system

Abstract

The system according to this embodiment aims to efficiently utilize device resources to train large-scale language models. [Solution] The system according to the embodiment comprises a monitoring unit, a decision unit, an allocation unit, and an aggregation unit. The monitoring unit monitors the resource utilization rate of the devices. The decision unit determines the proportion of resources to be used for learning based on the data collected by the monitoring unit. The allocation unit assigns distributed parallel learning tasks to the devices based on the proportion of resources determined by the decision unit. The aggregation unit aggregates the learning data for the tasks assigned by the allocation unit.

Description

Technical Field

[0001] The technology of the present disclosure relates to a system.

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] [[ID=关于传统技术,存在难以有效利用大规模语言模型学习所需计算资源的问题。]]

[0005] The system according to the embodiment aims to efficiently utilize the resources of a device to perform learning of a large language model.

Means for Solving the Problems

[0006] The system according to this embodiment comprises a monitoring unit, a decision unit, an allocation unit, and an aggregation unit. The monitoring unit monitors the resource utilization rate of the devices. The decision unit determines the proportion of resources to be used for learning based on the data collected by the monitoring unit. The allocation unit assigns distributed parallel learning tasks to the devices based on the proportion of resources determined by the decision unit. The aggregation unit aggregates the learning data for the tasks assigned by the allocation unit. [Effects of the Invention]

[0007] The system according to this embodiment can efficiently utilize device resources to train large-scale language models. [Brief explanation of the drawing]

[0008] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Modes for carrying out the invention]

[0009] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0010] First, let's explain the terminology used in the following explanation.

[0011] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), APU (Accelerated Processing Unit), or TPU (Tensor Processing Unit).

[0012] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0013] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0014] In the following embodiments, the labeled communication I / F (Interface) is an interface including a communication processor, an antenna, and the like. The communication I / F manages communication between a plurality of computers. Examples of communication standards applicable to the communication I / F include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0015] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B". That is, "A and / or B" means that it may be only A, only B, or a combination of A and B. Also, in this specification, when expressing three or more matters connected by "and / or", the same concept as "A and / or B" is applied.

[0016] [First Embodiment] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0017] As shown in FIG. 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0018] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. Also, the database 24 and the communication I / F 26 are connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0019] The smart device 14 comprises a computer 36, a receiving device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The receiving device 38, output device 40, and camera 42 are also connected to the bus 52.

[0020] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, and accepts user input. The touch panel 38A accepts user input via touch by detecting contact with an object (e.g., a pen or finger). The microphone 38B accepts user input via voice by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 (see Figure 2) acquires the data indicating the user input.

[0021] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user by outputting the data in a form perceptible to the user (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0022] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0023] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0024] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0025] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0026] In the smart device 14, specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The specific processing program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as a control unit 46A according to the specific processing program 60 executed on the RAM 48. The smart device 14 also has a data generation model 58 and an emotion identification model 59, similar to the data generation model and emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0027] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device (e.g., a generation server) may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device having the data generation model 58. The data processing device 12 may also be a server device or a terminal device owned by a user (e.g., a mobile phone, robot, home appliance, etc.). Next, an example of processing by the data processing system 10 according to the first embodiment will be described.

[0028] (Example of form 1) The learning system according to an embodiment of the present invention is a system that utilizes distributed parallel learning and advances learning using the computing resources of people's smartphones. This learning system works by using 1 to 10% of the device's resources to advance learning, such as displaying advertisements while the user is browsing a website. For example, while a user is browsing a website, the system monitors the device's resources and determines the percentage of resources that can be used for learning based on resource usage. If the CPU and memory usage of the device is low, 10% of the resources can be used for learning. Next, the system assigns distributed parallel learning tasks to the devices, and each device advances learning in a distributed manner. For example, while a user is browsing a website, the system uses the device's resources to train a large-scale language model. Furthermore, the data that has been trained is sent to the system and aggregated. This combines the distributed trained data into one, improving the accuracy of the model. For example, each device sends the trained data to a server, which aggregates it and updates the model. This mechanism allows users to advance learning by effectively utilizing their device's resources without any special operations. In addition, by efficiently using the device's resources, the computational cost of learning can be reduced. For example, by using the device's resources to perform learning while the user is browsing a website, the need for dedicated computing resources is eliminated. This allows the learning system to efficiently utilize the user's device resources for distributed parallel learning.

[0029] The learning system according to the embodiment comprises a monitoring unit, a decision unit, an allocation unit, and an aggregation unit. The monitoring unit monitors the resource utilization of the device. The resource utilization of the device includes, but is not limited to, CPU utilization, memory utilization, and disk I / O. For example, the monitoring unit monitors the CPU utilization of the device in real time and issues an alert if the utilization exceeds a certain threshold. The monitoring unit can also monitor memory utilization and issue a warning if memory usage increases. Furthermore, the monitoring unit can monitor disk I / O usage and issue a notification if the disk read / write speed decreases. The decision unit determines the proportion of resources to be used for learning based on the data collected by the monitoring unit. For example, the decision unit decides to use 10% of the resources for learning when the CPU utilization of the device is low. The decision unit can also reduce the proportion of resources used when memory utilization is high. Furthermore, the decision unit can adjust the proportion of resources used according to the disk I / O usage. The allocation unit assigns distributed parallel learning tasks to the device based on the proportion of resources determined by the decision unit. For example, the assignment unit dynamically adjusts task assignments according to the device's performance and network conditions. The assignment unit can assign tasks considering the device's CPU clock speed, memory capacity, network bandwidth, etc. Furthermore, the assignment unit can optimize task assignments based on information such as the device's battery level and temperature. The aggregation unit aggregates the training data for tasks assigned by the assignment unit. For example, the aggregation unit sends the completed training data to a server, which aggregates it and updates the model. The aggregation unit can use encryption protocols and authentication methods to ensure secure transmission and aggregation of data. Furthermore, the aggregation unit can calculate data checksums and hash values ​​to ensure data integrity. As a result, the learning system according to the embodiment can efficiently use the user's device resources to perform distributed parallel learning.

[0030] The monitoring unit can monitor the CPU or memory usage of the device. For example, the monitoring unit can monitor the device's CPU usage in real time. For example, the monitoring unit can issue an alert if the device's CPU usage exceeds a certain threshold. The monitoring unit can also monitor memory usage and issue a warning if memory usage increases. For example, the monitoring unit will issue a notification if memory usage exceeds 80%. The monitoring unit can also monitor disk I / O usage and issue a notification if the disk read / write speed decreases. For example, the monitoring unit will issue an alert if disk I / O usage exceeds 90%. This allows for an accurate understanding of the device's resource usage. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can monitor resource usage using an AI model that monitors the device's resource usage in real time and detects anomalies.

[0031] The decision unit can determine the proportion of resources to use for training based on the device's resource utilization. For example, the decision unit may decide to use 10% of the resources for training if the device's CPU utilization is low. For example, the decision unit may decide to use 10% of the resources for training if the CPU utilization is 20% or less. The decision unit can also reduce the proportion of resources used when memory utilization is high. For example, the decision unit may decide to reduce the proportion of resources used to 5% if memory utilization is 80% or higher. Furthermore, the decision unit can adjust the proportion of resources used according to disk I / O utilization. For example, the decision unit may decide to reduce the proportion of resources used to 2% if disk I / O utilization is 90% or higher. This allows for efficient use of the device's resources. Some or all of the above processing in the decision unit may be performed using AI, for example, or without AI. For example, the decision unit can determine the proportion of resources using an AI model that takes the device's resource utilization as input and outputs the proportion of resources to use for training.

[0032] The assignment unit may include an adjustment unit that dynamically adjusts task assignments according to device performance or network conditions. For example, the assignment unit can dynamically adjust task assignments according to device performance or network conditions. For example, the assignment unit can assign tasks considering the device's CPU clock speed, memory capacity, network bandwidth, etc. For example, the assignment unit can assign more tasks to devices with high CPU clock speeds. Also, the assignment unit can assign memory-intensive tasks to devices with large memory capacities. Furthermore, the assignment unit can assign tasks with high data transfer volumes to devices with wide network bandwidths. This enables optimal task assignment according to the device's status. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can adjust task assignments using an AI model that takes device performance and network conditions as input and outputs the optimal task assignment.

[0033] The aggregation unit can send the trained data to a server, which can then aggregate it and update the model. For example, the aggregation unit can send the trained data to a server, which can then aggregate it and update the model. For example, the aggregation unit can periodically send the trained data from each device to a server, which can then aggregate it and update the model. The aggregation unit can also use encryption protocols and authentication methods to ensure secure transmission and aggregation of data. For example, the aggregation unit can encrypt the data before transmission and decrypt it at the receiving end to ensure data security. Furthermore, the aggregation unit can calculate checksums or hash values ​​to ensure data integrity. For example, the aggregation unit can add a checksum to the data it transmits, and the receiving end can verify the checksum to confirm data integrity. This allows distributed trained data to be combined into a single dataset, improving the accuracy of the model. Some or all of the above-described processes in the aggregation unit may be performed using AI, or not. For example, the aggregation unit can perform data aggregation using an AI model that takes trained data as input and outputs aggregated data.

[0034] The learning system may include a permission section for obtaining user permission. The permission section may, for example, display a pop-up notification to obtain user permission. For instance, it might display a pop-up notification while a user is browsing a website, requesting permission to use device resources. The permission section may also obtain permission through a settings screen. For example, it might allow the user to select an option to allow the use of device resources in the settings screen. Furthermore, the permission section may provide a procedure for obtaining prior consent. For example, it might request consent regarding the use of device resources when the user agrees to the terms of service. This allows the system to use device resources after obtaining user permission. Some or all of the above processes in the permission section may be performed using, for example, AI, or not. For example, the permission section may obtain permission using an AI model that analyzes user behavior patterns and requests permission at the optimal time.

[0035] The learning system may include a security unit that performs secure data transmission and aggregation. The security unit may, for example, use encryption protocols to perform secure data transmission and aggregation. For instance, the security unit may encrypt and transmit data, and decrypt it at the receiving end to ensure data security. The security unit may also use authentication methods to verify the source and destination of the data. For example, the security unit may attach a digital signature when transmitting data, and verify the signature at the receiving end to verify the source. Furthermore, the security unit may calculate checksums or hash values ​​to ensure data integrity. For example, the security unit may attach a checksum to the data to be transmitted, and verify the checksum at the receiving end to ensure data integrity. This allows learning to proceed while ensuring data security. Some or all of the above processes in the security unit may be performed using AI, or not. For example, the security unit may use an AI model that evaluates data security risks and proposes optimal security measures to ensure data security.

[0036] The monitoring unit monitors the device's battery level and can temporarily suspend monitoring if the battery level drops. For example, the monitoring unit can temporarily suspend monitoring when the battery level falls below 20% to reduce the device's battery consumption. The monitoring unit can also completely stop monitoring when the battery level falls below 10% to protect the device's battery. Furthermore, the monitoring unit can resume monitoring and continue learning when the battery level recovers to 50% or higher. This reduces battery consumption and extends the device's operating time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input battery level data into a generating AI and have the generating AI execute a monitoring method to optimize battery consumption.

[0037] The monitoring unit can monitor the device temperature and adjust resource utilization to prevent overheating. For example, the monitoring unit can limit resource utilization to 50% if the device temperature exceeds 70 degrees. It can also limit resource utilization to 30% if the device temperature exceeds 80 degrees. Furthermore, the monitoring unit can temporarily suspend resource use if the device temperature exceeds 90 degrees and wait for the device temperature to decrease. This prevents the device from overheating and maintains stable operation. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input device temperature data into a generating AI and have the generating AI execute a method for adjusting resource utilization to prevent overheating.

[0038] The monitoring unit monitors the network connection status of the device and can temporarily suspend monitoring if the connection is unstable. For example, the monitoring unit can monitor the network connection status of the device and temporarily suspend monitoring if the connection is unstable. For example, the monitoring unit can temporarily suspend monitoring if the network connection is unstable and wait until the connection becomes stable. The monitoring unit can also stop monitoring if the network connection is completely lost and resume it after reconnection. Furthermore, the monitoring unit can continue monitoring and learning when the network connection is stable. This allows the device's performance to be maintained by temporarily suspending monitoring when the network connection is unstable. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input network connection status data into a generating AI and have the generating AI execute the optimal monitoring method when the connection is unstable.

[0039] The monitoring unit monitors the application usage of the device and can temporarily suspend monitoring if a specific application is being used. For example, the monitoring unit can temporarily suspend monitoring if a high-load game application is being used, prioritizing device performance. It can also temporarily suspend monitoring if a video playback application is being used to reduce battery consumption. Furthermore, it can temporarily suspend monitoring if a calling application is being used to maintain communication stability. This optimizes device performance by temporarily suspending monitoring when a specific application is being used. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input application usage data into a generating AI and have the generating AI execute the optimal monitoring method when a specific application is being used.

[0040] The decision unit can analyze the device's past resource usage history and determine the optimal resource ratio. For example, the decision unit can analyze the device's past resource usage history and determine the optimal resource ratio. For example, the decision unit can determine the most efficient resource ratio from past resource usage history. The decision unit can also set the optimal resource ratio for a specific time period based on past resource usage history. Furthermore, the decision unit can analyze past resource usage history and determine the ratio that minimizes resource waste. In this way, the device's performance can be optimized by determining the optimal resource ratio based on past resource usage history. Some or all of the above processing in the decision unit may be performed using AI, for example, or without AI. For example, the decision unit can input past resource usage history data into a generating AI and have the generating AI execute a method for determining the optimal resource ratio.

[0041] The decision unit can dynamically adjust resource allocation based on the device's current task priority. For example, the decision unit can dynamically adjust resource allocation based on the device's current task priority. For instance, if a high-priority task is running, the decision unit can set a lower resource allocation to prioritize task performance. Alternatively, if a low-priority task is running, the decision unit can set a higher resource allocation to prioritize learning progress. Furthermore, if the task priority changes, the decision unit can dynamically adjust resource allocation to maintain optimal performance. This allows for optimization of device performance by dynamically adjusting resource allocation based on the current task priority. Some or all of the above processing in the decision unit may be performed using AI, for example, or without AI. For example, the decision unit can input task priority data into a generating AI and have the generating AI execute a method for dynamically adjusting resource allocation.

[0042] The decision-making unit can adjust the resource allocation considering the device's geographical location. For example, the decision-making unit can set a lower resource allocation when the device is moving to conserve battery power. It can also set a higher resource allocation when the device is stationary to prioritize learning progress. Furthermore, if the device is in a specific location, the decision-making unit can set an optimal resource allocation for that location. This allows for optimization of device performance by adjusting resource allocation based on geographical location. Some or all of the above processing in the decision-making unit may be performed using AI, for example, or without AI. For example, the decision-making unit can input geographical location data into a generating AI and have the generating AI execute a method for adjusting resource allocation.

[0043] The decision-making unit can customize resource allocation based on the user's profile information. For example, if the user frequently uses high-load applications, the decision-making unit may set a lower resource allocation to prioritize device performance. Conversely, if the user frequently uses low-load applications, the decision-making unit may set a higher resource allocation to prioritize learning progress. Furthermore, the decision-making unit can customize the optimal resource allocation based on the user's profile information. This allows for optimization of device performance by customizing resource allocation based on the user's profile information. Some or all of the above processing in the decision-making unit may be performed using AI, for example, or without AI. For example, the decision-making unit can input user profile information data into a generating AI and have the generating AI execute a method for customizing resource allocation.

[0044] The assignment unit can analyze the device's past task processing history and determine the optimal task assignment method. For example, the assignment unit can analyze the device's past task processing history and determine the optimal task assignment method. For example, the assignment unit can determine the most efficient task assignment method from the past task processing history. The assignment unit can also set the optimal task assignment method for a specific time period based on the past task processing history. Furthermore, the assignment unit can analyze the past task processing history and determine an assignment method that minimizes task waste. This allows for the optimization of device performance by determining the optimal task assignment method based on the past task processing history. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can input past task processing history data into a generating AI and have the generating AI execute a method for determining the optimal task assignment method.

[0045] The assignment unit can adjust task assignments based on the device's current network bandwidth. For example, the assignment unit can increase task assignments and prioritize learning progress when network bandwidth is high. Conversely, the assignment unit can decrease task assignments and prioritize device performance when network bandwidth is low. Furthermore, the assignment unit can dynamically adjust task assignments to maintain optimal performance when network bandwidth fluctuates. This allows for optimization of device performance by adjusting task assignments based on the current network bandwidth. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can input network bandwidth data into a generating AI and have the generating AI execute methods for adjusting task assignments.

[0046] The assignment unit can adjust task assignments considering the device's geographical location. For example, the assignment unit can reduce task assignments to conserve battery power when the device is moving. It can also increase task assignments to prioritize learning progress when the device is stationary. Furthermore, the assignment unit can set optimal task assignments for a specific location when the device is in that location. This optimizes device performance by adjusting task assignments based on geographical location. Some or all of the above processing in the assignment unit may be performed using AI, for example, or not. For example, the assignment unit can input geographical location data into a generating AI and have the generating AI execute methods for adjusting task assignments.

[0047] The assignment unit can customize task assignments based on the user's profile information. For example, if the user frequently uses high-load applications, the assignment unit may reduce task assignments to prioritize device performance. Conversely, if the user frequently uses low-load applications, the assignment unit may increase task assignments to prioritize learning progress. Furthermore, the assignment unit can customize optimal task assignments based on the user's profile information. This allows for optimization of device performance by customizing task assignments based on the user's profile information. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can input user profile information data into a generating AI and have the generating AI execute a method for customizing task assignments.

[0048] The aggregation unit can analyze the device's past data transmission history and determine the optimal data aggregation method. For example, the aggregation unit can analyze the device's past data transmission history and determine the optimal data aggregation method. For example, the aggregation unit can determine the most efficient data aggregation method from the past data transmission history. The aggregation unit can also set the optimal data aggregation method for a specific time period based on the past data transmission history. Furthermore, the aggregation unit can analyze the past data transmission history and determine an aggregation method that minimizes data waste. This allows for the optimization of device performance by determining the optimal data aggregation method based on the past data transmission history. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input past data transmission history data into a generating AI and have the generating AI execute a method to determine the optimal data aggregation method.

[0049] The aggregation unit can adjust data aggregation based on the device's current network bandwidth. For example, if the network bandwidth is wide, the aggregation unit can set a higher data aggregation frequency to prioritize the progress of learning. Conversely, if the network bandwidth is narrow, the aggregation unit can set a lower data aggregation frequency to prioritize device performance. Furthermore, if the network bandwidth fluctuates, the aggregation unit can dynamically adjust the data aggregation frequency to maintain optimal performance. This allows for optimization of device performance by adjusting data aggregation based on the current network bandwidth. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input network bandwidth data into a generating AI and have the generating AI execute a method for adjusting data aggregation.

[0050] The aggregation unit can adjust data aggregation considering the device's geographical location. For example, the aggregation unit can set a lower data aggregation frequency when the device is moving to conserve battery power. It can also set a higher data aggregation frequency when the device is stationary to prioritize learning progress. Furthermore, the aggregation unit can set an optimal data aggregation method for a specific location when the device is in that location. This optimizes device performance by adjusting data aggregation based on geographical location. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input geographical location data into a generating AI and have the generating AI execute methods for adjusting data aggregation.

[0051] The aggregation unit can customize the data aggregation method based on the device user's profile information. For example, if the user frequently uses high-load applications, the aggregation unit can set the data aggregation frequency low to prioritize device performance. Conversely, if the user frequently uses low-load applications, the aggregation unit can set the data aggregation frequency high to prioritize learning progress. Furthermore, the aggregation unit can customize the optimal data aggregation method based on the user's profile information. This allows for optimization of device performance by customizing the data aggregation method based on the user's profile information. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input user profile information data into a generating AI and have the generating AI execute a method for customizing the data aggregation method.

[0052] The authorization unit can analyze a user's past authorization history and determine the optimal method for obtaining authorization. For example, the authorization unit can analyze a user's past authorization history and determine the optimal method for obtaining authorization. For example, the authorization unit can determine the most efficient method for obtaining authorization from past authorization history. The authorization unit can also set the optimal method for obtaining authorization for a specific time period based on past authorization history. Furthermore, the authorization unit can analyze past authorization history and determine an authorization method that minimizes wasted authorization. This reduces the burden on the user by determining the optimal method for obtaining authorization based on past authorization history. Some or all of the above processing in the authorization unit may be performed using AI, for example, or without AI. For example, the authorization unit can input past authorization history data into a generating AI and have the generating AI execute a method for determining the optimal method for obtaining authorization.

[0053] The permission unit can adjust the permission acquisition method considering the user's geographical location. For example, the permission unit can provide a simple permission acquisition method when the user is on the move, reducing the user's burden. It can also provide a detailed permission acquisition method when the user is stationary, giving the user options. Furthermore, if the user is in a specific location, the permission unit can set the optimal permission acquisition method for that location. This reduces the user's burden by adjusting the permission acquisition method based on geographical location information. Some or all of the above processing in the permission unit may be performed using AI, for example, or without AI. For example, the permission unit can input geographical location data into a generating AI and have the generating AI execute a method for adjusting the permission acquisition method.

[0054] The security department can analyze the device's past security incident history and determine the optimal security measures. For example, the security department can analyze the device's past security incident history and determine the optimal security measures. For example, the security department can determine the most efficient security measures from the past security incident history. The security department can also set the optimal security measures for a specific time period based on the past security incident history. Furthermore, the security department can analyze the past security incident history and determine measures that minimize security waste. In this way, the security of the device can be strengthened by determining the optimal security measures based on the past security incident history. Some or all of the above processes in the security department may be performed using AI, for example, or not using AI. For example, the security department can input past security incident history data into a generating AI and have the generating AI execute a method for determining the optimal security measures.

[0055] The security unit can adjust security measures considering the device's geographical location. For example, the security unit can provide simple security measures to reduce the user's burden when the device is in motion. It can also provide detailed security measures to the user when the device is stationary, giving the user options. Furthermore, the security unit can set optimal security measures for a specific location when the device is in that location. This enhances the device's security by adjusting security measures based on geographical location information. Some or all of the above processing in the security unit may be performed using AI, for example, or without AI. For example, the security unit can input geographical location data into a generating AI and have the generating AI execute methods for adjusting security measures.

[0056] The system according to the embodiment is not limited to the example described above, and various modifications are possible, for example, as follows.

[0057] The learning system includes a monitoring unit that monitors the device's resource usage, a determination unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a battery management unit that monitors the device's battery level and temporarily suspends resource usage when the battery level drops. For example, the battery management unit can temporarily suspend resource usage when the battery level falls below 20% to reduce battery consumption. It can also completely stop resource usage when the battery level falls below 10% to protect the device's battery. Furthermore, it can resume resource usage and continue learning when the battery level recovers to 50% or higher. This reduces battery consumption and extends the device's operating time.

[0058] The learning system includes a monitoring unit that monitors the device's resource usage, a determination unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a temperature management unit that monitors the device's temperature and adjusts the resource usage rate to prevent overheating. For example, the temperature management unit may limit the resource usage rate to 50% if the device temperature exceeds 70 degrees. It may also limit the resource usage rate to 30% if the device temperature exceeds 80 degrees. Moreover, it may temporarily suspend resource usage if the device temperature exceeds 90 degrees and wait for the device temperature to decrease. This prevents the device from overheating and maintains stable operation.

[0059] The learning system includes a monitoring unit that monitors the device's resource usage, a determination unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a network management unit that monitors the device's network connection status and temporarily suspends resource usage if the connection is unstable. For example, the network management unit can temporarily suspend resource usage if the network connection is unstable and wait until the connection stabilizes. It can also suspend resource usage if the network connection is completely lost and resume it after reconnection. Moreover, it can continue resource usage and advance learning if the network connection is stable. This allows the device's performance to be maintained by temporarily suspending resource usage when the network connection is unstable.

[0060] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a geographic information unit that adjusts the ratio of resource usage considering the device's geographic location. For example, the geographic information unit can set a lower ratio of resource usage when the device is moving to conserve battery power. Conversely, when the device is stationary, it can set a higher ratio of resource usage to prioritize the progress of learning. Moreover, if the device is in a specific location, it can set an optimal ratio of resource usage appropriate to that location. This allows for the optimization of device performance by adjusting the ratio of resource usage based on geographic location information.

[0061] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a profile management unit that customizes the ratio of resource usage based on the device user's profile information. For example, if the user frequently uses high-load applications, the profile management unit can set the ratio of resource usage low to prioritize device performance. Conversely, if the user frequently uses low-load applications, it can set the ratio of resource usage high to prioritize the progress of learning. In addition, it can customize the optimal ratio of resources based on the user's profile information. This allows for optimization of device performance by customizing the ratio of resource usage based on the user's profile information.

[0062] The following briefly describes the processing flow for example form 1.

[0063] Step 1: The monitoring unit monitors the device's resource usage. Device resource usage includes, for example, CPU usage, memory usage, and disk I / O. The monitoring unit monitors the device's CPU usage in real time and issues an alert if the usage exceeds a certain threshold. It can also monitor memory usage and issue a warning if memory usage increases. Furthermore, it can monitor disk I / O usage and notify if the disk read / write speed decreases. Step 2: The decision unit determines the proportion of resources to be used for training based on the data collected by the monitoring unit. For example, it may decide to use 10% of the resources for training if the device's CPU usage is low. It can also reduce the proportion of resources used if memory usage is high. Furthermore, it can adjust the proportion of resources used according to disk I / O usage. Step 3: The allocation unit assigns distributed parallel learning tasks to devices based on the resource ratio determined by the decision unit. For example, task assignment can be dynamically adjusted according to device performance and network conditions. Tasks can be assigned considering the device's CPU clock speed, memory capacity, network bandwidth, etc. Furthermore, task assignment can be optimized based on information such as the device's battery level and temperature. Step 4: The aggregation unit aggregates the training data for the tasks assigned by the assignment unit. For example, it sends the completed training data to the server, which aggregates it and updates the model. The aggregation unit can use encryption protocols and authentication methods to ensure secure transmission and aggregation of data. Furthermore, it can calculate checksums or hash values ​​of the data to ensure data integrity.

[0064] (Example of form 2) The learning system according to an embodiment of the present invention is a system that utilizes distributed parallel learning and advances learning using the computing resources of people's smartphones. This learning system works by using 1 to 10% of the device's resources to advance learning, such as displaying advertisements while the user is browsing a website. For example, while a user is browsing a website, the system monitors the device's resources and determines the percentage of resources that can be used for learning based on resource usage. If the CPU and memory usage of the device is low, 10% of the resources can be used for learning. Next, the system assigns distributed parallel learning tasks to the devices, and each device advances learning in a distributed manner. For example, while a user is browsing a website, the system uses the device's resources to train a large-scale language model. Furthermore, the data that has been trained is sent to the system and aggregated. This combines the distributed trained data into one, improving the accuracy of the model. For example, each device sends the trained data to a server, which aggregates it and updates the model. This mechanism allows users to advance learning by effectively utilizing their device's resources without any special operations. In addition, by efficiently using the device's resources, the computational cost of learning can be reduced. For example, by using the device's resources to perform learning while the user is browsing a website, the need for dedicated computing resources is eliminated. This allows the learning system to efficiently utilize the user's device resources for distributed parallel learning.

[0065] The learning system according to the embodiment comprises a monitoring unit, a decision unit, an allocation unit, and an aggregation unit. The monitoring unit monitors the resource utilization of the device. The resource utilization of the device includes, but is not limited to, CPU utilization, memory utilization, and disk I / O. For example, the monitoring unit monitors the CPU utilization of the device in real time and issues an alert if the utilization exceeds a certain threshold. The monitoring unit can also monitor memory utilization and issue a warning if memory usage increases. Furthermore, the monitoring unit can monitor disk I / O usage and issue a notification if the disk read / write speed decreases. The decision unit determines the proportion of resources to be used for learning based on the data collected by the monitoring unit. For example, the decision unit decides to use 10% of the resources for learning when the CPU utilization of the device is low. The decision unit can also reduce the proportion of resources used when memory utilization is high. Furthermore, the decision unit can adjust the proportion of resources used according to the disk I / O usage. The allocation unit assigns distributed parallel learning tasks to the device based on the proportion of resources determined by the decision unit. For example, the assignment unit dynamically adjusts task assignments according to the device's performance and network conditions. The assignment unit can assign tasks considering the device's CPU clock speed, memory capacity, network bandwidth, etc. Furthermore, the assignment unit can optimize task assignments based on information such as the device's battery level and temperature. The aggregation unit aggregates the training data for tasks assigned by the assignment unit. For example, the aggregation unit sends the completed training data to a server, which aggregates it and updates the model. The aggregation unit can use encryption protocols and authentication methods to ensure secure transmission and aggregation of data. Furthermore, the aggregation unit can calculate data checksums and hash values ​​to ensure data integrity. As a result, the learning system according to the embodiment can efficiently use the user's device resources to perform distributed parallel learning.

[0066] The monitoring unit can monitor the CPU or memory usage of the device. For example, the monitoring unit can monitor the device's CPU usage in real time. For example, the monitoring unit can issue an alert if the device's CPU usage exceeds a certain threshold. The monitoring unit can also monitor memory usage and issue a warning if memory usage increases. For example, the monitoring unit will issue a notification if memory usage exceeds 80%. The monitoring unit can also monitor disk I / O usage and issue a notification if the disk read / write speed decreases. For example, the monitoring unit will issue an alert if disk I / O usage exceeds 90%. This allows for an accurate understanding of the device's resource usage. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can monitor resource usage using an AI model that monitors the device's resource usage in real time and detects anomalies.

[0067] The decision unit can determine the proportion of resources to use for training based on the device's resource utilization. For example, the decision unit may decide to use 10% of the resources for training if the device's CPU utilization is low. For example, the decision unit may decide to use 10% of the resources for training if the CPU utilization is 20% or less. The decision unit can also reduce the proportion of resources used when memory utilization is high. For example, the decision unit may decide to reduce the proportion of resources used to 5% if memory utilization is 80% or higher. Furthermore, the decision unit can adjust the proportion of resources used according to disk I / O utilization. For example, the decision unit may decide to reduce the proportion of resources used to 2% if disk I / O utilization is 90% or higher. This allows for efficient use of the device's resources. Some or all of the above processing in the decision unit may be performed using AI, for example, or without AI. For example, the decision unit can determine the proportion of resources using an AI model that takes the device's resource utilization as input and outputs the proportion of resources to use for training.

[0068] The assignment unit may include an adjustment unit that dynamically adjusts task assignments according to device performance or network conditions. For example, the assignment unit can dynamically adjust task assignments according to device performance or network conditions. For example, the assignment unit can assign tasks considering the device's CPU clock speed, memory capacity, network bandwidth, etc. For example, the assignment unit can assign more tasks to devices with high CPU clock speeds. Also, the assignment unit can assign memory-intensive tasks to devices with large memory capacities. Furthermore, the assignment unit can assign tasks with high data transfer volumes to devices with wide network bandwidths. This enables optimal task assignment according to the device's status. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can adjust task assignments using an AI model that takes device performance and network conditions as input and outputs the optimal task assignment.

[0069] The aggregation unit can send the trained data to a server, which can then aggregate it and update the model. For example, the aggregation unit can send the trained data to a server, which can then aggregate it and update the model. For example, the aggregation unit can periodically send the trained data from each device to a server, which can then aggregate it and update the model. The aggregation unit can also use encryption protocols and authentication methods to ensure secure transmission and aggregation of data. For example, the aggregation unit can encrypt the data before transmission and decrypt it at the receiving end to ensure data security. Furthermore, the aggregation unit can calculate checksums or hash values ​​to ensure data integrity. For example, the aggregation unit can add a checksum to the data it transmits, and the receiving end can verify the checksum to confirm data integrity. This allows distributed trained data to be combined into a single dataset, improving the accuracy of the model. Some or all of the above-described processes in the aggregation unit may be performed using AI, or not. For example, the aggregation unit can perform data aggregation using an AI model that takes trained data as input and outputs aggregated data.

[0070] The learning system may include a permission section for obtaining user permission. The permission section may, for example, display a pop-up notification to obtain user permission. For instance, it might display a pop-up notification while a user is browsing a website, requesting permission to use device resources. The permission section may also obtain permission through a settings screen. For example, it might allow the user to select an option to allow the use of device resources in the settings screen. Furthermore, the permission section may provide a procedure for obtaining prior consent. For example, it might request consent regarding the use of device resources when the user agrees to the terms of service. This allows the system to use device resources after obtaining user permission. Some or all of the above processes in the permission section may be performed using, for example, AI, or not. For example, the permission section may obtain permission using an AI model that analyzes user behavior patterns and requests permission at the optimal time.

[0071] The learning system may include a security unit that performs secure data transmission and aggregation. The security unit may, for example, use encryption protocols to perform secure data transmission and aggregation. For instance, the security unit may encrypt and transmit data, and decrypt it at the receiving end to ensure data security. The security unit may also use authentication methods to verify the source and destination of the data. For example, the security unit may attach a digital signature when transmitting data, and verify the signature at the receiving end to verify the source. Furthermore, the security unit may calculate checksums or hash values ​​to ensure data integrity. For example, the security unit may attach a checksum to the data to be transmitted, and verify the checksum at the receiving end to ensure data integrity. This allows learning to proceed while ensuring data security. Some or all of the above processes in the security unit may be performed using AI, or not. For example, the security unit may use an AI model that evaluates data security risks and proposes optimal security measures to ensure data security.

[0072] The monitoring unit can estimate the user's emotions and adjust the monitoring frequency based on the estimated emotions. For example, if the user is stressed, the monitoring unit can lower the monitoring frequency to prioritize device performance. Conversely, if the user is relaxed, the monitoring unit can raise the monitoring frequency to prioritize learning progress. Furthermore, if the user is in a hurry, the monitoring unit can minimize the monitoring frequency and allocate device resources to other tasks. This optimizes device performance by adjusting the monitoring frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the monitoring unit may be performed using AI or not. For example, the monitoring unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0073] The monitoring unit monitors the device's battery level and can temporarily suspend monitoring if the battery level drops. For example, the monitoring unit can temporarily suspend monitoring when the battery level falls below 20% to reduce the device's battery consumption. The monitoring unit can also completely stop monitoring when the battery level falls below 10% to protect the device's battery. Furthermore, the monitoring unit can resume monitoring and continue learning when the battery level recovers to 50% or higher. This reduces battery consumption and extends the device's operating time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input battery level data into a generating AI and have the generating AI execute a monitoring method to optimize battery consumption.

[0074] The monitoring unit can monitor the device temperature and adjust resource utilization to prevent overheating. For example, the monitoring unit can limit resource utilization to 50% if the device temperature exceeds 70 degrees. It can also limit resource utilization to 30% if the device temperature exceeds 80 degrees. Furthermore, the monitoring unit can temporarily suspend resource use if the device temperature exceeds 90 degrees and wait for the device temperature to decrease. This prevents the device from overheating and maintains stable operation. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input device temperature data into a generating AI and have the generating AI execute a method for adjusting resource utilization to prevent overheating.

[0075] The monitoring unit can estimate the user's emotions and select the types of resources to monitor based on the estimated emotions. For example, if the user is stressed, the monitoring unit may prioritize CPU monitoring and reduce memory monitoring. If the user is relaxed, the monitoring unit may also monitor both CPU and memory to optimize learning progress. Furthermore, if the user is in a hurry, the monitoring unit may prioritize network monitoring and reduce monitoring of other resources. This allows for optimization of device performance by selecting resources to monitor according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the monitoring unit may be performed using AI or not. For example, the monitoring unit may input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0076] The monitoring unit monitors the network connection status of the device and can temporarily suspend monitoring if the connection is unstable. For example, the monitoring unit can monitor the network connection status of the device and temporarily suspend monitoring if the connection is unstable. For example, the monitoring unit can temporarily suspend monitoring if the network connection is unstable and wait until the connection becomes stable. The monitoring unit can also stop monitoring if the network connection is completely lost and resume it after reconnection. Furthermore, the monitoring unit can continue monitoring and learning when the network connection is stable. This allows the device's performance to be maintained by temporarily suspending monitoring when the network connection is unstable. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input network connection status data into a generating AI and have the generating AI execute the optimal monitoring method when the connection is unstable.

[0077] The monitoring unit monitors the application usage of the device and can temporarily suspend monitoring if a specific application is being used. For example, the monitoring unit can temporarily suspend monitoring if a high-load game application is being used, prioritizing device performance. It can also temporarily suspend monitoring if a video playback application is being used to reduce battery consumption. Furthermore, it can temporarily suspend monitoring if a calling application is being used to maintain communication stability. This optimizes device performance by temporarily suspending monitoring when a specific application is being used. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input application usage data into a generating AI and have the generating AI execute the optimal monitoring method when a specific application is being used.

[0078] The decision unit can estimate the user's emotions and adjust resource allocations based on the estimated emotions. For example, if the user is stressed, the decision unit may set a lower resource allocation to prioritize device performance. If the user is relaxed, the decision unit may set a higher resource allocation to prioritize learning progress. Furthermore, if the user is in a hurry, the decision unit may minimize resource allocation and allocate device resources to other tasks. This optimizes device performance by adjusting resource allocations according to the user's emotions. Emotion estimation is achieved using emotion estimation functions, such as emotion engines or generative AI. Generative AI includes, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the decision unit may be performed using AI or not. For example, the decision unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0079] The decision unit can analyze the device's past resource usage history and determine the optimal resource ratio. For example, the decision unit can analyze the device's past resource usage history and determine the optimal resource ratio. For example, the decision unit can determine the most efficient resource ratio from past resource usage history. The decision unit can also set the optimal resource ratio for a specific time period based on past resource usage history. Furthermore, the decision unit can analyze past resource usage history and determine the ratio that minimizes resource waste. In this way, the device's performance can be optimized by determining the optimal resource ratio based on past resource usage history. Some or all of the above processing in the decision unit may be performed using AI, for example, or without AI. For example, the decision unit can input past resource usage history data into a generating AI and have the generating AI execute a method for determining the optimal resource ratio.

[0080] The decision unit can dynamically adjust resource allocation based on the device's current task priority. For example, the decision unit can dynamically adjust resource allocation based on the device's current task priority. For instance, if a high-priority task is running, the decision unit can set a lower resource allocation to prioritize task performance. Alternatively, if a low-priority task is running, the decision unit can set a higher resource allocation to prioritize learning progress. Furthermore, if the task priority changes, the decision unit can dynamically adjust resource allocation to maintain optimal performance. This allows for optimization of device performance by dynamically adjusting resource allocation based on the current task priority. Some or all of the above processing in the decision unit may be performed using AI, for example, or without AI. For example, the decision unit can input task priority data into a generating AI and have the generating AI execute a method for dynamically adjusting resource allocation.

[0081] The decision unit can estimate the user's emotions and determine the resource usage time based on the estimated emotions. For example, if the user is stressed, the decision unit may set a shorter resource usage time to prioritize device performance. If the user is relaxed, the decision unit may also set a longer resource usage time to prioritize learning progress. Furthermore, if the user is in a hurry, the decision unit may minimize resource usage time and allocate device resources to other tasks. This optimizes device performance by determining resource usage time according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the decision unit may be performed using AI or not. For example, the decision unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0082] The decision-making unit can adjust the resource allocation considering the device's geographical location. For example, the decision-making unit can set a lower resource allocation when the device is moving to conserve battery power. It can also set a higher resource allocation when the device is stationary to prioritize learning progress. Furthermore, if the device is in a specific location, the decision-making unit can set an optimal resource allocation for that location. This allows for optimization of device performance by adjusting resource allocation based on geographical location. Some or all of the above processing in the decision-making unit may be performed using AI, for example, or without AI. For example, the decision-making unit can input geographical location data into a generating AI and have the generating AI execute a method for adjusting resource allocation.

[0083] The decision-making unit can customize resource allocation based on the user's profile information. For example, if the user frequently uses high-load applications, the decision-making unit may set a lower resource allocation to prioritize device performance. Conversely, if the user frequently uses low-load applications, the decision-making unit may set a higher resource allocation to prioritize learning progress. Furthermore, the decision-making unit can customize the optimal resource allocation based on the user's profile information. This allows for optimization of device performance by customizing resource allocation based on the user's profile information. Some or all of the above processing in the decision-making unit may be performed using AI, for example, or without AI. For example, the decision-making unit can input user profile information data into a generating AI and have the generating AI execute a method for customizing resource allocation.

[0084] The assignment unit can estimate the user's emotions and adjust the task assignment method based on the estimated user emotions. For example, if the user is stressed, the assignment unit may reduce the task assignment and prioritize device performance. Conversely, if the user is relaxed, the assignment unit may increase the task assignment and prioritize learning progress. Furthermore, if the user is in a hurry, the assignment unit may minimize the task assignment and allocate device resources to other tasks. This optimizes device performance by adjusting the task assignment method according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the assignment unit may be performed using AI or not. For example, the assignment unit may input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0085] The assignment unit can analyze the device's past task processing history and determine the optimal task assignment method. For example, the assignment unit can analyze the device's past task processing history and determine the optimal task assignment method. For example, the assignment unit can determine the most efficient task assignment method from the past task processing history. The assignment unit can also set the optimal task assignment method for a specific time period based on the past task processing history. Furthermore, the assignment unit can analyze the past task processing history and determine an assignment method that minimizes task waste. This allows for the optimization of device performance by determining the optimal task assignment method based on the past task processing history. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can input past task processing history data into a generating AI and have the generating AI execute a method for determining the optimal task assignment method.

[0086] The assignment unit can adjust task assignments based on the device's current network bandwidth. For example, the assignment unit can increase task assignments and prioritize learning progress when network bandwidth is high. Conversely, the assignment unit can decrease task assignments and prioritize device performance when network bandwidth is low. Furthermore, the assignment unit can dynamically adjust task assignments to maintain optimal performance when network bandwidth fluctuates. This allows for optimization of device performance by adjusting task assignments based on the current network bandwidth. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can input network bandwidth data into a generating AI and have the generating AI execute methods for adjusting task assignments.

[0087] The assignment unit can estimate the user's emotions and determine task priorities based on the estimated emotions. For example, if the user is stressed, the assignment unit may set a lower priority on the task, prioritizing device performance. Conversely, if the user is relaxed, the assignment unit may set a higher priority on the task, prioritizing learning progress. Furthermore, if the user is in a hurry, the assignment unit may minimize the task's priority, allocating device resources to other tasks. This optimizes device performance by determining task priorities according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the assignment unit may be performed using AI or not. For example, the assignment unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0088] The assignment unit can adjust task assignments considering the device's geographical location. For example, the assignment unit can reduce task assignments to conserve battery power when the device is moving. It can also increase task assignments to prioritize learning progress when the device is stationary. Furthermore, the assignment unit can set optimal task assignments for a specific location when the device is in that location. This optimizes device performance by adjusting task assignments based on geographical location. Some or all of the above processing in the assignment unit may be performed using AI, for example, or not. For example, the assignment unit can input geographical location data into a generating AI and have the generating AI execute methods for adjusting task assignments.

[0089] The assignment unit can customize task assignments based on the user's profile information. For example, if the user frequently uses high-load applications, the assignment unit may reduce task assignments to prioritize device performance. Conversely, if the user frequently uses low-load applications, the assignment unit may increase task assignments to prioritize learning progress. Furthermore, the assignment unit can customize optimal task assignments based on the user's profile information. This allows for optimization of device performance by customizing task assignments based on the user's profile information. Some or all of the above processing in the assignment unit may be performed using AI, for example, or without AI. For example, the assignment unit can input user profile information data into a generating AI and have the generating AI execute a method for customizing task assignments.

[0090] The aggregation unit can estimate the user's emotions and adjust the data aggregation method based on the estimated user emotions. For example, if the user is stressed, the aggregation unit can set the data aggregation frequency low to prioritize device performance. Conversely, if the user is relaxed, the aggregation unit can set the data aggregation frequency high to prioritize learning progress. Furthermore, if the user is in a hurry, the aggregation unit can minimize the data aggregation frequency and allocate device resources to other tasks. This optimizes device performance by adjusting the data aggregation method according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the aggregation unit may be performed using AI or not using AI. For example, the aggregation unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0091] The aggregation unit can analyze the device's past data transmission history and determine the optimal data aggregation method. For example, the aggregation unit can analyze the device's past data transmission history and determine the optimal data aggregation method. For example, the aggregation unit can determine the most efficient data aggregation method from the past data transmission history. The aggregation unit can also set the optimal data aggregation method for a specific time period based on the past data transmission history. Furthermore, the aggregation unit can analyze the past data transmission history and determine an aggregation method that minimizes data waste. This allows for the optimization of device performance by determining the optimal data aggregation method based on the past data transmission history. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input past data transmission history data into a generating AI and have the generating AI execute a method to determine the optimal data aggregation method.

[0092] The aggregation unit can adjust data aggregation based on the device's current network bandwidth. For example, if the network bandwidth is wide, the aggregation unit can set a higher data aggregation frequency to prioritize the progress of learning. Conversely, if the network bandwidth is narrow, the aggregation unit can set a lower data aggregation frequency to prioritize device performance. Furthermore, if the network bandwidth fluctuates, the aggregation unit can dynamically adjust the data aggregation frequency to maintain optimal performance. This allows for optimization of device performance by adjusting data aggregation based on the current network bandwidth. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input network bandwidth data into a generating AI and have the generating AI execute a method for adjusting data aggregation.

[0093] The aggregation unit can estimate the user's emotions and determine the timing of data transmission based on the estimated emotions. For example, if the user is stressed, the aggregation unit may delay data transmission to prioritize device performance. Conversely, if the user is relaxed, the aggregation unit may speed up data transmission to prioritize learning progress. Furthermore, if the user is in a hurry, the aggregation unit may minimize data transmission and allocate device resources to other tasks. This optimizes device performance by determining data transmission timing according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the aggregation unit may be performed using AI or not. For example, the aggregation unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0094] The aggregation unit can adjust data aggregation considering the device's geographical location. For example, the aggregation unit can set a lower data aggregation frequency when the device is moving to conserve battery power. It can also set a higher data aggregation frequency when the device is stationary to prioritize learning progress. Furthermore, the aggregation unit can set an optimal data aggregation method for a specific location when the device is in that location. This optimizes device performance by adjusting data aggregation based on geographical location. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input geographical location data into a generating AI and have the generating AI execute methods for adjusting data aggregation.

[0095] The aggregation unit can customize the data aggregation method based on the device user's profile information. For example, if the user frequently uses high-load applications, the aggregation unit can set the data aggregation frequency low to prioritize device performance. Conversely, if the user frequently uses low-load applications, the aggregation unit can set the data aggregation frequency high to prioritize learning progress. Furthermore, the aggregation unit can customize the optimal data aggregation method based on the user's profile information. This allows for optimization of device performance by customizing the data aggregation method based on the user's profile information. Some or all of the above processing in the aggregation unit may be performed using AI, for example, or without AI. For example, the aggregation unit can input user profile information data into a generating AI and have the generating AI execute a method for customizing the data aggregation method.

[0096] The permission unit can estimate the user's emotions and adjust the permission acquisition method based on the estimated emotions. For example, if the user is stressed, the permission unit can provide a simple permission acquisition method to reduce the user's burden. If the user is relaxed, the permission unit can also provide a detailed permission acquisition method to give the user choices. Furthermore, if the user is in a hurry, the permission unit can provide a quick permission acquisition method to save the user's time. In this way, the burden on the user can be reduced by adjusting the permission acquisition method according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the permission unit may be performed using AI or not using AI. For example, the permission unit can input the user's facial expression data into the generative AI and have the generative AI perform emotion estimation.

[0097] The authorization unit can analyze a user's past authorization history and determine the optimal method for obtaining authorization. For example, the authorization unit can analyze a user's past authorization history and determine the optimal method for obtaining authorization. For example, the authorization unit can determine the most efficient method for obtaining authorization from past authorization history. The authorization unit can also set the optimal method for obtaining authorization for a specific time period based on past authorization history. Furthermore, the authorization unit can analyze past authorization history and determine an authorization method that minimizes wasted authorization. This reduces the burden on the user by determining the optimal method for obtaining authorization based on past authorization history. Some or all of the above processing in the authorization unit may be performed using AI, for example, or without AI. For example, the authorization unit can input past authorization history data into a generating AI and have the generating AI execute a method for determining the optimal method for obtaining authorization.

[0098] The permission unit can estimate the user's emotions and determine the timing of obtaining permission based on the estimated emotions. For example, if the user is stressed, the permission unit may delay obtaining permission to reduce the user's burden. Conversely, if the user is relaxed, the permission unit may expedite obtaining permission to prioritize the learning process. Furthermore, if the user is in a hurry, the permission unit may minimize the timing of obtaining permission to save the user's time. This reduces the user's burden by determining the timing of permission acquisition according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the permission unit may be performed using AI or not using AI. For example, the permission unit may input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0099] The permission unit can adjust the permission acquisition method considering the user's geographical location. For example, the permission unit can provide a simple permission acquisition method when the user is on the move, reducing the user's burden. It can also provide a detailed permission acquisition method when the user is stationary, giving the user options. Furthermore, if the user is in a specific location, the permission unit can set the optimal permission acquisition method for that location. This reduces the user's burden by adjusting the permission acquisition method based on geographical location information. Some or all of the above processing in the permission unit may be performed using AI, for example, or without AI. For example, the permission unit can input geographical location data into a generating AI and have the generating AI execute a method for adjusting the permission acquisition method.

[0100] The security unit can estimate the user's emotions and adjust security measures based on those emotions. For example, if the user is stressed, the security unit can provide simple security measures to reduce the user's burden. If the user is relaxed, the security unit can also provide detailed security measures, giving the user choices. Furthermore, if the user is in a hurry, the security unit can provide rapid security measures to save the user's time. This reduces the user's burden by adjusting security measures according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the security unit may be performed using AI, for example, or not using AI. For example, the security unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0101] The security department can analyze the device's past security incident history and determine the optimal security measures. For example, the security department can analyze the device's past security incident history and determine the optimal security measures. For example, the security department can determine the most efficient security measures from the past security incident history. The security department can also set the optimal security measures for a specific time period based on the past security incident history. Furthermore, the security department can analyze the past security incident history and determine measures that minimize security waste. In this way, the security of the device can be strengthened by determining the optimal security measures based on the past security incident history. Some or all of the above processes in the security department may be performed using AI, for example, or not using AI. For example, the security department can input past security incident history data into a generating AI and have the generating AI execute a method for determining the optimal security measures.

[0102] The security unit can estimate the user's emotions and determine the priority of security measures based on the estimated emotions. For example, if the user is stressed, the security unit can set a lower priority for security measures to reduce the user's burden. Conversely, if the user is relaxed, the security unit can set a higher priority for security measures to prioritize enhanced security. Furthermore, if the user is in a hurry, the security unit can minimize the priority of security measures to save the user's time. In this way, the user's burden can be reduced by determining the priority of security measures according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the security unit may be performed using AI, for example, or not using AI. For example, the security unit can input user facial expression data into a generative AI and have the generative AI perform emotion estimation.

[0103] The security unit can adjust security measures considering the device's geographical location. For example, the security unit can provide simple security measures to reduce the user's burden when the device is in motion. It can also provide detailed security measures to the user when the device is stationary, giving the user options. Furthermore, the security unit can set optimal security measures for a specific location when the device is in that location. This enhances the device's security by adjusting security measures based on geographical location information. Some or all of the above processing in the security unit may be performed using AI, for example, or without AI. For example, the security unit can input geographical location data into a generating AI and have the generating AI execute methods for adjusting security measures. === Hard Collateral 1-1 === Each of the multiple elements described above, including the monitoring unit, decision unit, allocation unit, aggregation unit, authorization unit, and security unit, is implemented, for example, in at least one of the smart device 14 and the data processing unit 12. For example, the monitoring unit monitors the resource utilization rate of the smart device 14 by the control unit 46A of the smart device 14. The decision unit determines the proportion of resources to be used for learning by the specific processing unit 290 of the data processing unit 12. The allocation unit assigns distributed parallel learning tasks to the device by the control unit 46A of the smart device 14. The aggregation unit aggregates the learning data by the specific processing unit 290 of the data processing unit 12. The authorization unit obtains user authorization by the control unit 46A of the smart device 14. The security unit ensures data security by the specific processing unit 290 of the data processing unit 12. === Hard Collateral 1-2 === Each of the multiple elements described above, including the monitoring unit, decision unit, allocation unit, aggregation unit, authorization unit, and security unit, is implemented, for example, in at least one of the smart glasses 214 and the data processing unit 12. For example, the monitoring unit monitors the resource utilization rate of the device by the control unit 46A of the smart glasses 214. The decision unit determines the proportion of resources to be used for learning by the specific processing unit 290 of the data processing unit 12. The allocation unit assigns distributed parallel learning tasks to the device by the control unit 46A of the smart glasses 214. The aggregation unit aggregates the learning data by the specific processing unit 290 of the data processing unit 12. The authorization unit obtains user authorization by the control unit 46A of the smart glasses 214. The security unit ensures data security by the specific processing unit 290 of the data processing unit 12. === Hard Collateral 1-3 === Each of the multiple elements described above, including the monitoring unit, decision unit, allocation unit, aggregation unit, authorization unit, and security unit, is implemented, for example, in at least one of the headset terminal 314 and the data processing unit 12. For example, the monitoring unit monitors the resource utilization rate of the device by the control unit 46A of the headset terminal 314. The decision unit determines the proportion of resources to be used for learning by the specific processing unit 290 of the data processing unit 12. The allocation unit assigns distributed parallel learning tasks to the device by the control unit 46A of the headset terminal 314. The aggregation unit aggregates the learning data by the specific processing unit 290 of the data processing unit 12. The authorization unit obtains user authorization by the control unit 46A of the headset terminal 314. The security unit ensures data security by the specific processing unit 290 of the data processing unit 12. === Hard Collateral 1-4 === Each of the multiple elements described above, including the monitoring unit, decision unit, allocation unit, aggregation unit, authorization unit, and security unit, is implemented, for example, in at least one of the robot 414 and the data processing unit 12. For example, the monitoring unit monitors the resource utilization of the device by the control unit 46A of the robot 414. The decision unit determines the proportion of resources to be used for learning by the specific processing unit 290 of the data processing unit 12. The allocation unit assigns distributed parallel learning tasks to the device by the control unit 46A of the robot 414. The aggregation unit aggregates the learning data by the specific processing unit 290 of the data processing unit 12. The authorization unit obtains user authorization by the control unit 46A of the robot 414. The security unit ensures data security by the specific processing unit 290 of the data processing unit 12.

[0104] The system according to the embodiment is not limited to the example described above, and various modifications are possible, for example, as follows.

[0105] The learning system comprises a monitoring unit that monitors the resource usage of the user's device, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include an emotion adjustment unit that estimates the user's emotions and adjusts the ratio of resource usage based on the estimated emotions. For example, if the user is stressed, the emotion adjustment unit may set the ratio of resource usage low and prioritize device performance. Conversely, if the user is relaxed, it may set the ratio of resource usage high and prioritize the progress of learning. Furthermore, if the user is in a hurry, it may minimize the ratio of resource usage and allocate device resources to other tasks. In this way, the device performance can be optimized by adjusting the ratio of resource usage according to the user's emotions.

[0106] The learning system includes a monitoring unit that monitors the device's resource usage, a determination unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a battery management unit that monitors the device's battery level and temporarily suspends resource usage when the battery level drops. For example, the battery management unit can temporarily suspend resource usage when the battery level falls below 20% to reduce battery consumption. It can also completely stop resource usage when the battery level falls below 10% to protect the device's battery. Furthermore, it can resume resource usage and continue learning when the battery level recovers to 50% or higher. This reduces battery consumption and extends the device's operating time.

[0107] The learning system includes a monitoring unit that monitors the device's resource usage, a determination unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a temperature management unit that monitors the device's temperature and adjusts the resource usage rate to prevent overheating. For example, the temperature management unit may limit the resource usage rate to 50% if the device temperature exceeds 70 degrees. It may also limit the resource usage rate to 30% if the device temperature exceeds 80 degrees. Moreover, it may temporarily suspend resource usage if the device temperature exceeds 90 degrees and wait for the device temperature to decrease. This prevents the device from overheating and maintains stable operation.

[0108] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, a task assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include an emotion assignment unit that estimates the user's emotions and adjusts the task assignment method based on the estimated emotions. For example, if the user is stressed, the emotion assignment unit may reduce the task assignment and prioritize device performance. Conversely, if the user is relaxed, it may increase the task assignment and prioritize the progress of learning. Furthermore, if the user is in a hurry, it may minimize the task assignment and allocate device resources to other tasks. In this way, device performance can be optimized by adjusting the task assignment method according to the user's emotions.

[0109] The learning system includes a monitoring unit that monitors the device's resource usage, a determination unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a network management unit that monitors the device's network connection status and temporarily suspends resource usage if the connection is unstable. For example, the network management unit can temporarily suspend resource usage if the network connection is unstable and wait until the connection stabilizes. It can also suspend resource usage if the network connection is completely lost and resume it after reconnection. Moreover, it can continue resource usage and advance learning if the network connection is stable. This allows the device's performance to be maintained by temporarily suspending resource usage when the network connection is unstable.

[0110] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include an emotion aggregation unit that estimates the user's emotions and adjusts the data aggregation method based on the estimated emotions. For example, if the user is stressed, the emotion aggregation unit may set the data aggregation frequency low to prioritize device performance. Conversely, if the user is relaxed, it may set the data aggregation frequency high to prioritize the progress of learning. Furthermore, if the user is in a hurry, it may minimize the data aggregation frequency and allocate device resources to other tasks. In this way, device performance can be optimized by adjusting the data aggregation method according to the user's emotions.

[0111] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a geographic information unit that adjusts the ratio of resource usage considering the device's geographic location. For example, the geographic information unit can set a lower ratio of resource usage when the device is moving to conserve battery power. Conversely, when the device is stationary, it can set a higher ratio of resource usage to prioritize the progress of learning. Moreover, if the device is in a specific location, it can set an optimal ratio of resource usage appropriate to that location. This allows for the optimization of device performance by adjusting the ratio of resource usage based on geographic location information.

[0112] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include an emotion permission unit that estimates the user's emotions and adjusts the permission acquisition method based on the estimated emotions. For example, if the user is stressed, the emotion permission unit can provide a simple permission acquisition method to reduce the user's burden. If the user is relaxed, it can also provide a detailed permission acquisition method to give the user choices. Furthermore, if the user is in a hurry, it can provide a quick permission acquisition method to save the user's time. In this way, the user's burden can be reduced by adjusting the permission acquisition method according to the user's emotions.

[0113] The learning system comprises a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include a profile management unit that customizes the ratio of resource usage based on the device user's profile information. For example, if the user frequently uses high-load applications, the profile management unit can set the ratio of resource usage low to prioritize device performance. Conversely, if the user frequently uses low-load applications, it can set the ratio of resource usage high to prioritize the progress of learning. In addition, it can customize the optimal ratio of resources based on the user's profile information. This allows for optimization of device performance by customizing the ratio of resource usage based on the user's profile information.

[0114] The learning system includes a monitoring unit that monitors the device's resource usage, a decision unit that determines the ratio of resource usage, an assignment unit that assigns tasks, and an aggregation unit that aggregates learning data. Furthermore, the learning system may also include an emotion security unit that estimates the user's emotions and adjusts security measures based on the estimated emotions. For example, if the user is stressed, the emotion security unit can provide simple security measures to reduce the user's burden. If the user is relaxed, it can also provide detailed security measures and give the user choices. Moreover, if the user is in a hurry, it can provide rapid security measures to save the user's time. In this way, the user's burden can be reduced by adjusting security measures according to the user's emotions.

[0115] The following briefly describes the processing flow for example form 2.

[0116] Step 1: The monitoring unit monitors the device's resource usage. Device resource usage includes, for example, CPU usage, memory usage, and disk I / O. The monitoring unit monitors the device's CPU usage in real time and issues an alert if the usage exceeds a certain threshold. It can also monitor memory usage and issue a warning if memory usage increases. Furthermore, it can monitor disk I / O usage and notify if the disk read / write speed decreases. Step 2: The decision unit determines the proportion of resources to be used for training based on the data collected by the monitoring unit. For example, it may decide to use 10% of the resources for training if the device's CPU usage is low. It can also reduce the proportion of resources used if memory usage is high. Furthermore, it can adjust the proportion of resources used according to disk I / O usage. Step 3: The allocation unit assigns distributed parallel learning tasks to devices based on the resource ratio determined by the decision unit. For example, task assignment can be dynamically adjusted according to device performance and network conditions. Tasks can be assigned considering the device's CPU clock speed, memory capacity, network bandwidth, etc. Furthermore, task assignment can be optimized based on information such as the device's battery level and temperature. Step 4: The aggregation unit aggregates the training data for the tasks assigned by the assignment unit. For example, it sends the completed training data to the server, which aggregates it and updates the model. The aggregation unit can use encryption protocols and authentication methods to ensure secure transmission and aggregation of data. Furthermore, it can calculate checksums or hash values ​​of the data to ensure data integrity.

[0117] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0118] Data generation model 58 is a form of so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (for example, still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats from audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVMs), k-means clustering, convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), or Naive Bayes, and can perform a variety of operations, but is not limited to these examples. Furthermore, AI may also be an AI agent. Also, when the operations described above are performed by AI, the operations may be performed partially or entirely by AI, but is not limited to these examples. Additionally, operations performed by AI, including generative AI, may be replaced by rule-based operations, and rule-based operations may be replaced by operations performed by AI, including generative AI.

[0119] Furthermore, the processing performed by the data processing system 10 described above is carried out by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but it may also be carried out by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the smart device 14 or an external device, and the smart device 14 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0120] The correspondence between each part and the device or control unit is not limited to the examples described above, and various modifications are possible.

[0121] [Second Embodiment] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0122] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0123] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN and / or LAN.

[0124] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0125] The microphone 238 receives voice signals from the user and accepts instructions from the user. The microphone 238 captures the voice signals from the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0126] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, which captures images of the area around the user (for example, an imaging range defined by a field of view equivalent to the field of vision of a typical healthy person).

[0127] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0128] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing by the processor 28. The storage 32 stores the specific processing program 56.

[0129] The processor 28 reads a specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 acting as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0130] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0131] In the smart glasses 214, specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 acting as a control unit 46A according to the specific processing program 60 executed on the RAM 48. The smart glasses 214 also have a data generation model 58 and an emotion identification model 59, similar to the data generation model and emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0132] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device that has the data generation model 58. Also, the data processing device 12 may be a server device or a terminal device owned by the user (for example, a mobile phone, robot, home appliance, etc.).

[0133] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0134] The data generation model 58 is a so-called generative AI. An example of a data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each part described above is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example. Also, processing performed by an AI including a generative AI may be replaced by rule-based processing, and rule-based processing may be replaced by processing performed by an AI including a generative AI.

[0135] The data processing system 210 according to the second embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 210 is performed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but it may also be performed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the smart glasses 214 or an external device, and the smart glasses 214 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0136] The correspondence between each part and the device or control unit is not limited to the examples described above, and various modifications are possible.

[0137] [Third Embodiment] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0138] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0139] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN and / or LAN.

[0140] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0141] The microphone 238 receives voice signals from the user and accepts instructions from the user. The microphone 238 captures the voice signals from the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0142] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, which captures images of the area around the user (for example, an imaging range defined by a field of view equivalent to the field of vision of a typical healthy person).

[0143] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0144] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0145] The processor 28 reads a specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 acting as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0146] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0147] In the headset terminal 314, specific processing is performed by the processor 46. The storage 50 stores a specific program 60. The processor 46 reads the specific program 60 from the storage 50 and executes the read specific program 60 on the RAM 48. The specific processing is realized by the processor 46 acting as a control unit 46A according to the specific program 60 executed on the RAM 48. The headset terminal 314 also has a data generation model 58 and an emotion identification model 59, similar to the data generation model and emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0148] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device that has the data generation model 58. Also, the data processing device 12 may be a server device or a terminal device owned by the user (for example, a mobile phone, robot, home appliance, etc.).

[0149] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0150] The data generation model 58 is a so-called generative AI. An example of a data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each part described above is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example. Also, processing performed by an AI including a generative AI may be replaced by rule-based processing, and rule-based processing may be replaced by processing performed by an AI including a generative AI.

[0151] The data processing system 310 according to the third embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 310 is performed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the headset terminal 314, but may also be performed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the headset terminal 314. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the headset terminal 314 or an external device, and the headset terminal 314 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0152] The correspondence between each part and the device or control unit is not limited to the examples described above, and various modifications are possible.

[0153] [Fourth Embodiment] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0154] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0155] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN and / or LAN.

[0156] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0157] The microphone 238 receives voice signals from the user and accepts instructions from the user. The microphone 238 captures the voice signals from the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0158] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS image sensor or CCD image sensor, which captures images of the area around the user (for example, an imaging range defined by a field of view equivalent to the field of vision of a typical healthy person).

[0159] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0160] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. The robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0161] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0162] The processor 28 reads a specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 acting as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0163] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0164] In robot 414, specific processing is performed by processor 46. A specific program 60 is stored in storage 50. Processor 46 reads the specific program 60 from storage 50 and executes it on RAM 48. The specific processing is achieved by processor 46 acting as a control unit 46A according to the specific program 60 executed on RAM 48. Robot 414 also has data generation model 58 and emotion identification model 59, similar to those of the robot, and can perform processing similar to that of the specific processing unit 290 using these models.

[0165] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device that has the data generation model 58. Also, the data processing device 12 may be a server device or a terminal device owned by the user (for example, a mobile phone, robot, home appliance, etc.).

[0166] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0167] The data generation model 58 is a so-called generative AI. An example of a data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each part described above is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example. Also, processing performed by an AI including a generative AI may be replaced by rule-based processing, and rule-based processing may be replaced by processing performed by an AI including a generative AI.

[0168] The data processing system 410 according to the fourth embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 410 is performed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but it may also be performed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the robot 414 or an external device, and the robot 414 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0169] The correspondence between each part and the device or control unit is not limited to the examples described above, and various modifications are possible.

[0170] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0171] Figure 9 shows the emotion map 400, in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0172] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0173] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0174] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, and motorcycles, emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated based, for example, on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0175] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0176] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0177] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing method for the specific process may be used, which includes computer 22 and multiple other computers.

[0178] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0179] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0180] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0181] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0182] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0183] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0184] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0185] Furthermore, although the above-described examples were divided into four embodiments, some or all of these embodiments may be combined. Also, the smart device 14, smart glasses 214, headset terminal 314, and robot 414 are just examples, and they may be combined, or other devices may be used. Also, although the above-described examples were divided into two embodiments, Embodiment 1 and Embodiment 2, these may be combined.

[0186] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and other things that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0187] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0188] [Explanation of symbols]

[0189] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots

Claims

1. A monitoring unit that monitors the resource usage rate of the device, A determination unit that determines the proportion of resources to be used for learning based on the data collected by the monitoring unit, An allocation unit that assigns distributed parallel learning tasks to devices based on the ratio of resources determined by the determination unit, An aggregation unit that aggregates the learning data of the tasks assigned by the aforementioned assignment unit, Equipped with A system characterized by the following features.

2. The aforementioned monitoring unit, Monitor the CPU or memory usage of the device. The system according to feature 1.

3. The aforementioned determination unit, Determine the proportion of resources to use for training based on the device's resource usage. The system according to feature 1.

4. The aforementioned allocation unit is, It includes an adjustment unit that dynamically adjusts task assignments according to the device's performance or network conditions. The system according to feature 1.

5. The aforementioned aggregation unit is The trained data is sent to the server, which aggregates it and updates the model. The system according to feature 1.

6. It includes an authorization section for obtaining user permission. The system according to feature 1.

7. It is equipped with a security unit for secure data transmission and aggregation. The system according to feature 1.

8. The aforementioned monitoring unit, It estimates the user's emotions and adjusts the monitoring frequency based on the estimated emotions. The system according to feature 1.

9. The aforementioned monitoring unit, Monitor the device's battery level and pause monitoring if the battery is low. The system according to feature 1.

Images