Information processing device, information processing method, and program

Abstract

[Problem] To provide a mechanism capable of further improving the quality of user experience. [Solution] Provided is an information processing device comprising a control unit (116) that generates control information to be used by an inhalation device (100), the inhalation device heating an aerosol source to generate aerosol on the basis of the control information, wherein the control information defines a parameter relating to a temperature at which the aerosol source is heated, the control unit collects a plurality of training data including a combination of first control information, an evaluation set for the first control information, and second control information, the second control information is to be generated on the basis of the first control information and the evaluation set for the first control information, and the control unit generates the control information to be used by the inhalation device of a first user on the basis of a generation model for the control information trained on the basis of the collected plurality of training data.

Description

Technical Field

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

Background Art

[0002] Suction devices that generate substances to be inhaled by users, such as electronic cigarettes and nebulizers, are widely spread. For example, a suction device uses a base material including an aerosol source for generating an aerosol and a flavor source for imparting a flavor component to the generated aerosol, etc., to generate an aerosol to which a flavor component is imparted. A user can enjoy a flavor by inhaling the aerosol to which a flavor component is imparted, which is generated by the suction device. The operation of a user inhaling an aerosol is hereinafter also referred to as a puff or a puff operation.

[0003] Preferences for the flavor experienced when puffing vary from user to user. Therefore, it is preferable that the temperature for heating the aerosol source that directly affects the flavor can be customized by the user. Patent Document 1 below discloses a technique for a user to customize the temperature for heating the aerosol source.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, the technique according to Patent Document 1 above has been developed only recently, and there is still room for improvement from various viewpoints.

[0006] Therefore, the present disclosure has been made in view of the above problems, and an object of the present disclosure is to provide a mechanism capable of further improving the quality of the user experience. [Means for solving the problem]

[0007] To solve the above problems, according to one aspect of the present invention, an information processing device is provided which includes a control unit that generates control information used by a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters relating to the temperature at which the aerosol source is heated, wherein the control unit collects a plurality of training data including a combination of first control information, an evaluation set in the first control information, and second control information to be generated based on the first control information and the evaluation set in the first control information, and generates the control information used by a first user's suction device based on a control information generation model learned based on the plurality of training data collected.

[0008] The control unit may generate modified control information to be used by the first user's suction device by inputting the original control information used by the first user and the evaluation set by the first user for the original control information into the generation model.

[0009] The training data may include the first control information used by the first user with the suction device, the evaluation set for the first control information by the first user, and the second control information set for the first user to have a better evaluation than the first control information.

[0010] The control unit may collect the training data in a process in which a customization process is repeated, which includes generating modified control information to be used by the first user's suction device based on the original control information used by the first user's suction device and the evaluation set by the first user on the original control information.

[0011] The training data includes the control information before the modification in the first customization process as the first control information, and the control information after the modification in the second customization process as the second control information, and the second customization process may be the same as the first customization process or a customization process repeated after the first customization process.

[0012] The control unit may, during the process of repeating the customization process, replace the second control information included in the collected training data with the modified control information which has a better evaluation than the second control information.

[0013] The training data may include the first control information used by a suction device of a second user other than the first user, the evaluation set for the first control information by the second user, and the second control information for which the evaluation is better than that of the first control information, as set by the second user.

[0014] The aforementioned training data may further include evaluations set in the second control information.

[0015] The training data further includes information indicating the attributes of the user of the suction device using the first control information, and the control unit may generate the control information used by the first user of the suction device based on the training data which includes information indicating the same attributes as the first user's attributes.

[0016] The training data further includes information indicating the type of aerosol source heated based on the first control information, and the control unit may generate the control information used by the first user's suction device based on the training data which includes information indicating the same type as the type of aerosol source heated by the first user's suction device.

[0017] The training data further includes information indicating the type of suction device using the first control information, and the control unit may generate the control information used by the first user's suction device based on the training data which includes information indicating the same type as the first user's suction device.

[0018] Furthermore, in order to solve the above problems, according to another aspect of the present invention, there is an information processing method performed by a computer, the information processing method comprising generating control information used by a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters relating to the temperature at which the aerosol source is heated, the generating of the control information comprising collecting a plurality of training data including a combination of first control information, an evaluation set in the first control information, and second control information to be generated based on the first control information and the evaluation set in the first control information, and generating the control information used by a first user's suction device based on a control information generation model learned based on the plurality of training data collected.

[0019] Furthermore, in order to solve the above problems, according to another aspect of the present invention, a computer is provided as a control unit that generates control information used by a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters relating to the temperature at which the aerosol source is heated, the control unit collects a plurality of training data including a combination of first control information, an evaluation set in the first control information, and second control information to be generated based on the first control information and the evaluation set in the first control information, and generates the control information used by a first user's suction device based on a control information generation model learned based on the plurality of training data collected. [Effects of the Invention]

[0020] As described above, according to the present disclosure, it is possible to further improve the quality of the user experience.

Brief Description of the Drawings

[0021] [Figure 1] It is a diagram showing a configuration example of a system according to an embodiment of the present disclosure. [Figure 2] It is a schematic diagram schematically showing a configuration example of a suction device according to the same embodiment. [Figure 3] It is a block diagram showing a configuration example of a terminal device according to the same embodiment. [Figure 4] It is a block diagram showing a configuration example of a server according to the same embodiment. [Figure 5] It is a graph schematically showing an example of a heating profile. [Figure 6] It is a diagram for explaining a generation model according to the same embodiment. [Figure 7] It is a sequence diagram showing an example of the flow of a customization process executed by the system according to the same embodiment. [Figure 8] It is a flowchart showing an example of the flow of teacher data collection processing executed by the server according to the same embodiment.

Modes for Carrying Out the Invention

[0022] Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In the present specification and drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and redundant description is omitted.

[0023] Furthermore, in this specification and drawings, elements having substantially the same functional configuration may be distinguished by adding a different alphabet after the same reference numeral. For example, multiple elements having substantially the same functional configuration may be distinguished as suction device 100A and suction device 100B as needed. However, if there is no need to particularly distinguish each of multiple elements having substantially the same functional configuration, only the same reference numeral will be used. For example, if there is no need to particularly distinguish between suction device 100A and suction device 100B, they will simply be referred to as suction device 100.

[0024] <1. Example Configuration> Figure 1 shows an example configuration of System 1 according to this embodiment. As shown in Figure 1, System 1 includes a plurality of suction devices 100 (100A and 100B), a plurality of terminal devices 200 (200A and 200B), and a server 300.

[0025] The suction device 100 is a device that generates a substance to be aspirated by the user. In the following description, the substance generated by the suction device 100 will be described as an aerosol. The suction device 100 is an example of an aerosol generating device that generates an aerosol. In addition, the substance generated by the suction device may be a gas. The suction device 100 is capable of accommodating a stick-type substrate 150. The suction device 100 generates an aerosol using the accommodated stick-type substrate 150. The stick-type substrate 150 is an example of a substrate that contributes to aerosol generation. The stick-type substrate 150 contains an aerosol source. The suction device 100 generates an aerosol by heating the accommodated stick-type substrate 150.

[0026] Terminal device 200 is a device used by the user of suction device 100. Terminal device 200 is associated with suction device 100. Suction device 100 and terminal device 200 may be pre-paired for wireless communication, or the server 300 may be pre-registered to have the same user for suction device 100 and terminal device 200. Terminal device 200 may be any device such as a smartphone, tablet, wearable device, or PC (Personal Computer). Alternatively, terminal device 200 may be a charger for charging suction device 100.

[0027] Server 300 is a control device that manages information about each device included in System 1. Server 300 communicates with terminal device 200 via network 900. In particular, Server 300 communicates indirectly with suction device 100 via terminal device 200. Server 300 may perform various processes based on information collected from suction device 100 via terminal device 200. Alternatively, Server 300 may perform various processes based on user operations performed on terminal device 200.

[0028] System 1 includes multiple suction devices 100 and multiple terminal devices 200 used by multiple users. For example, a user who uses suction device 100A and terminal device 200A will also be referred to as User A. Similarly, a user who uses suction device 100B and terminal device 200B will also be referred to as User B.

[0029] (1) Example of suction device configuration Figure 2 is a schematic diagram illustrating an example configuration of the suction device 100. As shown in Figure 2, the suction device 100 according to this example configuration includes a power supply unit 111, a sensor unit 112, a notification unit 113, a storage unit 114, a communication unit 115, a control unit 116, a heating unit 121, a storage unit 140, and a heat insulation unit 144.

[0030] The power supply unit 111 stores power. Then, based on the control by the control unit 116, the power supply unit 111 supplies power to each component of the suction device 100. The power supply unit 111 may be composed of a rechargeable battery, such as a lithium-ion secondary battery.

[0031] The sensor unit 112 acquires various information related to the suction device 100. For example, the sensor unit 112 is composed of a pressure sensor such as a condenser microphone, a flow sensor, or a temperature sensor, and acquires values ​​associated with suction by the user. As another example, the sensor unit 112 is composed of an input device such as a button or switch that accepts information input from the user.

[0032] The notification unit 113 notifies the user of information. The notification unit 113 is composed of, for example, a light-emitting device that emits light, a display device that displays an image, a sound output device that emits sound, or a vibration device that vibrates.

[0033] The memory unit 114 stores various information for the operation of the suction device 100. The memory unit 114 is composed of a non-volatile storage medium, such as flash memory.

[0034] The communication unit 115 is a communication interface capable of performing communication in accordance with any wired or wireless communication standard. Examples of such communication standards include Wi-Fi®, Bluetooth®, BLE (Bluetooth Low Energy®), NFC (Near Field Communication), or LPWA (Low Power Wide Area). These may be adopted.

[0035] The control unit 116 functions as both an arithmetic processing unit and a control device, controlling the overall operation of the suction device 100 according to various programs. The control unit 116 is implemented by electronic circuits such as a CPU (Central Processing Unit) or a microprocessor.

[0036] The housing section 140 has an internal space 141 and holds the stick-type substrate 150 while housing a portion of the stick-type substrate 150 in the internal space 141. The housing section 140 has an opening 142 that communicates the internal space 141 with the outside and accommodates the stick-type substrate 150 inserted into the internal space 141 from the opening 142. For example, the housing section 140 is a cylindrical body with the opening 142 and bottom 143 as its base, defining a columnar internal space 141. An air passage is connected to the housing section 140 to supply air to the internal space 141. An air inlet, which is the air entrance to the air passage, is located, for example, on the side of the suction device 100. An air outlet, which is the air exit from the air passage to the internal space 141, is located, for example, on the bottom 143.

[0037] The stick-type base material 150 includes a base material portion 151 and a mouthpiece portion 152. The base material portion 151 includes an aerosol source. The aerosol source includes flavoring components derived from tobacco or non-tobacco. If the inhalation device 100 is a medical inhaler such as a nebulizer, the aerosol source may also include a drug. The aerosol source may be a liquid such as water, including polyhydric alcohols such as glycerin and propylene glycol, and flavoring components derived from tobacco or non-tobacco, or it may be a solid containing flavoring components derived from tobacco or non-tobacco. When the stick-type base material 150 is held in the housing portion 140, at least a part of the base material portion 151 is housed in the internal space 141, and at least a part of the mouthpiece portion 152 protrudes from the opening 142. When the user puts the mouthpiece portion 152 protruding from the opening 142 in their mouth and inhales, air flows into the internal space 141 via an air passage (not shown) and reaches the user's mouth together with the aerosol generated from the base material portion 151.

[0038] The heating unit 121 generates an aerosol by heating the aerosol source, thereby atomizing it. In the example shown in Figure 2, the heating unit 121 is configured in a film-like form and is positioned to cover the outer circumference of the containment unit 140. When the heating unit 121 generates heat, the base material portion 151 of the stick-type base material 150 is heated from the outer circumference, generating an aerosol. The heating unit 121 generates heat when power is supplied from the power supply unit 111. For example, power may be supplied when the sensor unit 112 detects that the user has started inhaling and / or that predetermined information has been input. Power may be stopped when the sensor unit 112 detects that the user has finished inhaling and / or that predetermined information has been input.

[0039] The heat insulating section 144 prevents heat transfer from the heating section 121 to other components. For example, the heat insulating section 144 is made of a vacuum insulating material or an aerogel insulating material.

[0040] The above describes an example configuration of the suction device 100. Of course, the configuration of the suction device 100 is not limited to the above, and it can take various configurations as exemplified below.

[0041] As an example, the heating section 121 may be configured in a blade shape and positioned to protrude from the bottom 143 of the housing section 140 into the internal space 141. In this case, the blade-shaped heating section 121 is inserted into the base material portion 151 of the stick-shaped base material 150 and heats the base material portion 151 of the stick-shaped base material 150 from the inside. As another example, the heating section 121 may be positioned to cover the bottom 143 of the housing section 140. Furthermore, the heating section 121 may be configured as a combination of two or more of the following: a first heating section covering the outer circumference of the housing section 140, a blade-shaped second heating section, and a third heating section covering the bottom 143 of the housing section 140.

[0042] As another example, the housing section 140 may include an opening and closing mechanism, such as a hinge, that opens and closes a part of the outer shell forming the internal space 141. The housing section 140 may then house the stick-shaped base material 150 inserted into the internal space 141 while clamping it by opening and closing the outer shell. In this case, the heating section 121 may be provided at the clamping location in the housing section 140 and may heat the stick-shaped base material 150 while pressing it.

[0043] Furthermore, the means for atomizing the aerosol source is not limited to heating by the heating unit 121. For example, the means for atomizing the aerosol source may be induction heating. In that case, the suction device 100 has at least an electromagnetic induction source, such as a coil that generates a magnetic field, instead of the heating unit 121. The susceptor that generates heat by induction heating may be provided in the suction device 100 or may be included in the stick-type substrate 150.

[0044] Furthermore, the suction device 100 works in cooperation with the stick-type substrate 150 to generate an aerosol that is aspirated by the user. Therefore, the combination of the suction device 100 and the stick-type substrate 150 may be considered as an aerosol generation system.

[0045] (2) Example of terminal device configuration Figure 3 is a block diagram showing an example configuration of the terminal device 200 according to this embodiment. As shown in Figure 3, the terminal device 200 includes an input unit 210, an output unit 220, a detection unit 230, a communication unit 240, a storage unit 250, and a control unit 260.

[0046] The input unit 210 has the function of receiving various types of information. The input unit 210 may include an input device that receives information from the user. Examples of input devices include buttons, keyboards, touch panels, and microphones. In addition, the input unit 210 may include various sensors such as image sensors.

[0047] The output unit 220 has the function of outputting information. The output unit 220 may include an output device that outputs information to the user. Examples of output devices include a display device that displays information, a light-emitting device that emits light, a vibration device that vibrates, and a sound output device that emits sound. An example of a display device is a display. An example of a light-emitting device is an LED (Light Emitting Diode). An example of a vibration device is an eccentric motor. An example of a sound output device is a speaker. The output unit 220 notifies the user of information by outputting the information input from the control unit 260.

[0048] The detection unit 230 has the function of detecting information related to the terminal device 200. The detection unit 230 may detect the location information of the terminal device 200. For example, the detection unit 230 receives GNSS signals from GNSS (Global Navigation Satellite System) satellites (for example, GPS signals from GPS (Global Positioning System) satellites) and detects the location information consisting of the latitude and longitude of the device. The detection unit 230 may also detect the movement of the terminal device 200. For example, the detection unit 230 includes a gyro sensor and an accelerometer and detects angular velocity and acceleration.

[0049] The communication unit 240 is a communication interface for sending and receiving information between the terminal device 200 and other devices. The communication unit 240 performs communication in accordance with any wired or wireless communication standard. Examples of such communication standards include those using USB (Universal Serial Bus), Wi-Fi (registered trademark), Bluetooth (registered trademark), NFC (Near Field Communication), or LPWA (Low Power Wide Area).

[0050] The memory unit 250 stores various types of information. The memory unit 250 is composed of a non-volatile storage medium, such as flash memory.

[0051] The control unit 260 functions as an arithmetic processing unit or control unit, and controls the overall operation within the terminal device 200 according to various programs. The control unit 260 is implemented by electronic circuits such as a CPU (Central Processing Unit) or a microprocessor. The unit 260 may include a ROM (Read Only Memory) for storing the program and calculation parameters to be used, and a RAM (Random Access Memory) for temporarily storing parameters that change as needed. The terminal device 200 performs various processes based on the control of the control unit 260. Processing of information input by the input unit 210, output of information by the output unit 220, detection of information by the detection unit 230, transmission and reception of information by the communication unit 240, and storage and reading of information by the storage unit 250 are examples of processes controlled by the control unit 260. Other processes performed by the terminal device 200, such as inputting information to each component and processing based on information output from each component, are also controlled by the control unit 260.

[0052] The functions of the control unit 260 may be implemented using an application. This application may be pre-installed or downloaded. Furthermore, the functions of the control unit 260 may be implemented using a Progressive Web App (PWA).

[0053] (3) Example of server configuration Figure 4 is a block diagram showing an example configuration of the server 300 according to this embodiment. As shown in Figure 4, the server 300 includes a communication unit 310, a storage unit 320, and a control unit 330.

[0054] The communication unit 310 is a communication interface for sending and receiving information between the server 300 and other devices. The communication unit 310 performs communication in accordance with any wired or wireless communication standard.

[0055] The storage unit 320 stores various information necessary for the operation of the server 300. The storage unit 320 is composed of non-volatile storage media such as an HDD (Hard Disk Drive) and an SSD (Solid State Drive).

[0056] The control unit 330 functions as an arithmetic processing unit and control unit, and controls the overall operation within the server 300 according to various programs. The control unit 330 is implemented by electronic circuits such as a CPU (Central Processing Unit) and a microprocessor. 330 may include a ROM (Read Only Memory) for storing the program and calculation parameters to be used, and a RAM (Random Access Memory) for temporarily storing parameters that change as needed. The server 300 executes various processes based on the control of the control unit 330. Sending and receiving information by the communication unit 310, and storing and reading information by the storage unit 320 are examples of processes controlled by the control unit 330. Other processes performed by the server 300, such as inputting information to each component and processing based on information output from each component, are also controlled by the control unit 330.

[0057] <2. Technical Features> (1) Heating profile The control unit 116 controls the operation of the heating unit 121 based on the heating profile. The control of the operation of the heating unit 121 is achieved by controlling the power supply from the power supply unit 111 to the heating unit 121. The heating unit 121 heats the stick-type substrate 150 using the power supplied from the power supply unit 111.

[0058] A heating profile is control information for controlling the temperature at which an aerosol source is heated. The heating profile defines parameters related to the temperature at which the aerosol source is heated. An example of the temperature at which the aerosol source is heated is the temperature of the heating unit 121. An example of parameters related to the temperature at which the aerosol source is the target temperature of the heating unit 121 (hereinafter also referred to as the target temperature). The temperature of the heating unit 121 may be controlled to change according to the elapsed time since the start of heating. In that case, the heating profile includes information that defines the time-series change of the target temperature. As another example, the heating profile may include parameters that define the method of supplying power to the heating unit 121 (hereinafter also referred to as power supply parameters). Power supply parameters include, for example, the voltage applied to the heating unit 121, ON / OFF of the power supply to the heating unit 121, or the method of feedback control to be adopted. ON / OFF of the power supply to the heating unit 121 may be considered as ON / OFF of the heating unit 121.

[0059] The control unit 116 controls the operation of the heating unit 121 so that its temperature (hereinafter also referred to as the actual temperature) progresses in a manner similar to the target temperature defined in the heating profile. The heating profile is typically designed to optimize the flavor the user experiences when inhaling the aerosol generated from the stick-type substrate 150. Therefore, by controlling the operation of the heating unit 121 based on the heating profile, the flavor the user experiences can be optimized.

[0060] Temperature control of the heating unit 121 can be achieved, for example, by known feedback control. The feedback control may be, for example, PID control (Proportional-Integral-Differential Controller). The control unit 116 may supply power from the power supply unit 111 to the heating unit 121 in the form of pulses obtained by pulse width modulation (PWM) or pulse frequency modulation (PFM). In this case, the control unit 116 can control the temperature of the heating unit 121 by adjusting the duty cycle or frequency of the power pulse in the feedback control. Alternatively, the control unit 116 may perform simple on / off control in the feedback control. For example, the control unit 116 may perform heating by the heating unit 121 until the actual temperature reaches the target temperature, interrupt heating by the heating unit 121 when the actual temperature reaches the target temperature, and resume heating by the heating unit 121 when the actual temperature falls below the target temperature.

[0061] The temperature of the heating section 121 can be quantified, for example, by measuring or estimating the electrical resistance of the heating section 121 (more precisely, the heat-generating resistor constituting the heating section 121). This is because the electrical resistance of the heat-generating resistor changes with temperature. The electrical resistance of the heat-generating resistor can be estimated, for example, by measuring the voltage drop across the heat-generating resistor. The voltage drop across the heat-generating resistor can be measured by a voltage sensor that measures the potential difference applied to the heat-generating resistor. In another example, the temperature of the heating section 121 can be measured by a temperature sensor, such as a thermistor, installed near the heating section 121.

[0062] The period from the start to the end of the process of generating aerosols using the stick-type substrate 150 will hereafter be referred to as the heating session. In other words, the heating session is the period during which power supply to the heating unit 121 is controlled based on the heating profile. The start of the heating session is the timing when heating based on the heating profile begins. The end of the heating session is the timing when a sufficient amount of aerosol is no longer generated. The heating session includes a preheating period in the first half and a puffing period in the second half. The puffing period is the period during which a sufficient amount of aerosol is expected to be generated. The preheating period is the period from the start of heating until the start of the puffing period. The heating performed during the preheating period is also referred to as preheating.

[0063] The notification unit 113 may notify the user of information indicating when preheating is complete. For example, the notification unit 113 may notify the user of information indicating the end of preheating before it is completed, or notify the user of information indicating that preheating is complete when it is completed. Notification to the user may be made by, for example, the lighting of an LED or vibration. The user can then perform puffing immediately after preheating is complete by referring to such notification.

[0064] Similarly, the notification unit 113 may notify the user of information indicating when the puffable period will end. For example, the notification unit 113 may notify the user of information that will precede the end of the puffable period before it ends, or notify the user of information indicating that the puffable period has ended when it has ended. Notification to the user may be made, for example, by the lighting or vibration of an LED. The user can then use such a notification as a reference to continue puffing until the puffable period ends.

[0065] An example of a heating profile will be explained with reference to Figure 5. Figure 5 is a schematic graph showing an example of a heating profile. The horizontal axis of Graph 20 is time. The vertical axis of Graph 20 is temperature. Line 21 shows the time series change of the target temperature. As shown in Figure 5, a heating session may sequentially include an initial heating period, an intermediate cooling period, and a reheating period. The initial heating period is the period after the start of heating during which the temperature of the heating section 121 rises rapidly and is maintained at a high temperature. The intermediate cooling period is the period after the initial heating period during which the temperature of the heating section 121 decreases. The reheating period is the period after the intermediate cooling period during which the temperature of the heating section 121 rises again. In the example shown in Figure 5, the target temperature rises rapidly to around 300°C during the initial heating period, then decreases to around 230°C during the intermediate cooling period, and then gradually rises to around 260°C during the reheating period. During the intermediate cooling period, power supply to the heating unit 121 may be interrupted and heating may be turned OFF. In the example shown in Figure 5, the preheating period is from the start of heating until partway through the initial heating period, and the puffing period is from partway through the initial heating period until the end of the reheating period.

[0066] (2) Customization process System 1 repeatedly performs a customization process. The customization process is the process of customizing (i.e., changing) the heating profile. In the customization process, System 1 modifies the heating profile to improve user evaluation. Therefore, by repeating the customization process, System 1 can gradually generate a heating profile that can provide the optimal user experience. The customization process is executed or controlled by the suction device 100, the terminal device 200, and the server 300, respectively.

[0067] The customization process includes at least the following: the suction device 100 generating an aerosol using a heating profile; setting an evaluation period; accepting evaluation settings from the user; modifying the heating profile based on the set evaluation; and setting the modified heating profile in the suction device 100. The customization process may be repeated until a heating profile that meets the user's intentions is generated. A heating profile that meets the user's intentions is one that receives a good evaluation throughout the entire heating session (i.e., for all puffs). Each process included in the customization process is described in detail below.

[0068] - Aerosol generation based on heating profile The suction device 100 generates an aerosol by heating the stick-shaped substrate 150 based on the heating profile (hereinafter also referred to as the heating profile before modification). The user inhales the aerosol generated by the suction device 100 to check the inhalation sensation. The user may take multiple puffs during the heating session.

[0069] The timing for performing a puff (hereinafter referred to as the puff timing) may be set in advance. In this case, the user performs a puff at the pre-set puff timing. For example, the terminal device 200 obtains information indicating the progress of heating from the suction device 100 and prompts the user to perform a puff at a predetermined timing during the heating session. The information indicating the progress of heating may include the elapsed time since the start of heating, or the temperature of the heating unit 121, etc. The terminal device 200 may obtain identification information of the heating profile used by the suction device 100 from the suction device 100, either together with or prior to the information indicating the progress of heating. This makes it possible to appropriately determine the arrival of the puff timing even if the puff timing differs for each heating profile. Of course, the puff timing does not have to be set in advance. In this case, the user performs a puff at any time they like. The suction device 100 may transmit information to the terminal device 200 to identify the actual puff timing. The information to identify the puff timing may be information indicating which puff has been performed during the heating session, or it may be information that identifies the puff timing by the elapsed time since the start of heating. Information for identifying the puffing timing may be transmitted as part of the information indicating the progress of heating.

[0070] - Setting the evaluation period The terminal device 200 divides the heating session and sets multiple evaluation periods. An evaluation period is a period during which the user is subject to evaluation. For example, the terminal device 200 sets the evaluation period based on the identification information of the heating profile used by the suction device 100 and information indicating the progress of heating.

[0071] The evaluation period may include multiple puff timings. That is, the user may set evaluations for multiple puffs at once. The puff timing here may be a pre-set puff timing or the actual puff timing. With this configuration, it becomes possible to roughly customize the heating profile. As a result, it is possible to reduce the burden on the user compared to setting evaluations for each puff.

[0072] Of course, the evaluation period may include one puff timing. That is, the user may set an evaluation for each puff. With such a configuration, it becomes possible to finely customize the heating profile.

[0073] The terminal device 200 may set the evaluation period based on the elapsed time from the start of heating. For example, the terminal device 200 may divide the puffing period into 30-second intervals and set multiple 30-second evaluation periods.

[0074] The terminal device 200 may set the evaluation period based on the number of puff timings. For example, the terminal device 200 may divide the puffable period for each puff timing and set an evaluation period for each puff timing. With such a configuration, it is possible to set the evaluation period appropriately even if the user's puff intervals are uneven.

[0075] - Acceptance of rating settings The terminal device 200 accepts user input to set an evaluation of the heating profile before the change. More specifically, the terminal device 200 accepts the setting of an evaluation of the aerosol inhaled by the user during each of several evaluation periods. In the following, the terminal device 200 accepts the setting of evaluations for each puff, which are performed multiple times during the heating session. For example, the terminal device 200 displays a screen for accepting the setting of evaluations for each puff and accepts tap operations on the screen. The evaluation set by the user is used to change the heating profile. That is, accepting the setting of an evaluation may be understood as accepting the setting of an instruction to change the heating profile (change value described later).

[0076] The terminal device 200 may accept evaluation settings in real time according to the progress of heating. If the terminal device 200 accepts evaluation settings in real time, it may obtain information indicating the progress of heating from the suction device 100 and prompt the user to set the evaluation immediately after puffing is performed. With this configuration, the user can set the evaluation for each puff in real time while puffing is being performed. Of course, the terminal device 200 may also accept a user operation to set the evaluation for each puff all at once after the heating session has ended.

[0077] The terminal device 200 may accept settings for evaluations of multiple evaluation items. With such a configuration, it becomes possible to improve evaluations from various perspectives. Examples of evaluation items include taste, smoke volume, tobacco feel, throat hit, odor, and smoking satisfaction. Taste refers to the overall taste of the aerosol. The stronger the taste, the more taste it is rated as; the weaker the taste, the less taste it is rated as. Smoke volume refers to the amount of aerosol. The more aerosol that reaches the user's mouth per puff, the more smoke it is rated as; the less aerosol that reaches the user's mouth per puff, the less smoke it is rated as. Tobacco feel refers to the closeness to the taste of a cigarette. The closer the taste of the aerosol itself or the intensity of the taste is to the taste of a cigarette, the stronger the tobacco feel is rated. On the other hand, the more refreshing the taste of the aerosol, due to factors such as a strong fruit or mint flavor, the weaker the tobacco feel is rated. Throat hit refers to the degree of stimulation to the throat. Typically, a higher nicotine content in the aerosol is rated as the stronger the throat hit. "Odor" refers to the degree of similarity to the smell of a cigarette. The closer the aerosol's odor is to that of a cigarette, the stronger the odor is rated. Conversely, the more refreshing the aerosol's odor (e.g., due to strong fruit or mint scents), the weaker the odor is rated. "Smoothness" refers to the degree of stimulation to the entire mouth. For these evaluation items, a positive rating (just right) or a negative rating (weak / strong, little / a lot) may be assigned.

[0078] In addition, the terminal device 200 may accept the setting of an evaluation for the entire heating session (i.e., the entirety of the multiple puffs performed during the heating session). For example, after the heating session ends, the terminal device 200 may ask a question such as, "Were you satisfied with this heating profile?" In that case, a positive evaluation such as "satisfied" or a negative evaluation such as "not satisfied" may be set.

[0079] - Change heating profile Server 300 (for example, control unit 330) is an example of an information processing device that generates a heating profile. Server 300 generates a new heating profile (hereinafter also referred to as the modified heating profile) by modifying the original heating profile based on evaluations set by the user. For example, Server 300 raises the target temperature at puff timings evaluated as having less flavor and lowers the target temperature at puff timings evaluated as having more flavor. With this configuration, it is possible to generate a modified heating profile that can improve the evaluation compared to the original heating profile.

[0080] When evaluations are set for multiple evaluation items, the server 300 generates a modified heating profile based on the evaluations for the multiple evaluation items. For example, the server 300 generates a modified heating profile by averaging and integrating the multiple target temperature change values ​​based on the evaluations for multiple evaluation items, and then applying this to the original heating profile. For example, if the change value based on the evaluation of taste is +30°C and the change value based on the evaluation of smoke volume is +10°C, the average of these, +20°C, may be adopted as the integrated change value. Then, the modified heating profile may be generated by raising the original target temperature by 20°C. With this configuration, it is possible to improve evaluations from various perspectives.

[0081] Server 300 may generate a modified heating profile based on a trained generative model for generating heating profiles. The generative model will be described with reference to Figure 6.

[0082] Figure 6 is a diagram illustrating the generation model according to this embodiment. As shown in Figure 6, the generation model M outputs a modified heating profile when it receives the original heating profile and the evaluation set for the original heating profile as input. The heating profile generation model M may be a model trained using known machine learning techniques such as SVM (Support Vector Machine) or a neural network. With this configuration, it is possible to automatically and accurately generate heating profiles that improve user evaluations. The accuracy of heating profile generation refers to the degree to which the generated heating profile matches the user's intentions. The higher the accuracy of heating profile generation, the higher the user evaluation set for the generated heating profile. The ability to easily generate and provide heating profiles that match the user's intentions greatly improves the quality of the user experience.

[0083] -Setting the modified heating profile The suction device 100 sets the modified heating profile. For example, the suction device 100 receives the modified heating profile generated by the server 300 via the terminal device 200 and stores it. This is expected to improve user evaluation in the next customization process.

[0084] - Repeat the customization process The above describes in detail each process included in the customization process. System 1 repeatedly executes the customization process described above until a heating profile matching the user's intentions is generated. The repetition of the customization process will be explained below with reference to Table 1.

[0085] [Table 1]

[0086] In Table 1, P represents the heating profile, E represents the evaluation, and the numbers following P and E indicate the index corresponding to the number of iterations of the customization process. According to Table 1, in the first customization process, heating profile P1 and the evaluation E1 set for heating profile P1 are input to the generation model, and heating profile P2 is output. Heating profile P2 is a heating profile to which changes have been made to heating profile P1 to improve the poor evaluation included in evaluation E1. This customization process is repeated, with the heating profile generated in the previous customization process being used as input to the generation model in the next customization process. In the 100th customization process, if the evaluation E100 set for heating profile P100 is a good evaluation for all puffs, the generation model outputs the input heating profile P100 as is, as shown in Table 1. Then, the iteration of the customization process stops. In this way, heating profile P100 that matches the user's intentions is generated.

[0087] (3) Training of generative models Server 300 collects multiple training data sets and learns a generative model for generating heating profiles based on the collected training data. Server 300 then generates heating profiles based on the learned generative model. The training data includes a combination of a first heating profile, the evaluation set for the first heating profile, and a second heating profile that should be generated based on the first heating profile and the evaluation set for the first heating profile. In other words, the training data is a desirable combination of the original heating profile and the evaluation set for the original heating profile, which are inputs to the generative model, and the modified heating profile, which is the output from the generative model. By collecting such training data, it is possible to learn a highly accurate generative model. The accuracy of the generative model corresponds to the accuracy of the heating profiles generated using the generative model.

[0088] In the following, it is assumed that server 300 generates a heating profile for user A (an example of the first user). The heating profile for user A is the heating profile used by the suction device 100A used by user A.

[0089] In this case, the server 300 inputs the original heating profile used by the suction device 100A and the evaluation set for the original heating profile by user A into the generation model. This allows the server 300 to generate the modified heating profile used by the suction device 100A. With this configuration, the server 300 can automatically and accurately generate the heating profile for user A using the generation model during the customization process. As a result, the heating profile that matches the user's intentions can be generated more quickly, reducing the number of iterations required for the customization process.

[0090] The training data used to train the generative model for generating a heating profile for user A may be training data in which user A was involved. Specifically, the training data may include a first heating profile used by the suction device 100A, an evaluation set by user A for the first heating profile, and a second heating profile in which user A set a better evaluation than the first heating profile. By generating a heating profile for user A using a generative model trained on training data in which user A was involved, it becomes possible to improve user A's evaluation more efficiently.

[0091] Server 300 may collect training data involving user A during the process of repeatedly performing a customization process to generate a heating profile for user A. The customization process to generate a heating profile for user A includes generating a modified heating profile to be used by the suction device 100A based on the original heating profile used by the suction device 100A and the evaluation set by user A for the original heating profile. With this configuration, the relationship between the change in target temperature and the change in evaluation, obtained during the trial-and-error process in which the user arrives at the desired heating profile, can be used as training data. This makes it possible to improve the learning efficiency of the generative model.

[0092] The training data may include the heating profile before modification in the first customization process as the first heating profile, and the heating profile after modification in the second customization process as the second heating profile. Here, the second customization process is a customization process that is the same as the first customization process or repeated after the first customization process. In the example shown in Table 1, as an example, if evaluation E2 is an improvement over evaluation E1, the server 300 may collect training data including heating profile P1 as the first heating profile, evaluation E1, and heating profile P2 as the second heating profile. As another example, the server 300 may collect training data including heating profile P1 as the first heating profile, evaluation E1, and heating profile P100 as the second heating profile. With such a configuration, it is possible to efficiently collect training data while repeating the customization process.

[0093] During the iterative customization process, Server 300 may replace a second heating profile included in the collected training data with a modified heating profile that has a better evaluation than the second heating profile. For example, suppose Server 300 has collected training data including heating profile P1 as the first heating profile, evaluation E1, and heating profile P2 as the second heating profile. Subsequently, if evaluation E3 is improved compared to evaluation E2, Server 300 may replace heating profile P2 as the second heating profile in the collected training data with heating profile P3. That is, Server 300 may update the training data including heating profile P1, evaluation E1, and heating profile P2 to training data including heating profile P1, evaluation E1, and heating profile P3. If the training data is updated during subsequent iterative customization processes, Server 300 will ultimately collect training data including heating profile P100 as the second heating profile. For example, the server 300 can collect training data including heating profile P1, evaluation E1, and heating profile P100. With this configuration, the collected training data can be updated to training data more suitable for learning. For example, a generative model trained on the updated training data will, upon input of heating profile P1 and evaluation E1, output a heating profile P100 that matches the user's intention. In this way, it is possible to improve the accuracy of the generative model.

[0094] Alternatively, training data may be collected that includes heating profiles during the iteration of the customization process as the first heating profile. For example, server 300 may collect training data that includes heating profile P2, evaluation E2, and heating profile P100.

[0095] (4) Processing flow - Customization process Figure 7 is a sequence diagram showing an example of the flow of customization processing performed by System 1 according to this embodiment. This sequence involves the suction device 100, the terminal device 200, and the server 300.

[0096] As shown in Figure 7, first the suction device 100 heats the stick-shaped substrate 150 based on the heating profile (step S102).

[0097] Next, the suction device 100 transmits identification information of the heating profile used for heating to the terminal device 200 (step S104).

[0098] Next, the terminal device 200 accepts the evaluation settings (step S106). Specifically, during the heating session, the suction device 100 transmits information indicating the progress of heating to the terminal device 200. Then, the terminal device 200 prompts the user to perform puffing at predetermined timings according to the progress of heating, prompts the user to set the evaluation immediately after puffing, and accepts the evaluation settings for each puff from the user.

[0099] Next, the terminal device 200 transmits to the server 300 identification information of the heating profile used by the suction device 100 for heating, and information indicating the evaluation set by the user (step S108). The information indicating the evaluation set by the user includes information for identifying multiple puff timings, and evaluations for each evaluation item at each puff timing.

[0100] Next, the server 300 modifies the heating profile using the trained generative model (step S110). Specifically, the server 300 generates the modified heating profile by inputting the heating profile received from the terminal device 200 and the evaluation set for that heating profile into the trained generative model.

[0101] Next, the server 300 sends the modified heating profile to the terminal device 200 (step S112). Upon receiving the modified heating profile from the server 300, the terminal device 200 forwards the received modified heating profile to the suction device 100 (step S114).

[0102] Then, when the suction device 100 receives the modified heating profile, it stores the received modified heating profile (step S116). As a result, in the next customization process, the stick-type substrate 150 will be heated based on the modified heating profile.

[0103] - Training data collection process Figure 8 is a flowchart showing an example of the flow of the training data collection process performed by the server 300 according to this embodiment.

[0104] As shown in Figure 8, first, the server 300 obtains the heating profile before modification, the evaluation of the heating profile before modification, the heating profile after modification, and the evaluation of the heating profile after modification during the repeated customization process (step S202). For example, the server 300 receives heating profile P1, evaluation E1, heating profile P2, and evaluation E2 from the terminal device 200.

[0105] Next, the server 300 determines whether the evaluation has improved before and after the change in the heating profile (step S204). For example, the server 300 determines that the evaluation has improved if the number of puffs that received a better evaluation than evaluation E1 in evaluation E2 has increased, and that it has not improved otherwise. If it is determined that the evaluation has not improved before and after the change in the heating profile (step S204: NO), the process proceeds to step S210.

[0106] If it is determined that the evaluation has improved before and after the change in the heating profile (step S204: YES), the server 300 generates training data (step S206). For example, the server 300 generates training data that includes heating profile P1 as the first heating profile, evaluation E1, and heating profile P2 as the second heating profile. If, during the process of repeated customization, a heating profile P100 is obtained with a better evaluation than heating profile P2, the server 300 may replace the second heating profile in the above training data with heating profile P100. That is, the server 300 may generate training data that includes heating profile P1 as the first heating profile, evaluation E1, and heating profile P100 as the second heating profile.

[0107] Next, the server 300 trains the generative model (step S208). For example, the server 300 trains the generative model based on the existing training data as well as the training data newly generated in step S206.

[0108] Next, the server 300 determines whether the repetition of the customization process has finished (step S210). Examples of conditions under which the repetition of the customization process is determined to have finished include all puffs receiving a good rating, the entire heating session receiving a good rating, and the user instructing the server to terminate the process.

[0109] If it is determined that the customization process has not been repeated (step S210: NO), the process returns to step S202. On the other hand, if it is determined that the customization process has been repeated (step S210: NO), the process terminates.

[0110] <3. Supplement> While preferred embodiments of the present disclosure have been described in detail above with reference to the attached drawings, the present disclosure is not limited to such examples. It is clear to any person with ordinary skill in the art to which the present disclosure pertains that various modifications or alterations may be conceived within the scope of the technical ideas described in the claims, and these will naturally be understood to fall within the technical scope of the present disclosure.

[0111] In the above embodiment, an example was described in which the training data used to train a generative model for generating a heating profile for user A is training data in which user A was involved, but this disclosure is not limited to such an example. The training data used to train a generative model for generating a heating profile for user A may include training data in which other users (e.g., user B) were involved. The training data in which user B was involved includes a first heating profile used by the suction device 100B used by user B, an evaluation set by user B for the first heating profile, and a second heating profile in which user B set a better evaluation than the first heating profile. The other users are not limited to one person, and training data in which multiple other users were involved may be used to train a generative model for generating a heating profile for user A. By training the generative model based on training data in which user A was involved and / or training data in which user B was involved, it is possible to further improve the accuracy of the generative model because it is possible to increase the amount of training data.

[0112] The above describes an example where one user (User A or User B) is involved in a single training data set, but this disclosure is not limited to such an example. Multiple users may be involved in a single training data set. For example, the user involved in the first heating profile may be different from the user involved in the second heating profile. As an example, the training data may include the top-ranked heating profiles on a web page where heating profiles are made publicly available for download as the second heating profile. As another example, the training data may include the heating profile that is used by a large number of users or has been rated as satisfactory by many users as the second heating profile. Such a configuration makes it possible to further improve the accuracy of the generative model.

[0113] In the embodiments described above, an example was described in which the training data includes a first heating profile, an evaluation set for the first heating profile, and a second heating profile, but this disclosure is not limited to such an example. Examples of other information that the training data may include are described below.

[0114] The training data may further include the evaluations set for the second heating profile. For example, the training data may include heating profile P1, evaluation E1, heating profile P2, and evaluation E2. With such a configuration, the difference between the first heating profile and the second heating profile can be associated with the difference in the evaluations set for these heating profiles. As a result, the server 300 can grasp the causal relationship between the changes in the heating profile and the changes in the evaluation in more detail and utilize this in generating the heating profile. For example, suppose the difference between heating profile P1 and heating profile P2 is that the target temperature at the third puff timing was increased by 10°C. Also, suppose the difference between evaluation E1 and evaluation E2 is that the evaluation of the taste for the third puff improved from weak to just right. In that case, the server 300 can grasp the detailed causal relationship that increasing the target temperature at the third puff timing by 10°C improved the evaluation of the taste for the third puff from weak to just right. By clarifying such causal relationships, it becomes possible to generate a more accurate heating profile. The generative model can be trained using the first heating profile, the evaluation set for the first heating profile, and the evaluation set for the second heating profile as inputs, and the second heating profile as the output. When the pre-modified heating profile, the evaluation set for the pre-modified heating profile, and the desired evaluation (for example, a good evaluation for all puffs) are input to the trained generative model in this way, a modified heating profile in which the desired evaluation can be set will be generated.

[0115] The training data may further include information indicating the user attributes of the suction device 100 using the first heating profile. The server 300 may then generate a heating profile for user A based on the training data which includes information indicating the same attributes as user A. That is, the server 300 may learn a generative model based on the training data which includes information indicating the same attributes as user A, and then use the learned generative model to generate a heating profile for user A. Examples of user attributes include gender, age, and place of residence. The user attributes may also include user browsing information on the web page where the heating profiles are made publicly available for download. An example of browsing information is an HTTP cookie. The suction device 100 can download and use the heating profile from the web page via the terminal device 200. With this configuration, it is possible to further improve the accuracy of the generative model according to the user's attributes.

[0116] The training data may further include information indicating the type of aerosol source heated based on the first heating profile, i.e., the type of stick-type substrate 150. The server 300 may then generate a heating profile for user A based on training data that includes information indicating the same type as the stick-type substrate 150 heated by the suction device 100A. As an example, suppose the suction device 100A heats a menthol-containing stick-type substrate 150. In this case, the server 300 learns a generative model based on training data collected when using a menthol-containing stick-type substrate 150. The server 300 then uses the learned generative model to generate a heating profile for user A and for a menthol-containing stick-type substrate 150. With this configuration, it is possible to further improve the accuracy of the generative model according to the type of stick-type substrate 150 used by the suction device 100A.

[0117] The training data may further include information indicating the type of suction device 100 that uses the first heating profile. The server 300 may then generate a heating profile for user A based on training data that includes information indicating the same type as suction device 100A. For example, if suction device 100A is a high-heating type, the server 300 learns a generative model based on training data collected when a high-heating type suction device 100 was used. The server 300 then uses the learned generative model to generate a heating profile for user A and for the high-heating type suction device 100A. Note that the type of suction device 100 may be the type of software (e.g., software version) in addition to the type of hardware. With this configuration, it is possible to further improve the accuracy of the generative model according to the type of suction device 100A.

[0118] The above explains some examples of other information that training data may contain.

[0119] The above embodiment describes an example in which training data is collected for the entire heating profile, but this disclosure is not limited to such an example. Training data may be collected for only a portion of the heating profile. For example, suppose a total of 15 puffs are performed during a heating session, and the customization process improves the evaluation of 10 of the 15 puffs. In this case, training data may be collected that includes the portion of the original heating profile corresponding to the 10 improved puffs, the evaluation of the original heating profile corresponding to the 10 improved puffs, and the portion of the modified heating profile corresponding to the 10 improved puffs. With such a configuration, training data can be collected even if the customization process is interrupted midway.

[0120] Each process performed by the terminal device 200 or server 300 as described in the above embodiment may be performed by any device. For example, training the generative model or modifying the heating profile may be performed by the terminal device 200.

[0121] In the above embodiment, an example was described in which evaluations are set for each puff for multiple evaluation items, that is, an example in which an evaluation period common to multiple evaluation items is set. However, this disclosure is not limited to such an example. The terminal device 200 may set multiple evaluation periods for each of the multiple evaluation items. For example, the terminal device 200 may set an evaluation period every 30 seconds for the taste and an evaluation period for each puff for the amount of smoke. With such a configuration, the evaluation period for each evaluation item can be flexibly set, thereby improving the ease of customization.

[0122] In the embodiments described above, changing the target temperature was given as an example of changing the heating profile, but this disclosure is not limited to such examples. Server 300 may change the time parameters of the heating profile. Examples of time parameters of the heating profile include the duration of the heating session, the duration of the initial heating period, the intermediate cooling period, and the duration of the reheating period. Another example of a time parameter of the heating profile is the puff timing.

[0123] In the above embodiment, an example was described in which the parameter relating to the temperature for heating the aerosol source, as defined in the heating profile, is a target value for the temperature of the heating unit 121. However, this disclosure is not limited to such an example. An example of a parameter relating to the temperature for heating the aerosol source is a target value for the electrical resistance of the heating unit 121. Furthermore, if the means for heating the aerosol source is induction heating, an example of a parameter relating to the temperature for heating the aerosol source, as defined in the heating profile, is a target value such as the temperature of the susceptor or the electrical resistance of the electromagnetic induction source.

[0124] In the above embodiment, an example was described in which the suction device 100 generates an aerosol by heating a stick-type substrate 150, but the present disclosure is not limited to such an example. The suction device 100 may be configured as a so-called liquid atomization type aerosol generator, which generates an aerosol by heating and atomizing an aerosol source as a liquid. The technology of the present disclosure can also be applied to liquid atomization type aerosol generators.

[0125] As described in the above embodiment, the evaluation settings are received by the terminal device 200. Here, the terminal device 200 receiving the evaluation settings may refer to receiving the evaluation settings via a native application installed on the terminal device 200. Alternatively, the terminal device 200 receiving the evaluation settings may refer to receiving the evaluation settings via a Progressive Web App (PWA) provided for the terminal device 200. For example, the server 300 may receive the evaluation settings via a PWA provided for the terminal device 200.

[0126] In the above embodiment, at least a portion of the functional configuration of the suction device 100 may be provided by other devices. An example of such other devices is a charging device for charging the suction device 100. The charging device has a mechanism for attaching and detaching the suction device 100, and can charge the suction device 100 or send and receive information with the suction device 100 while it is connected. For example, the charging device may have a wireless communication function and may relay information between the suction device 100 and a device such as a smartphone. As another example, the charging device may have a memory function and may store information received from or to be sent to the suction device 100. The combination of the suction device 100 and the charging device may be considered as an aerosol generation system. Furthermore, at least a portion of the functional configuration of the terminal device 200 described in the above embodiment may be provided by other devices such as a charging device for charging the suction device 100.

[0127] The series of processes performed by each device described herein may be implemented using software, hardware, or a combination of software and hardware. The programs constituting the software are pre-stored on a recording medium (more specifically, a non-temporary storage medium readable by a computer) located inside or outside each device. Each program is then loaded into RAM when executed by a computer controlling each device described herein, and executed by a processing circuit such as a CPU. The recording medium is, for example, a magnetic disk, an optical disk, a magneto-optical disk, or flash memory. The computer program may also be distributed via a network, for example, without using a recording medium. The computer may be an application-specific integrated circuit such as an ASIC, a general-purpose processor that performs functions by loading software programs, or a computer on a server used for cloud computing. Furthermore, the series of processes performed by each device described herein may be distributed and processed by multiple computers.

[0128] Furthermore, the processes described herein using flowcharts or sequence diagrams do not necessarily have to be executed in the order shown. Some processing steps may be executed in parallel. Additional processing steps may be adopted, and some processing steps may be omitted.

[0129] Furthermore, the following configurations also fall within the technical scope of this disclosure. (1) A control unit that generates the control information used in a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters related to the temperature at which the aerosol source is heated, Equipped with, The control unit, Collect multiple training data sets, each including a combination of a first control information, an evaluation set for the first control information, and a second control information to be generated based on the first control information and the evaluation set for the first control information. Based on the control information generation model learned from the collected training data, the control information used by the first user's suction device is generated. Information processing device. (2) The control unit generates modified control information to be used by the first user's suction device by inputting the original control information used by the first user and the evaluation set by the first user on the original control information into the generation model. The information processing device described in (1) above. (3) The training data includes the first control information used by the first user with the suction device, the evaluation set for the first control information by the first user, and the second control information set for the first user to have a better evaluation than the first control information. The information processing device described in (1) or (2) above. (4) The control unit collects the training data during a process in which a customization process is repeated, which includes generating modified control information to be used by the first user's suction device based on the original control information used by the first user's suction device and the evaluation set by the first user on the original control information. The information processing device described in (3) above. (5) The training data includes the control information before the modification in the first customization process as the first control information, and the control information after the modification in the second customization process as the second control information. The second customization process is the same as the first customization process, or a customization process that is repeated after the first customization process. The information processing device described in (4) above. (6) During the process in which the customization process is repeated, the control unit replaces the second control information included in the collected training data with the modified control information which has a better evaluation set than the second control information. The information processing device described in (5) above. (7) The training data includes the first control information used by a suction device of a second user other than the first user, the evaluation set for the first control information by the second user, and the second control information for which the second user has set a better evaluation than the first control information. An information processing device as described in any one of the above items (1) to (6). (8) The aforementioned training data further includes the evaluation set in the second control information, An information processing device as described in any one of the above items (1) to (7). (9) The training data further includes information indicating the user attributes of the suction device using the first control information, The control unit generates the control information used by the suction device of the first user based on the training data which includes information indicating the same attributes as the attributes of the first user. An information processing device as described in any one of the above items (1) to (8). (10) The training data further includes information indicating the type of aerosol source heated based on the first control information, The control unit generates the control information used by the first user's suction device based on the training data which includes information indicating the same type as the type of aerosol source heated by the first user's suction device. An information processing device as described in any one of the above items (1) to (9). (11) The training data further includes information indicating the type of suction device using the first control information, The control unit generates the control information used by the first user's suction device based on the training data which includes information indicating the same type as the first user's suction device. An information processing device as described in any one of the above items (1) to (10). (12) A method of information processing performed by a computer, The aforementioned information processing method is This includes generating control information used by a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters related to the temperature at which the aerosol source is heated, Generating the aforementioned control information means Collecting multiple training data including a combination of first control information, an evaluation set for the first control information, and second control information to be generated based on the first control information and the evaluation set for the first control information, Based on the control information generation model learned from the collected training data, the control information used by the first user's suction device is generated. Information processing methods, including those mentioned above. (13) Computers, A control unit that generates the control information used in a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters related to the temperature at which the aerosol source is heated, To make it function as, The control unit, Collect multiple training data sets, each including a combination of a first control information, an evaluation set for the first control information, and a second control information to be generated based on the first control information and the evaluation set for the first control information. Based on the control information generation model learned from the collected training data, the control information used by the first user's suction device is generated. program. [Explanation of symbols]

[0130] 1 System 100 Suction device 111 Power supply section 112 Sensor section 113 Notification Department 114 Storage section 115 Communications Department 116 Control Unit 121 Heating section 140 storage units 141 Interior space 142 Aperture 143 Bottom 144 Insulation section 150 Stick-type base material 151 Base material part 152 Mouthpiece 200 terminal devices 210 Input section 220 Output section 230 Detection unit 240 Communications Department 250 Storage section 260 Control Unit 300 servers 310 Communications Department 320 Storage section 330 Control Unit 900 Network

Claims

1. A control unit that generates the control information used in a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters related to the temperature at which the aerosol source is heated, Equipped with, The control unit, Collect multiple training data including a combination of first control information, an evaluation set for the first control information, and second control information to be generated based on the first control information and the evaluation set for the first control information. Based on the collected training data, the control information generation model is trained to output the second control information when the first control information and the evaluation set for the first control information are input. Based on the learned generative model, the control information used by the first user's suction device is generated. The control unit generates the control information used by the first user's suction device by inputting the original control information used by the first user's suction device and the evaluation set by the first user for the original control information into the generation model, thereby outputting the modified control information used by the first user's suction device from the generation model. Information processing device.

2. The training data includes the first control information used by the first user with the suction device, the evaluation set for the first control information by the first user, and the second control information set for the first user to have a better evaluation than the first control information. The information processing apparatus according to claim 1.

3. The control information includes an evaluation set in each of a plurality of evaluation intervals set by dividing the heating session, The number of evaluation intervals in which a good evaluation is set for the second control information is greater than the number of evaluation intervals in which a good evaluation is set for the first control information. The information processing apparatus according to claim 2.

4. The control unit collects the training data during a process in which a customization process is repeated, which includes generating modified control information to be used by the first user's suction device based on the original control information used by the first user's suction device and the evaluation set by the first user on the original control information. The information processing apparatus according to claim 2 or 3.

5. The training data includes the control information before the modification in the first customization process as the first control information, and the control information after the modification in the second customization process as the second control information. The second customization process is the same as the first customization process, or a customization process that is repeated after the first customization process. The information processing apparatus according to claim 4.

6. During the process in which the customization process is repeated, the control unit replaces the second control information included in the collected training data with the modified control information which has a better evaluation set than the second control information. The information processing apparatus according to claim 5.

7. The training data includes the first control information used by a suction device of a second user other than the first user, the evaluation set for the first control information by the second user, and the second control information for which the evaluation is better than that of the first control information, as set by the second user. The information processing apparatus according to claim 1.

8. The aforementioned training data further includes the evaluation set in the second control information, The information processing apparatus according to claim 1.

9. The training data further includes information indicating the user attributes of the suction device using the first control information, The control unit generates the control information used by the suction device of the first user based on the training data which includes information indicating the same attributes as the attributes of the first user. The information processing apparatus according to claim 1.

10. The training data further includes information indicating the type of aerosol source heated based on the first control information, The control unit generates the control information used by the first user's suction device based on the training data which includes information indicating the same type as the type of aerosol source heated by the first user's suction device. The information processing apparatus according to claim 1.

11. The training data further includes information indicating the type of suction device using the first control information, The control unit generates the control information used by the first user's suction device based on the training data which includes information indicating the same type as the first user's suction device. The information processing apparatus according to claim 1.

12. A method of information processing performed by a computer, The aforementioned information processing method is This includes generating control information used by a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters related to the temperature at which the aerosol source is heated, Generating the aforementioned control information means Collecting multiple training data including a combination of first control information, an evaluation set for the first control information, and second control information to be generated based on the first control information and the evaluation set for the first control information, Based on the collected training data, the control information generation model is trained to output the second control information when the first control information and the evaluation set for the first control information are input. To generate the control information used by the first user's suction device based on the learned generative model, Includes, Generating the control information used by the first user's suction device includes inputting the original control information used by the first user's suction device and the evaluation set by the first user on the original control information into the generation model, thereby causing the generation model to output the modified control information used by the first user's suction device. Information processing methods.

13. Computers A control unit that generates the control information used in a suction device that heats an aerosol source to generate an aerosol based on control information that defines parameters related to the temperature at which the aerosol source is heated, To make it function as, The control unit, Collect multiple training data including a combination of first control information, an evaluation set for the first control information, and second control information to be generated based on the first control information and the evaluation set for the first control information. Based on the collected training data, the control information generation model is trained to output the second control information when the first control information and the evaluation set for the first control information are input. Based on the learned generative model, the control information used by the first user's suction device is generated. The control unit generates the control information used by the first user's suction device by inputting the original control information used by the first user's suction device and the evaluation set by the first user for the original control information into the generation model, thereby outputting the modified control information used by the first user's suction device from the generation model. program.

Images