnormality determination unit 2125 acquires date and time information associated with the time when the user performed the behavior of going to bed, which is stored in the user behavior storage unit 2135. Then, the abnormality determination unit 2125 determines whether the user's bedtime indicated by the acquired date and time information, the sleeping time indicated by the information notified from the sleeping time calculation unit 2123, and the bathing interval indicated by the information notified from the bathing interval calculation unit 2124 satisfy the abnormality determination criteria corresponding to the bedtime, sleeping time, and bathing interval, respectively, stored in the abnormality determination criteria storage unit 2137 (step S2104). Here, if the abnormality determination unit 2125 determines that none of the bedtime, sleeping time, and bathing interval satisfy the abnormality determination criteria (step S2104: No), the processing from step S115 onwards is executed. On the other hand, it is assumed that the abnormality determination unit 2125 determines that any one of the bedtime, sleeping time, and bathing interval satisfies the abnormality determination criteria (step S2104: Yes). In this case, the abnormality notification unit 2126 generates abnormality notification information for notifying the user that an abnormality has occurred, and transmits the information to the terminal device 6 possessed by the administrator who manages the user (step S2105). Subsequently, the processes from step S115 onwards are executed.

[0061] According to this configuration, if an abnormality occurs to a user, the user's administrator can be notified of this, making it possible to quickly respond to the abnormality for the user.

[0062] In the embodiment, an example has been described in which the terminal device 5 is a smartphone, but the present invention is not limited to this, and the terminal device 5 may be, for example, a smart speaker. In this case, when the terminal device 5 acquires notification information from the cloud server 1, it converts the acquired notification information into audio information and notifies the user of the content of the notification information by audio from the speaker. Alternatively, the terminal device 5 may be replaced by a television that functions as the terminal device 5.

[0063] In the embodiment, an example has been described in which the estimation unit 117 of the cloud server 1 estimates the user's behavior from the first class identified by the device information classification unit 113 and the second class identified by the biological information classification unit 114 using the first trained model each time a preset medication timing calculation time arrives. However, this is not limiting, and the estimation unit 117 may estimate the user's behavior only from the first class identified by the device information classification unit 113. In this case, the model storage unit 134 may store a trained model for estimating the user's behavior only from the first class to which the combination of the state transitions of each of the plurality of devices 9 and the time indicated by the time information belongs, for example.

[0064] According to this configuration, there is no need to prepare an expensive smartwatch for obtaining biological indicators, and therefore the cost for realizing the notification information generating system can be reduced.

[0065] In the embodiment, an example has been described in which smartwatch 2 and blood glucose sensor 21 are capable of wireless communication to measure blood glucose levels, and smartwatch 2 acquires blood glucose level information from blood glucose sensor 21. However, this is not limiting, and for example, smartwatch 2 may have a built-in blood glucose sensor 21. Then, the user may be able to acquire blood glucose level information indicating the blood glucose level measured by blood glucose sensor 21 simply by wearing smartwatch 2 on their wrist.

[0066] Furthermore, the various functions of the cloud server 1, 2001 and terminal device 5 according to the present disclosure may be realized by software, firmware, or a combination of software and firmware. In this case, the software or firmware may be written as a program, and the program may be stored and distributed on a computer-readable recording medium such as a flexible disk, CD-ROM (Compact Disc Read Only Memory), DVD (Digital Versatile Disc), or MO (Magneto-Optical Disc). The program may then be read and installed on a computer to configure a computer capable of realizing each of the aforementioned functions. In addition, when each function is realized by sharing the work between an operating system (OS) and an application, or by cooperation between the OS and an application, only the parts other than the OS may be stored on the recording medium.

[0067] Furthermore, each program can be superimposed on a carrier wave and distributed over a network. For example, the program can be posted on a bulletin board system (BBS) on the network and distributed over the network. These programs can then be started and run under the control of the OS in the same way as other application programs, thereby enabling the above-mentioned processing to be performed. [Industrial Applicability]

[0068] The present disclosure is suitable as a system for allowing a user to manage a medication schedule. [Explanation of symbols]

[0069] 1,2001,3001 Cloud server, 2 Smartwatch, 3 Location management server, 5 Terminal device, 101,501 CPU, 102,502 Main memory unit, 103,503 Auxiliary memory unit, 106,506 Wide area communication unit, 107 Timekeeping unit, 109,509 Bus, 111 Device information acquisition unit, 112 Biometric information acquisition unit, 113 Device information classification unit, 114 Biometric information classification unit, 115 Location information acquisition unit, 116 Presence / absence determination unit, 117,2117 Estimation unit, 118 Medication timing calculation unit, 119 Medication timing notification unit, 120 Correction request acquisition unit, 121 Behavior identification unit, 122,2122 Model generation unit, 131 Device information storage unit, 132 Biometric information storage unit, 133 Classification information storage unit, 134, 2134 model storage unit, 135, 2135 user behavior storage unit, 136 medication method storage unit, 137 medication timing storage unit, 504 display unit, 505 input unit, 508 GNSS receiving unit, 511 reception unit, 512 display control unit, 513 positioning unit, 514 location information notification unit, 515 notification acquisition unit, 516 correction request generation unit, 517 correction request transmission unit, 521 location storage unit, 2123 sleep time calculation unit, 2124 bathing interval calculation unit, 2125 abnormality determination unit, 2126 abnormality notification unit, 2137 abnormality determination criterion storage unit, NW1 wide area network, NW2 local network

Claims

1. a device information acquisition unit that acquires device information indicating the status of at least one device used by a user; a biometric information acquisition unit that acquires biometric information indicating at least one type of biometric indicator that is preset for the user; a device information classifying unit that identifies a first class to which a state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one biometric indicator indicated by the biometric information belongs when the at least one biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication timing notification unit that generates notification information for notifying the user of the calculated medication timing; a location information acquisition unit that acquires location information of the user; a presence / absence determination unit that determines whether the user is present in a building in which the at least one device is installed based on the location information, When the presence / absence determination unit determines that the user is absent from the building, the estimation unit estimates the user's behavior from the second class using a second trained model for estimating the user's behavior from only the second class. Notification information generation device.

2. A device information acquisition unit that acquires device information indicating the status of at least one device used by a user; a biometric information acquisition unit that acquires biometric information indicating at least one type of biometric indicator that is preset for the user; a device information classifying unit that identifies a first class to which a state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one biometric indicator indicated by the biometric information belongs when the at least one biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication timing notification unit that generates notification information for notifying the user of the calculated medication timing; a correction request acquisition unit that acquires correction request information that requests a correction of the medication timing; a behavior identification unit that identifies a user behavior corresponding to the corrected medication timing indicated by the correction request information; a model generation unit that generates the first trained model based on the user behavior identified by the behavior identification unit, Notification information generation device.

3. an abnormality determination unit that determines whether an abnormality has occurred in the user based on the behavior of the user; an abnormality notification unit that, when it is determined by the abnormality determination unit that an abnormality has occurred in the user, generates notification information that notifies the user that an abnormality has occurred.

3. The notification information generating device according to claim 1 or 2.

4. a biosensor attached to a user, which measures at least one type of biomarker set in advance of the user, and generates biometric information indicating the at least one type of biomarker obtained by the measurement; a terminal device carried by the user; a device information acquisition unit that acquires device information indicating a status of at least one device used by the user; a biometric information acquisition unit that acquires the biometric information from the biometric sensor; a device information classifying unit that identifies a first class to which a state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one biometric indicator indicated by the biometric information belongs when the at least one biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication time notifying unit that generates notification information for notifying the user of the calculated medication time and transmits the notification information to the terminal device; a location information acquisition unit that acquires location information of the user; a presence / absence determination unit that determines whether the user is present in a building in which the at least one device is installed based on the location information, When the presence / absence determination unit determines that the user is absent from the building, the estimation unit estimates the user's behavior from the second class using a second trained model for estimating the user's behavior from only the second class. Notification information generation system.

5. A biosensor that is attached to a user, measures at least one type of biometric indicator that is preset for the user, and generates biometric information that indicates the at least one type of biometric indicator obtained by the measurement; a terminal device carried by the user; a device information acquisition unit that acquires device information indicating a status of at least one device used by the user; a biometric information acquisition unit that acquires the biometric information from the biometric sensor; a device information classifying unit that identifies a first class to which a state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one biometric indicator indicated by the biometric information belongs when the at least one biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication time notifying unit that generates notification information for notifying the user of the calculated medication time and transmits the notification information to the terminal device; a correction request acquisition unit that acquires correction request information that requests a correction of the medication timing; a behavior identification unit that identifies a user behavior corresponding to the corrected medication timing indicated by the correction request information; a model generation unit that generates the first trained model based on the user behavior identified by the behavior identification unit, Notification information generation system.

6. acquiring device information indicating the status of at least one device used by the user; acquiring biometric information indicating at least one type of biometric indicator set in advance of the user; a step of identifying a first class to which the state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; identifying a second class to which the at least one type of biomarker indicated by the biometric information belongs when the at least one type of biomarker is classified into a plurality of preset classes; a step of estimating the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; Calculating the timing of administering medication to the user based on the user's behavior; generating notification information for notifying the user of the calculated medication timing; acquiring location information of the user; and determining whether the user is present in a building in which the at least one device is installed based on the location information; In the step of estimating the user's behavior, when it is determined that the user is absent from the building, the user's behavior is estimated from the second class using a second trained model for estimating the user's behavior from only the second class. A computer-implemented method for generating notification information.

7. A step of acquiring device information indicating the status of at least one device used by a user; acquiring biometric information indicating at least one type of biometric indicator set in advance of the user; a step of identifying a first class to which the state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; identifying a second class to which the at least one type of biomarker indicated by the biometric information belongs when the at least one type of biomarker is classified into a plurality of preset classes; a step of estimating the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; Calculating the timing of administering medication to the user based on the user's behavior; generating notification information for notifying the user of the calculated medication timing; acquiring correction request information requesting a correction of the medication timing; Identifying a user's behavior corresponding to the corrected medication timing indicated by the correction request information; generating the first trained model based on the identified user behavior; A computer-implemented method for generating notification information.

8. Computer, a device information acquisition unit that acquires device information indicating the status of at least one device used by the user; a biometric information acquisition unit that acquires biometric information indicating at least one type of biometric indicator that has been preset for the user; a device information classifying unit that identifies a first class to which a state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one type of biometric indicator indicated by the biometric information belongs when the at least one type of biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication timing notification unit that generates notification information for notifying the user of the calculated medication timing; a location information acquisition unit that acquires location information of the user; a presence / absence determination unit that determines whether the user is present in a building in which the at least one device is installed based on the location information; A program for functioning as When the presence / absence determination unit determines that the user is absent from the building, the estimation unit estimates the user's behavior from the second class using a second trained model for estimating the user's behavior from only the second class. program.

9. A computer, a device information acquisition unit that acquires device information indicating the status of at least one device used by the user; a biometric information acquisition unit that acquires biometric information indicating at least one type of biometric indicator that has been preset for the user; a device information classifying unit that identifies a first class to which a state of the at least one device indicated by the device information belongs when the state of the at least one device is classified into a plurality of preset classes; a biometric information classifying unit that identifies a second class to which the at least one type of biometric indicator indicated by the biometric information belongs when the at least one type of biometric indicator is classified into a plurality of preset classes; an estimation unit that estimates the user's behavior from the first class and the second class using a first trained model for estimating the user's behavior from the first class and the second class; a medication timing calculation unit that calculates a medication timing for the user based on the behavior of the user; a medication timing notification unit that generates notification information for notifying the user of the calculated medication timing; a correction request acquisition unit that acquires correction request information that requests a correction of the medication timing; a behavior identification unit that identifies a user behavior corresponding to the corrected medication timing indicated by the correction request information; a model generation unit that generates the first trained model based on the user's behavior identified by the behavior identification unit; A program to function as a

Images