system

Abstract

The system according to this embodiment aims to prevent the generating AI from learning incorrect information and to prevent information from being altered due to cyberattacks. [Solution] The system according to the embodiment comprises a detection unit, an analysis unit, and a recovery unit. The detection unit detects the timing at which the generating AI learns incorrect information and the state before a cyberattack. The analysis unit analyzes and verifies the data detected by the detection unit. The recovery unit performs recovery based on the results obtained by the analysis 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, the method including 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 as a 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] In the conventional technology, sufficient measures have not been taken to prevent a generative AI from learning incorrect information or from information rewriting due to a cyber attack, and there is room for improvement.

[0005] The system according to the embodiment aims to prevent a generative AI from learning incorrect information or from information rewriting due to a cyber attack.

Means for Solving the Problems

[0006] The system according to this embodiment comprises a detection unit, an analysis unit, and a recovery unit. The detection unit detects the timing at which the generating AI learns incorrect information and the state prior to a cyberattack. The analysis unit analyzes and verifies the data detected by the detection unit. The recovery unit performs recovery based on the results obtained by the analysis unit. [Effects of the Invention]

[0007] The system according to this embodiment can prevent the generating AI from learning incorrect information and prevent information from being altered due to cyberattacks. [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, etc. The communication I / F manages communication between multiple computers. Examples of communication standards applied 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 generation AI error prevention system according to an embodiment of the present invention is a system for preventing the generation AI from learning incorrect information, leading to the derivation of incorrect answers, and preventing the rewriting of information due to cyberattacks targeting the generation AI. This system detects the timing at which the generation AI learns incorrect information and the state before a cyberattack, and uses tools that can visualize and analyze internet searches, trending topics, trends, search data, posted data, etc., to analyze and verify the timing and content of the appearance of incorrect information, deviations from trend predictions, etc. Furthermore, it performs recovery of the timing of learning incorrect information and the state before a server attack. This mechanism can prevent the generation AI from learning incorrectly and the rewriting of information due to cyberattacks, thereby improving the reliability of the generation AI.

[0029] The system for preventing mislearning of a generative AI according to this embodiment comprises a detection unit, an analysis unit, and a recovery unit. The detection unit detects the timing at which the generative AI learns incorrect information and the state prior to a cyberattack. For example, the detection unit closely monitors the content and timing of the data that the generative AI learns and detects abnormal data and timing. For example, when the generative AI learns homonyms or Japanese words whose meanings change over time, it can detect the timing at which it learns the wrong meaning. The analysis unit analyzes and verifies the data detected by the detection unit. For example, the analysis unit analyzes internet search and posting data to identify the timing and content of the appearance of misinformation. The analysis unit can also predict the appearance of misinformation by analyzing deviations from trend predictions. The recovery unit performs recovery based on the results obtained by the analysis unit. For example, if the generative AI learns incorrect information, the recovery unit eliminates the impact of the misinformation by returning it to the state before it learned the information. Furthermore, if information is overwritten due to a cyberattack, the recovery unit ensures the reliability of the information by recovering it to the state before the attack. As a result, the system for preventing mislearning of the generation AI according to the embodiment can prevent mislearning of the generation AI and alteration of information due to cyberattacks, thereby improving the reliability of the generation AI.

[0030] The detection unit detects when the generative AI learns incorrect information and the state prior to a cyberattack. Specifically, it closely monitors the content and timing of the data the generative AI learns, detecting abnormal data and timing. For example, when the generative AI learns homonyms or Japanese words whose meanings change over time, it can detect when it learns incorrect meanings. The detection unit uses natural language processing technology to analyze the content of the data and monitors changes in context and meaning in real time. Furthermore, it performs time-series analysis to monitor the timing of the data and detects abnormal patterns and sudden changes. For example, if the generative AI learns a specific keyword that increases sharply at a particular time, it can detect the possibility that that keyword is misinformation. In addition, to detect precursors to cyberattacks, it monitors network traffic and system logs to detect abnormal access and data tampering. This allows the detection unit to detect anomalies before the generative AI learns incorrect information and respond quickly. Furthermore, to improve the accuracy of anomaly detection, the detection unit can learn from past data using machine learning algorithms and automatically identify abnormal patterns. This allows the detection unit to minimize the risk of mislearning by the generative AI and cyberattacks.

[0031] The analysis unit analyzes and verifies the data detected by the detection unit. Specifically, it analyzes internet search and posting data to identify the timing and content of misinformation. The analysis unit uses natural language processing technology to analyze the content of text data in detail and extract the characteristics of misinformation. For example, it identifies the timing of sudden increases in specific keywords or phrases and verifies whether the content is misinformation. The analysis unit can also predict the appearance of misinformation by analyzing deviations from trend predictions. Specifically, it constructs a trend prediction model based on past data and detects abnormal patterns and deviations by comparing it with current data. Furthermore, the analysis unit can evaluate the quality of data learned by the generation AI and quantitatively assess the impact of misinformation. For example, it uses evaluation metrics to measure the proportion and impact of misinformation in order to evaluate the quality of text generated by the generation AI. This allows the analysis unit to accurately grasp the impact when the generation AI learns incorrect information and take appropriate countermeasures. In addition, the analysis unit can use anomaly detection algorithms to detect unusual patterns and abnormal data and issue warnings early. This allows the analysis department to play a crucial role in minimizing the impact of mislearning in the generated AI and cyberattacks.

[0032] The recovery unit performs recovery based on the results obtained by the analysis unit. Specifically, if the generating AI learns incorrect information, it eliminates the impact of the misinformation by reverting to the state before the learning occurred. The recovery unit meticulously records the learning history of the generating AI and identifies when the misinformation was learned. This allows it to revert to the state before the misinformation was learned. Furthermore, if information is overwritten due to a cyberattack, the recovery unit ensures the reliability of the information by recovering to the state before the attack. Specifically, it takes backups regularly and quickly restores from the backup in the event of an attack. In addition, the recovery unit monitors the learning process of the generating AI and can automatically perform recovery if an anomaly is detected. For example, if the generating AI begins to learn abnormal data, it invalidates that data and replaces it with normal data. The recovery unit can also adjust the learning algorithm of the generating AI and take measures to minimize the impact of misinformation. In this way, the recovery unit can quickly and effectively eliminate the impact of mislearning and cyberattacks on the generating AI, improving the reliability of the generating AI. Furthermore, the recovery unit can improve the reliability and security of the entire system by evaluating the effectiveness of the recovery process and continuously making improvements.

[0033] The monitoring unit can closely monitor the content and timing of the data that the generating AI learns. For example, the monitoring unit can closely monitor the content and timing of the data that the generating AI learns and detect abnormal data or timing. For example, the monitoring unit can detect the timing at which the generating AI learns an incorrect meaning when learning homonyms or Japanese words whose meanings change over time. In this way, by closely monitoring the content and timing of the data that the generating AI learns, it is possible to prevent the learning of misinformation. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can input the content and timing of the data that the generating AI learns into an AI model and have the AI ​​perform the detection of abnormal data or timing.

[0034] The analysis unit can analyze internet search and social media posting data to identify the timing and content of misinformation. For example, the analysis unit can analyze internet search and social media posting data to identify the timing and content of misinformation. The analysis unit can analyze internet search data to identify the timing and content of misinformation. The analysis unit can also analyze social media posting data to identify the timing and content of misinformation. By identifying the timing and content of misinformation, it is possible to prevent the generation AI from learning incorrectly. Some or all of the above-described processes in the analysis unit may be performed using the generation AI, or they may be performed without the generation AI. For example, the analysis unit can input internet search and social media posting data into the generation AI and have the generation AI perform the task of identifying the timing and content of misinformation.

[0035] The recovery unit can revert to the state before the learning of erroneous information. For example, if the generating AI learns erroneous information, the recovery unit can eliminate the impact of the erroneous information by reverting to the state before that learning. The recovery unit can also recover to the state before the cyberattack if information has been overwritten due to a cyberattack. This improves the reliability of the generating AI by reverting to the state before the learning of erroneous information. Some or all of the above processing in the recovery unit may be performed using AI, for example, or without AI. For example, if the generating AI learns erroneous information, the recovery unit can have the AI ​​perform the process of reverting to the state before that learning.

[0036] The detection unit can detect when a generating AI learns an incorrect meaning when learning homophones or Japanese words whose meanings change over time. For example, the detection unit can detect when a generating AI learns an incorrect meaning when learning homophones or Japanese words whose meanings change over time. For example, the detection unit can detect when a generating AI learns an incorrect meaning when learning homophones such as "bridge" and "chopsticks." The detection unit can also detect when a generating AI learns an incorrect meaning when learning Japanese words whose meanings change over time, such as when the meaning of "telephone" changes from landline to mobile phone. By detecting when a generating AI learns an incorrect meaning, mislearning can be prevented. Some or all of the above processing in the detection unit is implemented using sentiment estimation functions, for example, using a sentiment engine or a generating AI. The generating AI is a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. For example, the detection unit can instruct the generating AI to perform actions that cause it to learn incorrect meanings when it is learning homonyms or Japanese words whose meanings change over time.

[0037] The recovery unit can recover information to its pre-attack state if it has been altered due to a cyberattack. For example, if information has been altered due to a cyberattack, the recovery unit will recover it to its pre-attack state. For example, if data has been tampered with or deleted due to a cyberattack, the recovery unit can recover it to its pre-attack state. This ensures the reliability of information by recovering it to its pre-attack state if it has been altered due to a cyberattack. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, if information has been altered due to a cyberattack, the recovery unit can have AI perform the process of recovering it to its pre-attack state.

[0038] The detection unit evaluates the reliability of the data that the generating AI learns during detection and can prioritize the detection of unreliable data. For example, if the source of the data is unknown, the detection unit will determine that it is unreliable and prioritize its detection. For example, if the data has been previously detected as misinformation, the detection unit will determine that it is unreliable and prioritize its detection. The detection unit can also determine that data is unreliable and prioritize its detection if it increases rapidly. By prioritizing the detection of unreliable data, the accuracy of misinformation detection can be improved. Some or all of the above processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can make the AI ​​evaluate the reliability of the data that the generating AI learns and perform the process of prioritizing the detection of unreliable data.

[0039] The detection unit can identify the source of the data that the generating AI learns from during detection and improve the detection accuracy based on the source. For example, the detection unit can increase the detection accuracy if the data source is reliable. For example, the detection unit can set the detection accuracy lower if the data source is unknown. The detection unit can also adjust the detection accuracy if the data source has previously provided misinformation. In this way, the accuracy of misinformation detection can be improved by improving the detection accuracy based on the data source. Some or all of the above processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can make the AI ​​perform the process of identifying the source of the data that the generating AI learns from and improving the detection accuracy based on the source.

[0040] The detection unit can apply different detection algorithms depending on the format of the data being learned by the generating AI during detection. For example, in the case of text data, the detection unit can apply a natural language processing algorithm. For example, in the case of image data, the detection unit can apply an image recognition algorithm. Furthermore, in the case of audio data, the detection unit can apply a speech recognition algorithm. This allows for improved detection accuracy by applying different detection algorithms depending on the data format. Some or all of the above-described processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can cause the AI ​​to perform the process of applying different detection algorithms depending on the format of the data being learned by the generating AI.

[0041] The detection unit can adjust the detection frequency based on the amount of data the generating AI learns during detection. For example, the detection unit can increase the detection frequency when the amount of data is large. For example, the detection unit can decrease the detection frequency when the amount of data is small. The detection unit can also temporarily increase the detection frequency when the amount of data increases rapidly. By adjusting the detection frequency based on the amount of data, efficient detection becomes possible. Some or all of the above processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can cause the AI ​​to perform the process of adjusting the detection frequency based on the amount of data the generating AI learns.

[0042] The analysis unit can evaluate the impact of misinformation during analysis and prioritize the analysis of information with a high impact. For example, if misinformation is widely disseminated, the analysis unit will determine that it has a high impact and prioritize its analysis. For example, if misinformation is related to an important topic, the analysis unit will determine that it has a high impact and prioritize its analysis. Furthermore, if misinformation affects many users, the analysis unit will determine that it has a high impact and prioritize its analysis. In this way, the impact of misinformation can be minimized by prioritizing the analysis of information with a high impact. Some or all of the above processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of evaluating the impact of misinformation and prioritizing the analysis of information with a high impact.

[0043] The analysis unit can improve the accuracy of its analysis by considering the frequency of misinformation occurrences. For example, the analysis unit can increase the accuracy of its analysis if misinformation occurs frequently. For example, the analysis unit can set the accuracy of its analysis lower if misinformation occurs rarely. Furthermore, the analysis unit can dynamically adjust the accuracy of its analysis if the frequency of misinformation occurrences fluctuates. In this way, the accuracy of the analysis can be improved by considering the frequency of misinformation occurrences. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform a process to evaluate the frequency of misinformation occurrences and improve the accuracy of the analysis.

[0044] The analysis unit can apply different analysis methods based on the location where the misinformation appears during analysis. For example, in the case of misinformation on a website, the analysis unit can apply web analysis methods. For example, in the case of misinformation on social media, the analysis unit can apply social media analysis methods. Furthermore, in the case of misinformation on a forum, the analysis unit can apply forum analysis methods. By applying different analysis methods based on the location where the misinformation appears, the accuracy of the analysis can be improved. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of applying different analysis methods based on the location where the misinformation appears.

[0045] The analysis unit can adjust the order of analysis based on the relevance of the misinformation during the analysis process. For example, if the misinformation is strongly related to other information, the analysis unit will prioritize its analysis. If the misinformation is independent, the analysis unit can analyze it in the normal order. Furthermore, if the misinformation is related to multiple pieces of information, the analysis unit can analyze them in order of relevance. This allows for efficient analysis by adjusting the order of analysis based on the relevance of the misinformation. Some or all of the above processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of adjusting the order of analysis based on the relevance of the misinformation.

[0046] The recovery unit can evaluate the scope of impact of misinformation during recovery and prioritize the recovery of information with a wide impact. For example, if misinformation is widely disseminated, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. For example, if misinformation is related to an important topic, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. Furthermore, if misinformation affects many users, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. In this way, the impact of misinformation can be minimized by prioritizing the recovery of information with a wide impact. Some or all of the above processing in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of evaluating the scope of impact of misinformation and prioritizing the recovery of information with a wide impact.

[0047] The recovery unit can improve the accuracy of recovery by considering the timing of the appearance of misinformation during recovery. For example, the recovery unit can increase the accuracy of recovery if misinformation appears frequently. For example, the recovery unit can set the accuracy of recovery to a lower level if misinformation appears rarely. Furthermore, the recovery unit can dynamically adjust the accuracy of recovery if the timing of the appearance of misinformation fluctuates. In this way, the accuracy of recovery can be improved by considering the timing of the appearance of misinformation. Some or all of the above processing in the recovery unit may be performed using AI, for example, or without using AI. For example, the recovery unit can have AI perform a process to evaluate the timing of the appearance of misinformation and improve the accuracy of recovery.

[0048] The recovery unit can apply different recovery methods based on the location where the misinformation appears during recovery. For example, in the case of misinformation on a website, the recovery unit can apply a web analysis method. For example, in the case of misinformation on social media, the recovery unit can apply a social media analysis method. Furthermore, in the case of misinformation on a forum, the recovery unit can apply a forum analysis method. By applying different recovery methods based on the location where the misinformation appears, the accuracy of recovery can be improved. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of applying different recovery methods based on the location where the misinformation appears.

[0049] The recovery unit can adjust the recovery order based on the relationships between misinformation during the recovery process. For example, if misinformation is strongly related to other information, the recovery unit will prioritize its recovery. If misinformation is independent, the recovery unit can recover it in the normal order. Furthermore, if misinformation is related to multiple pieces of information, the recovery unit can recover them in order of increasing relevance. This allows for efficient recovery by adjusting the recovery order based on the relationships between misinformation. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of adjusting the recovery order based on the relationships between misinformation.

[0050] The monitoring unit can monitor fluctuations in the data that the generating AI learns from in real time during monitoring and immediately detect anomalies. For example, the monitoring unit can immediately detect an anomaly if the data fluctuations increase rapidly. For example, the monitoring unit can immediately detect an anomaly if the data fluctuations exceed the normal range. The monitoring unit can also immediately detect an anomaly if the data fluctuations differ from predictions. In this way, anomalies can be detected immediately by monitoring data fluctuations in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can monitor fluctuations in the data that the generating AI learns from in real time and have the AI ​​perform anomaly detection.

[0051] The monitoring unit can identify the source of the data that the generating AI learns from during monitoring and improve the accuracy of monitoring based on the source. For example, the monitoring unit can increase the accuracy of monitoring if the data source is reliable. For example, the monitoring unit can set the accuracy of monitoring lower if the data source is unknown. The monitoring unit can also adjust the accuracy of monitoring if the data source has previously provided misinformation. In this way, the accuracy of monitoring for misinformation can be improved by improving the accuracy of monitoring based on the data source. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can have the AI ​​perform the process of identifying the source of the data that the generating AI learns from and improving the accuracy of monitoring based on the source.

[0052] The monitoring unit can apply different monitoring algorithms depending on the format of the data being learned by the generating AI during monitoring. For example, the monitoring unit can apply a natural language processing algorithm in the case of text data. For example, the monitoring unit can apply an image recognition algorithm in the case of image data. Furthermore, the monitoring unit can apply a speech recognition algorithm in the case of audio data. This improves monitoring accuracy by applying different monitoring algorithms depending on the data format. Some or all of the above-described processes in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can cause the AI ​​to perform the process of applying different monitoring algorithms depending on the format of the data being learned by the generating AI.

[0053] The monitoring unit can adjust the monitoring frequency based on the amount of data the generating AI learns during monitoring. For example, the monitoring unit can increase the monitoring frequency when the amount of data is large. For example, the monitoring unit can decrease the monitoring frequency when the amount of data is small. The monitoring unit can also temporarily increase the monitoring frequency if the amount of data increases rapidly. This allows for efficient monitoring by adjusting the monitoring frequency based on the amount of data. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can have the AI ​​perform the process of adjusting the monitoring frequency based on the amount of data the generating AI learns.

[0054] The analysis unit can evaluate the impact of misinformation during analysis and prioritize the analysis of information with a high impact. For example, if misinformation is widely disseminated, the analysis unit can determine that it has a high impact and prioritize its analysis. For example, if misinformation is related to an important topic, the analysis unit can determine that it has a high impact and prioritize its analysis. Furthermore, if misinformation affects many users, the analysis unit can determine that it has a high impact and prioritize its analysis. In this way, the impact of misinformation can be minimized by prioritizing the analysis of information with a high impact. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of evaluating the impact of misinformation and prioritizing the analysis of information with a high impact.

[0055] The analysis unit can improve the accuracy of its analysis by considering the frequency of misinformation occurrences during the analysis. For example, the analysis unit can increase the accuracy of the analysis if misinformation occurs frequently. For example, the analysis unit can set the accuracy of the analysis lower if misinformation occurs rarely. Furthermore, the analysis unit can dynamically adjust the accuracy of the analysis if the frequency of misinformation occurrences fluctuates. In this way, the accuracy of the analysis can be improved by considering the frequency of misinformation occurrences. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without using AI. For example, the analysis unit can have AI perform a process to evaluate the frequency of misinformation occurrences and improve the accuracy of the analysis.

[0056] The analysis unit can apply different analysis methods based on the location where the misinformation appears during analysis. For example, in the case of misinformation on a website, the analysis unit can apply a web analysis method. For example, in the case of misinformation on social media, the analysis unit can apply a social media analysis method. Furthermore, in the case of misinformation on a forum, the analysis unit can apply a forum analysis method. By applying different analysis methods based on the location where the misinformation appears, the accuracy of the analysis can be improved. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of applying different analysis methods based on the location where the misinformation appears.

[0057] The analysis unit can adjust the order of analysis based on the relationships between misinformation during the analysis process. For example, if misinformation is strongly related to other information, the analysis unit will prioritize its analysis. If misinformation is independent, the analysis unit can analyze it in the normal order. Furthermore, if misinformation is related to multiple pieces of information, the analysis unit can analyze them in order of increasing relevance. This allows for efficient analysis by adjusting the order of analysis based on the relationships between misinformation. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of adjusting the order of analysis based on the relationships between misinformation.

[0058] The recovery unit can evaluate the scope of impact of misinformation during recovery and prioritize the recovery of information with a wide impact. For example, if misinformation is widely disseminated, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. For example, if misinformation is related to an important topic, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. Furthermore, if misinformation affects many users, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. In this way, the impact of misinformation can be minimized by prioritizing the recovery of information with a wide impact. Some or all of the above processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of evaluating the scope of impact of misinformation and prioritizing the recovery of information with a wide impact.

[0059] The recovery unit can improve the accuracy of recovery by considering the timing of the appearance of misinformation during recovery. For example, the recovery unit can increase the accuracy of recovery if misinformation appears frequently. For example, the recovery unit can set the accuracy of recovery to a lower level if misinformation appears rarely. Furthermore, the recovery unit can dynamically adjust the accuracy of recovery if the timing of the appearance of misinformation fluctuates. In this way, the accuracy of recovery can be improved by considering the timing of the appearance of misinformation. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without using AI. For example, the recovery unit can have AI perform a process to evaluate the timing of the appearance of misinformation and improve the accuracy of recovery.

[0060] The recovery unit can apply different recovery methods based on the location where the misinformation appeared during recovery. For example, in the case of misinformation on a website, the recovery unit can apply a web analysis method. For example, in the case of misinformation on social media, the recovery unit can apply a social media analysis method. Furthermore, in the case of misinformation on a forum, the recovery unit can apply a forum analysis method. By applying different recovery methods based on the location where the misinformation appeared, the accuracy of recovery can be improved. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of applying different recovery methods based on the location where the misinformation appeared.

[0061] The recovery unit can adjust the recovery order based on the relationships between misinformation during the recovery process. For example, if misinformation is strongly related to other information, the recovery unit will prioritize its recovery. If misinformation is independent, the recovery unit can recover it in the normal order. Furthermore, if misinformation is related to multiple pieces of information, the recovery unit can recover them in order of relative relevance. This allows for efficient recovery by adjusting the recovery order based on the relationships between misinformation. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of adjusting the recovery order based on the relationships between misinformation.

[0062] The system according to the embodiment is not limited to the example described above, and various modifications are possible, for example, as follows.

[0063] The generative AI's error prevention system can further evaluate data reliability and prioritize the detection of unreliable data. For example, if the source of the data is unknown, it will be judged as unreliable and prioritized for detection. Also, if the data has been previously detected as misinformation, it will be judged as unreliable and prioritized for detection. Furthermore, if the data increases rapidly, it will be judged as unreliable and prioritized for detection. In this way, the accuracy of misinformation detection can be improved by prioritizing the detection of unreliable data.

[0064] The generative AI's mislearning prevention system can further identify the source of data and improve detection accuracy based on that source. For example, if the data source is reliable, detection accuracy can be increased. Conversely, if the data source is unknown, detection accuracy can be set lower. Furthermore, if the data source has previously provided misinformation, detection accuracy can be adjusted. In this way, by improving detection accuracy based on the data source, the accuracy of misinformation detection can be improved.

[0065] The generation AI's error prevention system further evaluates the scope of misinformation's impact and prioritizes the recovery of information with a wide impact. For example, if misinformation is widely disseminated, it is judged to have a wide impact and prioritized for recovery. Similarly, if misinformation relates to an important topic, it is judged to have a wide impact and prioritized for recovery. Furthermore, if misinformation affects many users, it is judged to have a wide impact and prioritized for recovery. In this way, the impact of misinformation can be minimized by prioritizing the recovery of information with a wide impact.

[0066] The error prevention system for generative AI can further apply different detection algorithms depending on the data format. For example, a natural language processing algorithm can be applied to text data. An image recognition algorithm can be applied to image data. Furthermore, a speech recognition algorithm can be applied to audio data. This allows for improved detection accuracy by applying different detection algorithms depending on the data format.

[0067] The error prevention system for generative AI can further improve the accuracy of analysis by considering the frequency of misinformation. For example, if misinformation appears frequently, the accuracy of the analysis can be increased. Conversely, if misinformation appears rarely, the accuracy of the analysis can be set lower. Furthermore, if the frequency of misinformation fluctuates, the accuracy of the analysis can be dynamically adjusted. In this way, the accuracy of the analysis can be improved by considering the frequency of misinformation.

[0068] The following briefly describes the processing flow for example form 1.

[0069] Step 1: The detection unit detects when the generating AI learns incorrect information and the state prior to a cyberattack. For example, it closely monitors the content and timing of the data the generating AI learns and detects abnormal data and timing. Specifically, it can detect when the generating AI learns homonyms or Japanese words whose meanings change over time and when it learns incorrect meanings. Step 2: The analysis unit analyzes and verifies the data detected by the detection unit. For example, it analyzes internet search and posting data to identify the timing and content of misinformation. It can also predict the appearance of misinformation by analyzing deviations from trend predictions. Step 3: The recovery unit performs recovery based on the results obtained by the analysis unit. For example, if the generating AI learns incorrect information, the unit eliminates the impact of the incorrect information by reverting to the state before it learned the information. Also, if information is overwritten due to a cyberattack, the unit ensures the reliability of the information by recovering to the state before the attack.

[0070] (Example of form 2) The generation AI error prevention system according to an embodiment of the present invention is a system for preventing the generation AI from learning incorrect information, leading to the derivation of incorrect answers, and preventing the rewriting of information due to cyberattacks targeting the generation AI. This system detects the timing at which the generation AI learns incorrect information and the state before a cyberattack, and uses tools that can visualize and analyze internet searches, trending topics, trends, search data, posted data, etc., to analyze and verify the timing and content of the appearance of incorrect information, deviations from trend predictions, etc. Furthermore, it performs recovery of the timing of learning incorrect information and the state before a server attack. This mechanism can prevent the generation AI from learning incorrectly and the rewriting of information due to cyberattacks, thereby improving the reliability of the generation AI.

[0071] The system for preventing mislearning of a generative AI according to this embodiment comprises a detection unit, an analysis unit, and a recovery unit. The detection unit detects the timing at which the generative AI learns incorrect information and the state prior to a cyberattack. For example, the detection unit closely monitors the content and timing of the data that the generative AI learns and detects abnormal data and timing. For example, when the generative AI learns homonyms or Japanese words whose meanings change over time, it can detect the timing at which it learns the wrong meaning. The analysis unit analyzes and verifies the data detected by the detection unit. For example, the analysis unit analyzes internet search and posting data to identify the timing and content of the appearance of misinformation. The analysis unit can also predict the appearance of misinformation by analyzing deviations from trend predictions. The recovery unit performs recovery based on the results obtained by the analysis unit. For example, if the generative AI learns incorrect information, the recovery unit eliminates the impact of the misinformation by returning it to the state before it learned the information. Furthermore, if information is overwritten due to a cyberattack, the recovery unit ensures the reliability of the information by recovering it to the state before the attack. As a result, the system for preventing mislearning of the generation AI according to the embodiment can prevent mislearning of the generation AI and alteration of information due to cyberattacks, thereby improving the reliability of the generation AI.

[0072] The detection unit detects when the generative AI learns incorrect information and the state prior to a cyberattack. Specifically, it closely monitors the content and timing of the data the generative AI learns, detecting abnormal data and timing. For example, when the generative AI learns homonyms or Japanese words whose meanings change over time, it can detect when it learns incorrect meanings. The detection unit uses natural language processing technology to analyze the content of the data and monitors changes in context and meaning in real time. Furthermore, it performs time-series analysis to monitor the timing of the data and detects abnormal patterns and sudden changes. For example, if the generative AI learns a specific keyword that increases sharply at a particular time, it can detect the possibility that that keyword is misinformation. In addition, to detect precursors to cyberattacks, it monitors network traffic and system logs to detect abnormal access and data tampering. This allows the detection unit to detect anomalies before the generative AI learns incorrect information and respond quickly. Furthermore, to improve the accuracy of anomaly detection, the detection unit can learn from past data using machine learning algorithms and automatically identify abnormal patterns. This allows the detection unit to minimize the risk of mislearning by the generative AI and cyberattacks.

[0073] The analysis unit analyzes and verifies the data detected by the detection unit. Specifically, it analyzes internet search and posting data to identify the timing and content of misinformation. The analysis unit uses natural language processing technology to analyze the content of text data in detail and extract the characteristics of misinformation. For example, it identifies the timing of sudden increases in specific keywords or phrases and verifies whether the content is misinformation. The analysis unit can also predict the appearance of misinformation by analyzing deviations from trend predictions. Specifically, it constructs a trend prediction model based on past data and detects abnormal patterns and deviations by comparing it with current data. Furthermore, the analysis unit can evaluate the quality of data learned by the generation AI and quantitatively assess the impact of misinformation. For example, it uses evaluation metrics to measure the proportion and impact of misinformation in order to evaluate the quality of text generated by the generation AI. This allows the analysis unit to accurately grasp the impact when the generation AI learns incorrect information and take appropriate countermeasures. In addition, the analysis unit can use anomaly detection algorithms to detect unusual patterns and abnormal data and issue warnings early. This allows the analysis department to play a crucial role in minimizing the impact of mislearning in the generated AI and cyberattacks.

[0074] The recovery unit performs recovery based on the results obtained by the analysis unit. Specifically, if the generating AI learns incorrect information, it eliminates the impact of the misinformation by reverting to the state before the learning occurred. The recovery unit meticulously records the learning history of the generating AI and identifies when the misinformation was learned. This allows it to revert to the state before the misinformation was learned. Furthermore, if information is overwritten due to a cyberattack, the recovery unit ensures the reliability of the information by recovering to the state before the attack. Specifically, it takes backups regularly and quickly restores from the backup in the event of an attack. In addition, the recovery unit monitors the learning process of the generating AI and can automatically perform recovery if an anomaly is detected. For example, if the generating AI begins to learn abnormal data, it invalidates that data and replaces it with normal data. The recovery unit can also adjust the learning algorithm of the generating AI and take measures to minimize the impact of misinformation. In this way, the recovery unit can quickly and effectively eliminate the impact of mislearning and cyberattacks on the generating AI, improving the reliability of the generating AI. Furthermore, the recovery unit can improve the reliability and security of the entire system by evaluating the effectiveness of the recovery process and continuously making improvements.

[0075] The monitoring unit can closely monitor the content and timing of the data that the generating AI learns. For example, the monitoring unit can closely monitor the content and timing of the data that the generating AI learns and detect abnormal data or timing. For example, the monitoring unit can detect the timing at which the generating AI learns an incorrect meaning when learning homonyms or Japanese words whose meanings change over time. In this way, by closely monitoring the content and timing of the data that the generating AI learns, it is possible to prevent the learning of misinformation. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can input the content and timing of the data that the generating AI learns into an AI model and have the AI ​​perform the detection of abnormal data or timing.

[0076] The analysis unit can analyze internet search and social media posting data to identify the timing and content of misinformation. For example, the analysis unit can analyze internet search and social media posting data to identify the timing and content of misinformation. The analysis unit can analyze internet search data to identify the timing and content of misinformation. The analysis unit can also analyze social media posting data to identify the timing and content of misinformation. By identifying the timing and content of misinformation, it is possible to prevent the generation AI from learning incorrectly. Some or all of the above-described processes in the analysis unit may be performed using the generation AI, or they may be performed without the generation AI. For example, the analysis unit can input internet search and social media posting data into the generation AI and have the generation AI perform the task of identifying the timing and content of misinformation.

[0077] The recovery unit can revert to the state before the learning of erroneous information. For example, if the generating AI learns erroneous information, the recovery unit can eliminate the impact of the erroneous information by reverting to the state before that learning. The recovery unit can also recover to the state before the cyberattack if information has been overwritten due to a cyberattack. This improves the reliability of the generating AI by reverting to the state before the learning of erroneous information. Some or all of the above processing in the recovery unit may be performed using AI, for example, or without AI. For example, if the generating AI learns erroneous information, the recovery unit can have the AI ​​perform the process of reverting to the state before that learning.

[0078] The detection unit can detect when a generating AI learns an incorrect meaning when learning homophones or Japanese words whose meanings change over time. For example, the detection unit can detect when a generating AI learns an incorrect meaning when learning homophones or Japanese words whose meanings change over time. For example, the detection unit can detect when a generating AI learns an incorrect meaning when learning homophones such as "bridge" and "chopsticks." The detection unit can also detect when a generating AI learns an incorrect meaning when learning Japanese words whose meanings change over time, such as when the meaning of "telephone" changes from landline to mobile phone. By detecting when a generating AI learns an incorrect meaning, mislearning can be prevented. Some or all of the above processing in the detection unit is implemented using sentiment estimation functions, for example, using a sentiment engine or a generating AI. The generating AI is a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. For example, the detection unit can instruct the generating AI to perform actions that cause it to learn incorrect meanings when it is learning homonyms or Japanese words whose meanings change over time.

[0079] The recovery unit can recover information to its pre-attack state if it has been altered due to a cyberattack. For example, if information has been altered due to a cyberattack, the recovery unit will recover it to its pre-attack state. For example, if data has been tampered with or deleted due to a cyberattack, the recovery unit can recover it to its pre-attack state. This ensures the reliability of information by recovering it to its pre-attack state if it has been altered due to a cyberattack. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, if information has been altered due to a cyberattack, the recovery unit can have AI perform the process of recovering it to its pre-attack state.

[0080] The detection unit can estimate the user's emotions and adjust the timing of misinformation detection based on the estimated user emotions. For example, if the user is stressed, the detection unit can quickly detect misinformation. If the user is relaxed, the detection unit can detect misinformation at a normal time. The detection unit can also prioritize the detection of misinformation if the user is in a hurry. By adjusting the timing of misinformation detection according to the user's emotions, misinformation can be detected at a more appropriate time. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or a generative AI. The generative AI is, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above processing in the detection unit may be performed using AI, or not using AI. For example, the detection unit can input user emotion data into the generative AI and have the generative AI adjust the timing of misinformation detection.

[0081] The detection unit evaluates the reliability of the data that the generating AI learns during detection and can prioritize the detection of unreliable data. For example, if the source of the data is unknown, the detection unit will determine that it is unreliable and prioritize its detection. For example, if the data has been previously detected as misinformation, the detection unit will determine that it is unreliable and prioritize its detection. The detection unit can also determine that data is unreliable and prioritize its detection if it increases rapidly. By prioritizing the detection of unreliable data, the accuracy of misinformation detection can be improved. Some or all of the above processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can make the AI ​​evaluate the reliability of the data that the generating AI learns and perform the process of prioritizing the detection of unreliable data.

[0082] The detection unit can identify the source of the data that the generating AI learns from during detection and improve the detection accuracy based on the source. For example, the detection unit can increase the detection accuracy if the data source is reliable. For example, the detection unit can set the detection accuracy lower if the data source is unknown. The detection unit can also adjust the detection accuracy if the data source has previously provided misinformation. In this way, the accuracy of misinformation detection can be improved by improving the detection accuracy based on the data source. Some or all of the above processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can make the AI ​​perform the process of identifying the source of the data that the generating AI learns from and improving the detection accuracy based on the source.

[0083] The detection unit can estimate the user's emotions and determine the priority of data to detect based on the estimated user emotions. For example, if the user is stressed, the detection unit will prioritize detecting important data. If the user is relaxed, the detection unit can prioritize detecting data with normal priority. Also, if the user is in a hurry, the detection unit can prioritize detecting data with high urgency. In this way, by prioritizing data according to the user's emotions, more appropriate data can be detected preferentially. 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 detection unit may be performed using AI, for example, or without AI. For example, the detection unit can input user emotion data into a generative AI and have the generative AI determine the priority of data to detect.

[0084] The detection unit can apply different detection algorithms depending on the format of the data being learned by the generating AI during detection. For example, in the case of text data, the detection unit can apply a natural language processing algorithm. For example, in the case of image data, the detection unit can apply an image recognition algorithm. Furthermore, in the case of audio data, the detection unit can apply a speech recognition algorithm. This allows for improved detection accuracy by applying different detection algorithms depending on the data format. Some or all of the above-described processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can cause the AI ​​to perform the process of applying different detection algorithms depending on the format of the data being learned by the generating AI.

[0085] The detection unit can adjust the detection frequency based on the amount of data the generating AI learns during detection. For example, the detection unit can increase the detection frequency when the amount of data is large. For example, the detection unit can decrease the detection frequency when the amount of data is small. The detection unit can also temporarily increase the detection frequency when the amount of data increases rapidly. By adjusting the detection frequency based on the amount of data, efficient detection becomes possible. Some or all of the above processing in the detection unit may be performed using AI, for example, or without AI. For example, the detection unit can cause the AI ​​to perform the process of adjusting the detection frequency based on the amount of data the generating AI learns.

[0086] The analysis unit can estimate the user's emotions and adjust the presentation of the analysis based on the estimated emotions. For example, if the user is nervous, the analysis unit can provide a simple and highly visual presentation. If the user is relaxed, the analysis unit can provide a presentation that includes detailed information. Furthermore, if the user is in a hurry, the analysis unit can provide a concise presentation. By adjusting the presentation of the analysis according to the user's emotions, more appropriate analysis results can be provided. 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 analysis unit may be performed using AI, for example, or not using AI. For example, the analysis unit can input user emotion data into the generative AI and have the generative AI adjust the presentation of the analysis.

[0087] The analysis unit can evaluate the impact of misinformation during analysis and prioritize the analysis of information with a high impact. For example, if misinformation is widely disseminated, the analysis unit will determine that it has a high impact and prioritize its analysis. For example, if misinformation is related to an important topic, the analysis unit will determine that it has a high impact and prioritize its analysis. Furthermore, if misinformation affects many users, the analysis unit will determine that it has a high impact and prioritize its analysis. In this way, the impact of misinformation can be minimized by prioritizing the analysis of information with a high impact. Some or all of the above processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of evaluating the impact of misinformation and prioritizing the analysis of information with a high impact.

[0088] The analysis unit can improve the accuracy of its analysis by considering the frequency of misinformation occurrences. For example, the analysis unit can increase the accuracy of its analysis if misinformation occurs frequently. For example, the analysis unit can set the accuracy of its analysis lower if misinformation occurs rarely. Furthermore, the analysis unit can dynamically adjust the accuracy of its analysis if the frequency of misinformation occurrences fluctuates. In this way, the accuracy of the analysis can be improved by considering the frequency of misinformation occurrences. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform a process to evaluate the frequency of misinformation occurrences and improve the accuracy of the analysis.

[0089] The analysis unit can estimate the user's emotions and determine the priority of analysis based on the estimated emotions. For example, if the user is stressed, the analysis unit will prioritize analyzing important information. If the user is relaxed, the analysis unit can analyze information with normal priority. Also, if the user is in a hurry, the analysis unit can prioritize analyzing information with high urgency. In this way, by determining the priority of analysis according to the user's emotions, more important information can be analyzed first. 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 analysis unit may be performed using AI, for example, or not using AI. For example, the analysis unit can input user emotion data into a generative AI and have the generative AI determine the priority of analysis.

[0090] The analysis unit can apply different analysis methods based on the location where the misinformation appears during analysis. For example, in the case of misinformation on a website, the analysis unit can apply web analysis methods. For example, in the case of misinformation on social media, the analysis unit can apply social media analysis methods. Furthermore, in the case of misinformation on a forum, the analysis unit can apply forum analysis methods. By applying different analysis methods based on the location where the misinformation appears, the accuracy of the analysis can be improved. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of applying different analysis methods based on the location where the misinformation appears.

[0091] The analysis unit can adjust the order of analysis based on the relevance of the misinformation during the analysis process. For example, if the misinformation is strongly related to other information, the analysis unit will prioritize its analysis. If the misinformation is independent, the analysis unit can analyze it in the normal order. Furthermore, if the misinformation is related to multiple pieces of information, the analysis unit can analyze them in order of relevance. This allows for efficient analysis by adjusting the order of analysis based on the relevance of the misinformation. Some or all of the above processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of adjusting the order of analysis based on the relevance of the misinformation.

[0092] The recovery unit can estimate the user's emotions and adjust the recovery method based on the estimated emotions. For example, if the user is stressed, the recovery unit can provide a rapid recovery method. For example, if the user is relaxed, the recovery unit can provide a normal recovery method. Furthermore, if the user is in a hurry, the recovery unit can provide a highly urgent recovery method. By adjusting the recovery method according to the user's emotions, a more appropriate recovery becomes possible. 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 recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can input user emotion data into the generative AI and have the generative AI adjust the recovery method.

[0093] The recovery unit can evaluate the scope of impact of misinformation during recovery and prioritize the recovery of information with a wide impact. For example, if misinformation is widely disseminated, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. For example, if misinformation is related to an important topic, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. Furthermore, if misinformation affects many users, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. In this way, the impact of misinformation can be minimized by prioritizing the recovery of information with a wide impact. Some or all of the above processing in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of evaluating the scope of impact of misinformation and prioritizing the recovery of information with a wide impact.

[0094] The recovery unit can improve the accuracy of recovery by considering the timing of the appearance of misinformation during recovery. For example, the recovery unit can increase the accuracy of recovery if misinformation appears frequently. For example, the recovery unit can set the accuracy of recovery to a lower level if misinformation appears rarely. Furthermore, the recovery unit can dynamically adjust the accuracy of recovery if the timing of the appearance of misinformation fluctuates. In this way, the accuracy of recovery can be improved by considering the timing of the appearance of misinformation. Some or all of the above processing in the recovery unit may be performed using AI, for example, or without using AI. For example, the recovery unit can have AI perform a process to evaluate the timing of the appearance of misinformation and improve the accuracy of recovery.

[0095] The recovery unit can estimate the user's emotions and determine recovery priorities based on the estimated emotions. For example, if the user is stressed, the recovery unit will prioritize recovering important information. If the user is relaxed, the recovery unit can recover information with normal priorities. Furthermore, if the user is in a hurry, the recovery unit can prioritize recovering information of high urgency. In this way, by determining recovery priorities according to the user's emotions, more important information can be recovered preferentially. 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 recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can input user emotion data into a generative AI and have the generative AI determine the recovery priorities.

[0096] The recovery unit can apply different recovery methods based on the location where the misinformation appears during recovery. For example, in the case of misinformation on a website, the recovery unit can apply a web analysis method. For example, in the case of misinformation on social media, the recovery unit can apply a social media analysis method. Furthermore, in the case of misinformation on a forum, the recovery unit can apply a forum analysis method. By applying different recovery methods based on the location where the misinformation appears, the accuracy of recovery can be improved. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of applying different recovery methods based on the location where the misinformation appears.

[0097] The recovery unit can adjust the recovery order based on the relationships between misinformation during the recovery process. For example, if misinformation is strongly related to other information, the recovery unit will prioritize its recovery. If misinformation is independent, the recovery unit can recover it in the normal order. Furthermore, if misinformation is related to multiple pieces of information, the recovery unit can recover them in order of increasing relevance. This allows for efficient recovery by adjusting the recovery order based on the relationships between misinformation. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of adjusting the recovery order based on the relationships between misinformation.

[0098] 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 increase the monitoring frequency. If the user is relaxed, for example, the monitoring unit can monitor at the normal frequency. The monitoring unit can also temporarily increase the monitoring frequency if the user is in a hurry. This allows for more appropriate monitoring by adjusting the monitoring frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. The 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, for example, or without AI. For example, the monitoring unit can input user emotion data into the generative AI and have the generative AI adjust the monitoring frequency.

[0099] The monitoring unit can monitor fluctuations in the data that the generating AI learns from in real time during monitoring and immediately detect anomalies. For example, the monitoring unit can immediately detect an anomaly if the data fluctuations increase rapidly. For example, the monitoring unit can immediately detect an anomaly if the data fluctuations exceed the normal range. The monitoring unit can also immediately detect an anomaly if the data fluctuations differ from predictions. In this way, anomalies can be detected immediately by monitoring data fluctuations in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can monitor fluctuations in the data that the generating AI learns from in real time and have the AI ​​perform anomaly detection.

[0100] The monitoring unit can identify the source of the data that the generating AI learns from during monitoring and improve the accuracy of monitoring based on the source. For example, the monitoring unit can increase the accuracy of monitoring if the data source is reliable. For example, the monitoring unit can set the accuracy of monitoring lower if the data source is unknown. The monitoring unit can also adjust the accuracy of monitoring if the data source has previously provided misinformation. In this way, the accuracy of monitoring for misinformation can be improved by improving the accuracy of monitoring based on the data source. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can have the AI ​​perform the process of identifying the source of the data that the generating AI learns from and improving the accuracy of monitoring based on the source.

[0101] The monitoring unit can estimate the user's emotions and determine the priority of data to monitor based on the estimated user emotions. For example, if the user is stressed, the monitoring unit will prioritize monitoring important data. If the user is relaxed, the monitoring unit can monitor data with normal priority. Also, if the user is in a hurry, the monitoring unit can prioritize monitoring data with high urgency. In this way, by determining the priority of data according to the user's emotions, more important data can be monitored first. 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 monitoring unit may be performed using AI or not using AI. For example, the monitoring unit can input user emotion data into a generative AI and have the generative AI determine the priority of data to monitor.

[0102] The monitoring unit can apply different monitoring algorithms depending on the format of the data being learned by the generating AI during monitoring. For example, the monitoring unit can apply a natural language processing algorithm in the case of text data. For example, the monitoring unit can apply an image recognition algorithm in the case of image data. Furthermore, the monitoring unit can apply a speech recognition algorithm in the case of audio data. This improves monitoring accuracy by applying different monitoring algorithms depending on the data format. Some or all of the above-described processes in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can cause the AI ​​to perform the process of applying different monitoring algorithms depending on the format of the data being learned by the generating AI.

[0103] The monitoring unit can adjust the monitoring frequency based on the amount of data the generating AI learns during monitoring. For example, the monitoring unit can increase the monitoring frequency when the amount of data is large. For example, the monitoring unit can decrease the monitoring frequency when the amount of data is small. The monitoring unit can also temporarily increase the monitoring frequency if the amount of data increases rapidly. This allows for efficient monitoring by adjusting the monitoring frequency based on the amount of data. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can have the AI ​​perform the process of adjusting the monitoring frequency based on the amount of data the generating AI learns.

[0104] The analysis unit can estimate the user's emotions and adjust the presentation of the analysis based on the estimated emotions. For example, if the user is tense, the analysis unit can provide a simple and highly visual presentation. If the user is relaxed, the analysis unit can provide a presentation that includes detailed information. Furthermore, if the user is in a hurry, the analysis unit can provide a concise presentation. By adjusting the presentation of the analysis according to the user's emotions, more appropriate analysis results can be provided. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or a generative AI. The generative AI is, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input user emotion data into the generative AI and have the generative AI adjust the presentation of the analysis.

[0105] The analysis unit can evaluate the impact of misinformation during analysis and prioritize the analysis of information with a high impact. For example, if misinformation is widely disseminated, the analysis unit can determine that it has a high impact and prioritize its analysis. For example, if misinformation is related to an important topic, the analysis unit can determine that it has a high impact and prioritize its analysis. Furthermore, if misinformation affects many users, the analysis unit can determine that it has a high impact and prioritize its analysis. In this way, the impact of misinformation can be minimized by prioritizing the analysis of information with a high impact. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of evaluating the impact of misinformation and prioritizing the analysis of information with a high impact.

[0106] The analysis unit can improve the accuracy of its analysis by considering the frequency of misinformation occurrences during the analysis. For example, the analysis unit can increase the accuracy of the analysis if misinformation occurs frequently. For example, the analysis unit can set the accuracy of the analysis lower if misinformation occurs rarely. Furthermore, the analysis unit can dynamically adjust the accuracy of the analysis if the frequency of misinformation occurrences fluctuates. In this way, the accuracy of the analysis can be improved by considering the frequency of misinformation occurrences. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without using AI. For example, the analysis unit can have AI perform a process to evaluate the frequency of misinformation occurrences and improve the accuracy of the analysis.

[0107] The analysis unit can estimate the user's emotions and determine the priority of analysis based on the estimated emotions. For example, if the user is stressed, the analysis unit will prioritize analyzing important information. If the user is relaxed, the analysis unit can analyze information with normal priority. Also, if the user is in a hurry, the analysis unit can prioritize analyzing information with high urgency. In this way, by determining the priority of analysis according to the user's emotions, more important information can be analyzed preferentially. 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 analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input user emotion data into a generative AI and have the generative AI determine the priority of analysis.

[0108] The analysis unit can apply different analysis methods based on the location where the misinformation appears during analysis. For example, in the case of misinformation on a website, the analysis unit can apply a web analysis method. For example, in the case of misinformation on social media, the analysis unit can apply a social media analysis method. Furthermore, in the case of misinformation on a forum, the analysis unit can apply a forum analysis method. By applying different analysis methods based on the location where the misinformation appears, the accuracy of the analysis can be improved. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of applying different analysis methods based on the location where the misinformation appears.

[0109] The analysis unit can adjust the order of analysis based on the relationships between misinformation during the analysis process. For example, if misinformation is strongly related to other information, the analysis unit will prioritize its analysis. If misinformation is independent, the analysis unit can analyze it in the normal order. Furthermore, if misinformation is related to multiple pieces of information, the analysis unit can analyze them in order of increasing relevance. This allows for efficient analysis by adjusting the order of analysis based on the relationships between misinformation. Some or all of the above-described processes in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can have AI perform the process of adjusting the order of analysis based on the relationships between misinformation.

[0110] The recovery unit can estimate the user's emotions and adjust the recovery method based on the estimated emotions. For example, if the user is stressed, the recovery unit can provide a rapid recovery method. For example, if the user is relaxed, the recovery unit can provide a normal recovery method. Furthermore, if the user is in a hurry, the recovery unit can provide a highly urgent recovery method. This allows for more appropriate recovery by adjusting the recovery 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 recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can input user emotion data into a generative AI and have the generative AI adjust the recovery method.

[0111] The recovery unit can evaluate the scope of impact of misinformation during recovery and prioritize the recovery of information with a wide impact. For example, if misinformation is widely disseminated, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. For example, if misinformation is related to an important topic, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. Furthermore, if misinformation affects many users, the recovery unit can determine that the scope of impact is wide and prioritize its recovery. In this way, the impact of misinformation can be minimized by prioritizing the recovery of information with a wide impact. Some or all of the above processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of evaluating the scope of impact of misinformation and prioritizing the recovery of information with a wide impact.

[0112] The recovery unit can improve the accuracy of recovery by considering the timing of the appearance of misinformation during recovery. For example, the recovery unit can increase the accuracy of recovery if misinformation appears frequently. For example, the recovery unit can set the accuracy of recovery to a lower level if misinformation appears rarely. Furthermore, the recovery unit can dynamically adjust the accuracy of recovery if the timing of the appearance of misinformation fluctuates. In this way, the accuracy of recovery can be improved by considering the timing of the appearance of misinformation. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without using AI. For example, the recovery unit can have AI perform a process to evaluate the timing of the appearance of misinformation and improve the accuracy of recovery.

[0113] The recovery unit can estimate the user's emotions and determine recovery priorities based on the estimated emotions. For example, if the user is stressed, the recovery unit will prioritize recovering important information. If the user is relaxed, the recovery unit can recover information with normal priorities. If the user is in a hurry, the recovery unit can also prioritize recovering information of high urgency. In this way, by determining recovery priorities according to the user's emotions, more important information can be recovered first. 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 recovery unit may be performed using AI, for example, or not using AI. For example, the recovery unit can input user emotion data into a generative AI and have the generative AI perform the determination of recovery priorities.

[0114] The recovery unit can apply different recovery methods based on the location where the misinformation appeared during recovery. For example, in the case of misinformation on a website, the recovery unit can apply a web analysis method. For example, in the case of misinformation on social media, the recovery unit can apply a social media analysis method. Furthermore, in the case of misinformation on a forum, the recovery unit can apply a forum analysis method. By applying different recovery methods based on the location where the misinformation appeared, the accuracy of recovery can be improved. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of applying different recovery methods based on the location where the misinformation appeared.

[0115] The recovery unit can adjust the recovery order based on the relationships between misinformation during the recovery process. For example, if misinformation is strongly related to other information, the recovery unit will prioritize its recovery. If misinformation is independent, the recovery unit can recover it in the normal order. Furthermore, if misinformation is related to multiple pieces of information, the recovery unit can recover them in order of relative relevance. This allows for efficient recovery by adjusting the recovery order based on the relationships between misinformation. Some or all of the above-described processes in the recovery unit may be performed using AI, for example, or without AI. For example, the recovery unit can have AI perform the process of adjusting the recovery order based on the relationships between misinformation.

[0116] The system according to the embodiment is not limited to the example described above, and various modifications are possible, for example, as follows.

[0117] The generative AI's mislearning prevention system can further estimate the user's emotions and adjust the timing of misinformation detection based on those emotions. For example, if the user is stressed, misinformation can be detected quickly. If the user is relaxed, misinformation can be detected at the normal timing. Furthermore, if the user is in a hurry, misinformation detection can be prioritized. By adjusting the timing of misinformation detection according to the user's emotions, misinformation can be detected at a more appropriate time.

[0118] The generative AI's error prevention system can further evaluate data reliability and prioritize the detection of unreliable data. For example, if the source of the data is unknown, it will be judged as unreliable and prioritized for detection. Also, if the data has been previously detected as misinformation, it will be judged as unreliable and prioritized for detection. Furthermore, if the data increases rapidly, it will be judged as unreliable and prioritized for detection. In this way, the accuracy of misinformation detection can be improved by prioritizing the detection of unreliable data.

[0119] The generative AI's error-prevention system can further estimate the user's emotions and determine the priority of data to detect based on those emotions. For example, if the user is stressed, important data can be prioritized for detection. If the user is relaxed, data can be detected with normal priority. Furthermore, if the user is in a hurry, highly urgent data can be prioritized for detection. In this way, by prioritizing data according to the user's emotions, more important data can be detected preferentially.

[0120] The generative AI's mislearning prevention system can further identify the source of data and improve detection accuracy based on that source. For example, if the data source is reliable, detection accuracy can be increased. Conversely, if the data source is unknown, detection accuracy can be set lower. Furthermore, if the data source has previously provided misinformation, detection accuracy can be adjusted. In this way, by improving detection accuracy based on the data source, the accuracy of misinformation detection can be improved.

[0121] The generative AI's error-prevention system can further estimate the user's emotions and adjust the recovery method based on those emotions. For example, if the user is stressed, it can provide a rapid recovery method. If the user is relaxed, it can provide a normal recovery method. Furthermore, if the user is in a hurry, it can provide a high-priority recovery method. This allows for more appropriate recovery by adjusting the recovery method according to the user's emotions.

[0122] The generation AI's error prevention system further evaluates the scope of misinformation's impact and prioritizes the recovery of information with a wide impact. For example, if misinformation is widely disseminated, it is judged to have a wide impact and prioritized for recovery. Similarly, if misinformation relates to an important topic, it is judged to have a wide impact and prioritized for recovery. Furthermore, if misinformation affects many users, it is judged to have a wide impact and prioritized for recovery. In this way, the impact of misinformation can be minimized by prioritizing the recovery of information with a wide impact.

[0123] The generation AI's error-prevention system can further estimate the user's emotions and adjust the monitoring frequency based on those estimates. For example, if the user is stressed, the monitoring frequency can be increased. Conversely, if the user is relaxed, monitoring can be performed at the normal frequency. Furthermore, if the user is in a hurry, the monitoring frequency can be temporarily increased. This allows for more appropriate monitoring by adjusting the monitoring frequency according to the user's emotions.

[0124] The error prevention system for generative AI can further apply different detection algorithms depending on the data format. For example, a natural language processing algorithm can be applied to text data. An image recognition algorithm can be applied to image data. Furthermore, a speech recognition algorithm can be applied to audio data. This allows for improved detection accuracy by applying different detection algorithms depending on the data format.

[0125] The generative AI's error-prevention system can further estimate the user's emotions and adjust the presentation of the analysis based on those emotions. For example, if the user is nervous, it can provide a simple and highly visual presentation. If the user is relaxed, it can provide a presentation that includes detailed information. Furthermore, if the user is in a hurry, it can provide a presentation that gets straight to the point. By adjusting the presentation of the analysis according to the user's emotions, it can provide more appropriate analysis results.

[0126] The error prevention system for generative AI can further improve the accuracy of analysis by considering the frequency of misinformation. For example, if misinformation appears frequently, the accuracy of the analysis can be increased. Conversely, if misinformation appears rarely, the accuracy of the analysis can be set lower. Furthermore, if the frequency of misinformation fluctuates, the accuracy of the analysis can be dynamically adjusted. In this way, the accuracy of the analysis can be improved by considering the frequency of misinformation.

[0127] The following briefly describes the processing flow for example form 2.

[0128] Step 1: The detection unit detects when the generating AI learns incorrect information and the state prior to a cyberattack. For example, it closely monitors the content and timing of the data the generating AI learns and detects abnormal data and timing. Specifically, it can detect when the generating AI learns homonyms or Japanese words whose meanings change over time and when it learns incorrect meanings. Step 2: The analysis unit analyzes and verifies the data detected by the detection unit. For example, it analyzes internet search and posting data to identify the timing and content of misinformation. It can also predict the appearance of misinformation by analyzing deviations from trend predictions. Step 3: The recovery unit performs recovery based on the results obtained by the analysis unit. For example, if the generating AI learns incorrect information, the unit eliminates the impact of the incorrect information by reverting to the state before it learned the information. Also, if information is overwritten due to a cyberattack, the unit ensures the reliability of the information by recovering to the state before the attack.

[0129] 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.

[0130] 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 text generation AI, image generation AI, and multimodal generation AI. 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 with 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 from audio data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. 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 various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each of the above parts 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.Furthermore, processing performed by AI, including generative AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by AI, including generative AI.

[0131] 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.

[0132] Each of the multiple elements described above, including the detection unit, analysis unit, recovery unit, monitoring unit, analysis unit, and recovery unit, is implemented in at least one of the smart device 14 and the data processing unit 12. For example, the detection unit is implemented by the control unit 46A of the smart device 14 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented by the specific processing unit 290 of the data processing unit 12. The monitoring unit is implemented by the control unit 46A of the smart device 14 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented by the specific processing unit 290 of the data processing unit 12. The correspondence between each unit and the device or control unit is not limited to the examples described above and can be changed in various ways.

[0133] [Second Embodiment] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0134] 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.

[0135] 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.

[0136] 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.

[0137] 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.

[0138] 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).

[0139] 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.

[0140] 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.

[0141] 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.

[0142] 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.

[0143] 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.

[0144] 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.).

[0145] 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.

[0146] 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. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. 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.

[0147] 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.

[0148] Each of the multiple elements described above, including the detection unit, analysis unit, recovery unit, monitoring unit, analysis unit, and restoration unit, is implemented, for example, in at least one of the smart glasses 214 and the data processing unit 12. For example, the detection unit is implemented by the control unit 46A of the smart glasses 214 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The monitoring unit is implemented, for example, by the control unit 46A of the smart glasses 214 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The restoration unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The correspondence between each unit and the device or control unit is not limited to the examples described above, and various modifications are possible.

[0149] [Third Embodiment] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0150] 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.

[0151] 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.

[0152] 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.

[0153] 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.

[0154] 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).

[0155] 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.

[0156] 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.

[0157] 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.

[0158] 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.

[0159] 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.

[0160] 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.).

[0161] 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.

[0162] 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. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. 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.

[0163] 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.

[0164] Each of the multiple elements described above, including the detection unit, analysis unit, recovery unit, monitoring unit, analysis unit, and recovery unit, is implemented in at least one of the headset terminal 314 and the data processing unit 12. For example, the detection unit is implemented by the control unit 46A of the headset terminal 314 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented by the specific processing unit 290 of the data processing unit 12. The monitoring unit is implemented by the control unit 46A of the headset terminal 314 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented by the specific processing unit 290 of the data processing unit 12. The correspondence between each unit and the device or control unit is not limited to the examples described above and can be modified in various ways.

[0165] [Fourth Embodiment] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0166] 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.

[0167] 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.

[0168] 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.

[0169] 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.

[0170] 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).

[0171] 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.

[0172] 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.

[0173] 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.

[0174] 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.

[0175] 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.

[0176] 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.

[0177] 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.).

[0178] 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.

[0179] 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. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. 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.

[0180] 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.

[0181] Each of the multiple elements described above, including the detection unit, analysis unit, recovery unit, monitoring unit, analysis unit, and recovery unit, is implemented, for example, in at least one of the robot 414 and the data processing unit 12. For example, the detection unit is implemented by the control unit 46A of the robot 414 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The monitoring unit is implemented, for example, by the control unit 46A of the robot 414 or the specific processing unit 290 of the data processing unit 12. The analysis unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The recovery unit is implemented, for example, by the specific processing unit 290 of the data processing unit 12. The correspondence between each unit and the device or control unit is not limited to the examples described above, and various modifications are possible.

[0182] 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.

[0183] 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.

[0184] 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.

[0185] 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.

[0186] 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.

[0187] 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."

[0188] 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.

[0189] 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.

[0190] 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.

[0191] 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.

[0192] 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.

[0193] 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.

[0194] 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.

[0195] 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.

[0196] 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.

[0197] 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.

[0198] 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.

[0199] 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.

[0200] (Note 1) A detection unit that detects when the generating AI learns incorrect information and the state before a cyberattack, An analysis unit analyzes and verifies the data detected by the detection unit, The system includes a recovery unit that performs recovery based on the results obtained by the analysis unit. A system characterized by the following features. (Note 2) It includes a monitoring unit that closely monitors the content and timing of the data that the generating AI learns from. The system described in Appendix 1, characterized by the features described herein. (Note 3) It includes an analysis unit that analyzes internet search and social media posting data to identify the timing and content of misinformation. The system described in Appendix 1, characterized by the features described herein. (Note 4) It includes a recovery unit that restores the state to what it was before the misinformation was learned. The system described in Appendix 1, characterized by the features described herein. (Note 5) The detection unit is When a generative AI learns homophones or Japanese words whose meanings change over time, it detects when it might learn an incorrect meaning. The system described in Appendix 1, characterized by the features described herein. (Note 6) The recovery unit is If information is altered due to a cyberattack, the system will recover to its state before the attack. The system described in Appendix 1, characterized by the features described herein. (Note 7) The detection unit is It estimates the user's emotions and adjusts the timing of misinformation detection based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 8) The detection unit is During detection, the reliability of the data used by the generating AI for learning is evaluated, and data with low reliability is prioritized for detection. The system described in Appendix 1, characterized by the features described herein. (Note 9) The detection unit is During detection, the generator AI identifies the source of the data it learns from and improves the accuracy of detection based on the source. The system described in Appendix 1, characterized by the features described herein. (Note 10) The detection unit is It estimates the user's emotions and determines the priority of data to detect based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 11) The detection unit is During detection, different detection algorithms are applied depending on the format of the data the generating AI is learning from. The system described in Appendix 1, characterized by the features described herein. (Note 12) The detection unit is During detection, the detection frequency is adjusted based on the amount of data the generating AI learns from. The system described in Appendix 1, characterized by the features described herein. (Note 13) The aforementioned analysis unit is It estimates the user's emotions and adjusts the way the analysis is presented based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 14) The aforementioned analysis unit is During analysis, assess the impact of misinformation and prioritize analyzing information with a high impact. The system described in Appendix 1, characterized by the features described herein. (Note 15) The aforementioned analysis unit is When analyzing data, consider the frequency of misinformation to improve the accuracy of the analysis. The system described in Appendix 1, characterized by the features described herein. (Note 16) The aforementioned analysis unit is We estimate the user's emotions and prioritize the analysis based on those estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 17) The aforementioned analysis unit is During analysis, different analytical methods are applied based on where the misinformation appears. The system described in Appendix 1, characterized by the features described herein. (Note 18) The aforementioned analysis unit is During analysis, adjust the order of analysis based on the relevance of the misinformation. The system described in Appendix 1, characterized by the features described herein. (Note 19) The recovery unit is It estimates the user's emotions and adjusts the recovery method based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 20) The recovery unit is During recovery, the scope of the misinformation's impact is evaluated, and information with a wide impact is prioritized for recovery. The system described in Appendix 1, characterized by the features described herein. (Note 21) The recovery unit is During recovery, the timing of misinformation's appearance is taken into consideration to improve the accuracy of the recovery process. The system described in Appendix 1, characterized by the features described herein. (Note 22) The recovery unit is The system estimates the user's emotions and determines recovery priorities based on those estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 23) The recovery unit is During recovery, different recovery methods are applied based on the location where the misinformation appears. The system described in Appendix 1, characterized by the features described herein. (Note 24) The recovery unit is During recovery, the recovery order is adjusted based on the relevance of the misinformation. The system described in Appendix 1, characterized by the features described herein. (Note 25) The aforementioned monitoring unit, It estimates the user's emotions and adjusts the monitoring frequency based on the estimated emotions. The system described in Appendix 2, characterized by the features described herein. (Note 26) The aforementioned monitoring unit, During monitoring, the system monitors changes in the data that the generating AI learns from in real time, and detects anomalies immediately. The system described in Appendix 2, characterized by the features described herein. (Note 27) The aforementioned monitoring unit, During monitoring, the source of the data used by the generating AI for learning is identified, and the accuracy of monitoring is improved based on the source. The system described in Appendix 2, characterized by the features described herein. (Note 28) The aforementioned monitoring unit, It estimates user sentiment and prioritizes data to monitor based on the estimated user sentiment. The system described in Appendix 2, characterized by the features described herein. (Note 29) The aforementioned monitoring unit, During monitoring, different monitoring algorithms are applied depending on the format of the data the generating AI learns from. The system described in Appendix 2, characterized by the features described herein. (Note 30) The aforementioned monitoring unit, During monitoring, the monitoring frequency is adjusted based on the amount of data the generating AI learns from. The system described in Appendix 2, characterized by the features described herein. (Note 31) The aforementioned analysis unit, The system estimates the user's emotions and adjusts the representation of the analysis based on the estimated emotions. The system described in Appendix 3, characterized by the features described herein. (Note 32) The aforementioned analysis unit, During analysis, the impact of misinformation is evaluated, and information with a high impact is prioritized for analysis. The system described in Appendix 3, characterized by the features described herein. (Note 33) The aforementioned analysis unit, During analysis, the frequency of misinformation is taken into consideration to improve the accuracy of the analysis. The system described in Appendix 3, characterized by the features described herein. (Note 34) The aforementioned analysis unit, The system estimates the user's emotions and determines the priority of analysis based on the estimated emotions. The system described in Appendix 3, characterized by the features described herein. (Note 35) The aforementioned analysis unit, During analysis, different analytical methods are applied based on the location where the misinformation appears. The system described in Appendix 3, characterized by the features described herein. (Note 36) The aforementioned analysis unit, During analysis, the order of analysis is adjusted based on the relevance of the misinformation. The system described in Appendix 3, characterized by the features described herein. (Note 37) The aforementioned recovery unit is It estimates the user's emotions and adjusts the recovery method based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 38) The aforementioned recovery unit is During recovery, the scope of the misinformation's impact will be evaluated, and information with a wide-ranging impact will be prioritized for recovery. The system described in Appendix 1, characterized by the features described herein. (Note 39) The aforementioned recovery unit is During recovery, we improve the accuracy of the recovery process by taking into account the timing of the appearance of misinformation. The system described in Appendix 1, characterized by the features described herein. (Note 40) The aforementioned recovery unit is The system estimates the user's emotions and determines recovery priorities based on those estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 41) The aforementioned recovery unit is During recovery, different recovery methods are applied based on the location where the misinformation appeared. The system described in Appendix 1, characterized by the features described herein. (Note 42) The aforementioned recovery unit is During recovery, the recovery order will be adjusted based on the relevance of the misinformation. The system described in Appendix 1, characterized by the features described herein. [Explanation of Symbols]

[0201] 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 detection unit that detects when the generating AI learns incorrect information and the state before a cyberattack, An analysis unit analyzes and verifies the data detected by the detection unit, The system includes a recovery unit that performs recovery based on the results obtained by the analysis unit. A system characterized by the following features.

2. It includes a monitoring unit that closely monitors the content and timing of the data that the generating AI learns from. The system according to feature 1.

3. It includes an analysis unit that analyzes internet search and social media posting data to identify the timing and content of misinformation. The system according to feature 1.

4. It includes a recovery unit that restores the state to what it was before the misinformation was learned. The system according to feature 1.

5. The detection unit is When a generative AI learns homophones or Japanese words whose meanings change over time, it detects when it might learn an incorrect meaning. The system according to feature 1.

6. The recovery unit is If information is altered due to a cyberattack, the system will recover to its state before the attack. The system according to feature 1.

7. The detection unit is It estimates the user's emotions and adjusts the timing of misinformation detection based on the estimated user emotions. The system according to feature 1.

8. The detection unit is During detection, the reliability of the data used by the generating AI for learning is evaluated, and data with low reliability is prioritized for detection. The system according to feature 1.

9. The detection unit is During detection, the source of the data used by the generating AI for learning is identified, and the accuracy of the detection is improved based on the source. The system according to feature 1.

Images