system

Abstract

To provide a system for instructing a user on what action to take on the basis of a type of disaster.SOLUTION: A system includes: means for receiving an emergency bulletin mail including disaster evacuation information such as an earthquake, a tsunami, a fire, and a flood; means for analyzing contents of the received emergency bulletin mail to determine a type of disaster; means for generating a prompt sentence for instructing generation of an instruction of an action to be taken by a user on the basis of the determined type of disaster; means for generating an instruction of an action to be taken by the user using the generated prompt sentence and a generative AI model; means for displaying the generated instruction of the action on a portable terminal of the user; means for identifying a feeling state of the user using a feeling engine; means for generating a prompt sentence for instructing correction of the instruction of the action to be taken by the user on the basis of the feeling state of the identified user; means for correcting the instruction of the action to be taken by the user using the generated prompt sentence and the generative AI model; and means for displaying the generated corrected instruction of the action on the portable terminal of the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Conventional emergency alert emails only inform users of the type of disaster and evacuation information, but do not provide specific instructions on what to do, which can confuse users. Another issue is that they cannot provide personalized responses that take into account the user's location information or past behavioral history. [Means for solving the problem]

[0005] This invention receives emergency alert emails and analyzes their contents. It then uses AI to determine the type of disaster and generate appropriate actions based on that type. The generated actions are then displayed on the user's smartphone. This allows the user to receive specific instructions on how to act, avoiding confusion. It also makes it possible to respond individually, taking into account the user's location information and past behavioral history. [Brief explanation of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 2 is a sequence diagram showing a flow of processing in the data processing system according to the first embodiment of the first form example. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Embodiment 1. [Figure 13] FIG. 10 is a sequence diagram showing a processing flow of a data processing system in a second embodiment of the second form example. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Embodiment Example 2. [Figure 15] FIG. 10 is a sequence diagram showing the flow of processing in a data processing system according to a third embodiment of the third embodiment. [Figure 16] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Embodiment 3. [Figure 17] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when an emotion engine is combined. [Figure 19] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the second embodiment of the second form example when an emotion engine is combined. [Figure 20] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when an emotion engine is combined. [Figure 21] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the third embodiment of the third form example when an emotion engine is combined. [Figure 22] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0007] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, the processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0011] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0012] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 1, a 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.

[0017] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. 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. The database 24 and the communication I / F 26 are also 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).

[0018] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0019] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0020] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0022] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0023] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0024] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0025] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] One embodiment of the present invention is a system using a smartphone. This system has the function of receiving emergency alert emails, such as earthquake, tsunami, and disaster evacuation information. The content of the received emergency alert email is analyzed by an AI inside the system to determine the type of disaster. The AI ​​generates actions to be taken based on the determined type of disaster and displays the action instructions on the smartphone screen.

[0029] "Example 2"

[0030] Furthermore, another embodiment of the present invention is a system that takes into account the user's location information. In this system, AI acquires the user's location information and generates an action to be taken based on that location information. For example, if the user is in an area where a tsunami warning has been issued, the AI ​​generates an action such as "evacuate to the nearest high ground."

[0031] "Example 3"

[0032] Another embodiment of the present invention is a system that learns the user's past behavioral history. In this system, AI learns the user's past behavioral history and generates appropriate actions based on the learning results. For example, it is possible to prioritize routes to evacuation shelters that the user has visited in the past during evacuation drills.

[0033] The processing flow of each embodiment will be described below.

[0034] "Example 1"

[0035] Step 1: Your smartphone will receive emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc.

[0036] Step 2: The contents of the received emergency alert email are analyzed by the AI ​​within the system.

[0037] Step 3: The AI ​​determines the type of disaster and generates appropriate actions based on that type.

[0038] Step 4: Display the generated action instructions on the smartphone screen.

[0039] "Example 2"

[0040] Step 1: AI obtains the user's location information.

[0041] Step 2: Based on the acquired location information, the AI ​​generates the action to be taken.

[0042] Step 3: Display the generated action instructions on the smartphone screen.

[0043] "Example 3"

[0044] Step 1: The AI ​​learns the user's past behavioral history.

[0045] Step 2: Based on the learning results, the AI ​​generates the action to be taken.

[0046] Step 3: Display the generated action instructions on the smartphone screen.

[0047] Example 1

[0048] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0049] Conventional emergency alert systems required users to analyze the content of the received emergency alert email and determine the appropriate course of action, making it difficult to respond quickly and accurately. Furthermore, because they were unable to provide individualized responses that took into account the user's location information or past behavioral history, they could only provide uniform instructions to all users. This made it difficult for users to take appropriate evacuation actions.

[0050] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0051] In this invention, the server includes: means for receiving emergency alert emails for earthquakes, tsunamis, disaster evacuation information, etc.; means for analyzing the content of the received emergency alert emails and determining the type of disaster; artificial intelligence means for generating response actions based on the determined type of disaster; means for displaying the generated actions on the user's mobile information terminal; means for the artificial intelligence means to analyze the content of the emergency alert emails using natural language processing technology; and means for the artificial intelligence means to display the generated action instructions on the screen of the mobile information terminal. This eliminates the need for the user to analyze the content of the received emergency alert emails themselves, allowing them to receive quick and accurate action instructions. Furthermore, it is possible to respond individually to evacuation actions taking into account the user's location information and past behavioral history, allowing for more appropriate evacuation actions.

[0052] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0053] "Mobile information terminal" refers to a portable information processing device such as a smartphone or tablet.

[0054] "Artificial intelligence means" refers to a system that uses technologies such as machine learning and deep learning to analyze data and make judgments and predictions.

[0055] "Natural language processing technology" is a technology that allows computers to understand, interpret, and generate human language.

[0056] "Action instructions" are instructions that indicate specific actions that a user should take in a particular situation.

[0057] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[0058] "Past behavior history" is a record of actions taken by a user in the past, and is used to predict future actions based on this.

[0059] This invention is a system that receives emergency emails about earthquakes, tsunamis, disaster evacuation information, etc., analyzes the contents of the emails, and generates appropriate instructions for actions. A specific embodiment of this system will be described below.

[0060] First, let's assume that a user has a mobile information terminal (smartphone). The terminal has the function to receive emergency alert emails. Emergency alert emails are sent by official organizations such as the government or the Japan Meteorological Agency. The terminal receives these emails through a built-in communication module (e.g., a 4G / 5G modem).

[0061] Next, the server analyzes the content of the received emergency alert email. Specifically, it extracts the text data from the email and analyzes it using natural language processing (NLP) technology. The generative AI model used here is, for example, BERT or GPT-3 (registered trademark). The server obtains the content of the email in text format and inputs it into the generative AI model. For example, it analyzes the text, "A large earthquake has occurred. Please evacuate to a safe place immediately."

[0062] The server uses a generative AI model to determine the type of disaster from the email content. The model analyzes keywords and context within the text to identify disasters such as earthquakes, tsunamis, and fires. For example, if the text "A large earthquake has occurred" is input, the model will determine this as an "earthquake."

[0063] Next, the server generates actions that the user should take based on the type of disaster determined. The generative AI model generates appropriate action instructions according to the type of disaster. For example, in the case of an earthquake, it generates specific action instructions such as "hide under a desk," and in the case of a tsunami, it generates specific action instructions such as "evacuate to higher ground." The server inputs a prompt statement to the generative AI model saying, "Please generate action instructions in the case of an earthquake," and the model generates the action instruction, "Hide under a desk or evacuate outside the building."

[0064] Finally, the terminal displays the generated action instructions on the smartphone screen. The user checks the action instructions displayed on the screen and takes appropriate action. For example, a message such as "Hide under a desk or evacuate outside the building" is displayed on the smartphone screen.

[0065] As a concrete example, consider the case where a user receives an emergency alert email on their smartphone. The email message reads, "A large earthquake has occurred. Please evacuate to a safe location immediately." In this case, the server inputs the following prompt sentence into the generative AI model:

[0066] Prompt text: "An emergency alert email has been received stating, 'A large earthquake has occurred. Please evacuate to a safe place immediately.'" Analyze the contents of this email and generate the action the user should take.

[0067] Based on this prompt, the generative AI model generates the following action instructions:

[0068] Action Instructions: "Hide under a desk or leave the building."

[0069] This instruction for action is displayed on the smartphone screen, prompting the user to take appropriate action.

[0070] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0071] Step 1: Receiving emergency alert emails

[0072] The device receives emergency alert emails. Specifically, it receives emergency alert emails sent by official organizations such as the government and the Japan Meteorological Agency via the device's communications module (4G / 5G modem).

[0073] Input: Emergency alert email sent by official organization

[0074] Output: Text data of the received emergency alert email

[0075] Specific operation: If a user has a smartphone, the device will automatically receive an emergency alert email. For example, if an earthquake alert is sent, the device will receive the email.

[0076] Step 2: Analyzing the email content

[0077] The server analyzes the content of the received emergency alert email. Specifically, it extracts the text data from the email and analyzes it using natural language processing technology. The generative AI models used here are BERT and GPT-3.

[0078] Input: Text data of received emergency alert email

[0079] Output: Parsed email content text data

[0080] Specific operation: The server retrieves the contents of the email in text format and inputs it into the generative AI model. For example, it analyzes the text, "A large earthquake has occurred. Please evacuate to a safe place immediately."

[0081] Step 3: Determine the type of disaster

[0082] The server uses a generative AI model to determine the type of disaster from the email content. The model analyzes keywords and context within the text to identify disasters such as earthquakes, tsunamis, and fires.

[0083] Input: Parsed email content text data

[0084] Output: Type of disaster determined

[0085] Specific operation: The server inputs the text "A large earthquake has occurred" into the generative AI model, and the model analyzes this and determines that it is an "earthquake."

[0086] Step 4: Generate action instructions

[0087] The server generates the actions that the user should take based on the determined type of disaster, and the generative AI model generates appropriate action instructions according to the type of disaster.

[0088] Input: Type of disaster determined

[0089] Output: Generated action instructions

[0090] Specific operation: The server inputs a prompt statement to the generative AI model saying, "Generate action instructions in the event of an earthquake." The model then generates the action instruction, "Hide under a desk or evacuate outside the building."

[0091] Step 5: Displaying Action Instructions

[0092] The terminal displays the generated action instructions on the smartphone screen, and the user checks the action instructions displayed on the screen and takes appropriate action.

[0093] Input: Generated action instructions

[0094] Output: Instructions displayed on the smartphone screen

[0095] Specific operation: The device displays the action instructions received from the server on the screen. For example, a message such as "Hide under a desk or evacuate outside the building" will be displayed on the smartphone screen.

[0096] (Application example 1)

[0097] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0098] Conventional disaster response systems only received emergency alert emails and were unable to provide users with specific response actions. It was also difficult to generate appropriate action instructions in real time depending on the type of disaster, making it difficult for users to act quickly and appropriately. Furthermore, because they were unable to respond individually based on the user's location information or past behavioral history, they could only provide generic instructions.

[0099] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0100] In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating appropriate actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile information terminal, means for determining the type of disaster using a generative AI model and generating appropriate actions, and means for generating action instructions by inputting a prompt sentence into the generative AI model. This allows the user, upon receiving the emergency alert email, to quickly learn specific appropriate actions based on the type of disaster analyzed by the AI, enabling individual responses that take into account location information and past behavioral history.

[0101] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0102] "Analysis" is the process of examining the contents of the received emergency alert email in detail to determine the type of disaster and its impact.

[0103] "Type of disaster" refers to the specific type of disaster that occurred, such as earthquake, tsunami, fire, or flood.

[0104] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[0105] "AI means" refers to means that use artificial intelligence technology to analyze data, determine the type of disaster, and generate appropriate actions to respond.

[0106] The term "mobile information terminal" refers to an information processing device that is carried and used by a user, such as a smartphone or tablet.

[0107] A "generative AI model" is a generative artificial intelligence model, an AI model that has the ability to generate new data and information based on input data.

[0108] A "prompt" is text data input to a generative AI model that contains instructions or questions to obtain a specific output.

[0109] A system for implementing this invention is configured as follows: A server has means for receiving emergency alert emails, such as earthquake, tsunami, and disaster evacuation information. The content of the received emergency alert emails is analyzed in detail by analysis means within the server, and the type of disaster is determined. The analysis means determines the type of disaster using a generative AI model and generates actions to be taken in response.

[0110] The generated action instructions are displayed on the user's mobile information device, which is an information processing device such as a smartphone or tablet that the user always carries with them. The generative AI model operates by inputting prompt sentences, which contain instructions or questions to obtain a specific output.

[0111] As a specific example, suppose the server receives an emergency alert email with the content "Emergency Alert: An earthquake has occurred. The epicenter is in Tokyo." The analysis means analyzes the content of this email and inputs the following prompt sentence into the generative AI model.

[0112] Please analyze the contents of the emergency alert email below and determine the type of disaster.

[0113] Breaking News: An earthquake has occurred. The epicenter is in Tokyo.

[0114] Based on this prompt, the generative AI model determines that it is an "earthquake" and then inputs the following prompt to generate the action to be taken:

[0115] If the disaster type is an earthquake, generate the actions to be taken.

[0116] The generative AI model generates specific instructions, such as "hide under the desk," which are displayed on the user's mobile device. The user can then act quickly by following the instructions displayed on the mobile device.

[0117] The system is implemented using Python and the OpenAI (registered trademark) API. Python is a programming language well-suited for data analysis and operating AI models, and OpenAI's API provides an interface for operating the generative AI model. This allows users to quickly learn specific response actions based on the type of disaster analyzed by the AI ​​when they receive an emergency alert email.

[0118] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0119] Step 1:

[0120] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the content of the emergency alert email, and the output is the received email data. Specifically, the server retrieves the emergency alert email from the mail server and stores its content in memory.

[0121] Step 2:

[0122] The server analyzes the content of the received emergency alert email. The input is the received email data, and the output is the analyzed disaster type. Specifically, the server inputs the following prompt sentence into the generative AI model:

[0123] Please analyze the contents of the emergency alert email below and determine the type of disaster.

[0124] Breaking News: An earthquake has occurred. The epicenter is in Tokyo.

[0125] Based on this prompt, the generative AI model determines the type of disaster to be "earthquake."

[0126] Step 3:

[0127] The server generates the action to be taken based on the type of disaster determined. The input is the type of disaster, and the output is the action to be taken. Specifically, the server inputs the following prompt sentence into the generation AI model:

[0128] If the disaster type is an earthquake, generate the actions to be taken.

[0129] Based on this prompt, the generative AI model generates specific action instructions, such as "hide under the desk."

[0130] Step 4:

[0131] The server sends the generated action instructions to the user's mobile information device. The input is the action instructions to be handled, and the output is the action instructions to be displayed on the user's mobile information device. Specifically, the server formats the action instructions as a text message and sends it to the user's mobile information device.

[0132] Step 5:

[0133] The user acts according to the instructions displayed on the mobile information terminal. The input is the instructions displayed on the mobile information terminal, and the output is the user's specific actions. As a specific action, the user follows instructions such as "hide under a desk" to evacuate to a safe place.

[0134] Example 2

[0135] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0136] Conventional disaster evacuation systems provide uniform evacuation information without considering the user's location information, making it difficult to suggest optimal evacuation actions for each individual user. Furthermore, because they do not consider the user's past behavioral history, they are unable to suggest appropriate actions based on the user's characteristics. This has led to the issue of not being able to adequately ensure the safety of users in emergencies.

[0137] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0138] In this invention, the server includes a means for acquiring user location information, a means for transmitting the acquired location information to the server, and a means for the server to analyze the location information received and generate an action to be taken using a generative AI model. This makes it possible to propose appropriate evacuation actions based on the user's current location in real time. In addition, the generative AI model learns the user's past behavioral history and generates an action to be taken based on the learning results, making it possible to propose more appropriate actions according to the user's characteristics.

[0139] An "emergency message" is a message that includes information about an emergency, such as an earthquake, tsunami, or disaster evacuation information.

[0140] "Location information" means data that indicates a user's current geographic location, including latitude and longitude information.

[0141] The "server" is a computer system that receives and analyzes the user's location information and uses a generative AI model to generate an action to be taken.

[0142] A "generative AI model" is an artificial intelligence model that generates appropriate behavior based on a user's location information and past behavioral history.

[0143] A "prompt" is a textual instruction that is input into a generative AI model to generate appropriate behavior based on the user's location and situation.

[0144] "Terminal" refers to a device used by a user, including a smartphone or tablet.

[0145] "Analysis" is the process of interpreting information and deriving meaning from received data.

[0146] "Action suggestions" are specific instructions for actions that a user should take, generated by a generative AI model based on the user's location information and past behavioral history.

[0147] "Notification" is the process of sending the action suggestions generated by the server to the user's terminal to inform the user.

[0148] This invention is a system that suggests appropriate actions based on the user's location information. This system is realized using the user's device, a server, and a generative AI model.

[0149] First, the user's device is a device with GPS functionality, such as a smartphone or tablet. The device uses the GPS sensor to obtain the user's current location information. The obtained location information is expressed as latitude and longitude data.

[0150] The device then sends the acquired location information to the server using the HTTPS protocol to ensure data security. The device converts the location information into JSON format and sends it to the server via an HTTPS request.

[0151] The server analyzes the received location information. A generative AI model, such as a natural language processing model, is used for the analysis. The server identifies the area the user is in based on the location information and obtains risk information related to that area.

[0152] The server then uses the generative AI model to generate suggested actions based on the user's location. Specifically, the server inputs prompts into the generative AI model to generate appropriate actions. Examples of prompts include:

[0153] "The user is currently near the coast. A tsunami warning has been issued. Please suggest what action the user should take."

[0154] Based on this prompt, the generative AI model generates the suggested action, "Evacuate to the nearest high ground."

[0155] Finally, the server notifies the user's device of the generated action suggestions using a push notification service. The server converts the action suggestions into JSON format and sends them to the user's smartphone via the push notification service. The user's smartphone receives the notification and displays it on the screen.

[0156] In this way, the system can take into account the user's location and suggest appropriate actions in real time, allowing users to take prompt and appropriate action even in emergencies.

[0157] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0158] Step 1:

[0159] The user's device uses the GPS sensor to obtain current location information. The input is data from the GPS sensor, and the output is latitude and longitude information. Specifically, when a user enables location services on their smartphone, the device periodically obtains GPS data.

[0160] Step 2:

[0161] The device sends the acquired location information to the server. The input is latitude and longitude information, and the output is data sent to the server as an HTTPS request. Specifically, the device converts the location information into JSON format and sends it to the server using the HTTPS protocol.

[0162] Step 3:

[0163] The server analyzes the received location information. The input is the location information sent from the device, and the output is the result of identifying the area the user is in. Specifically, the server checks the location information against a database to see if the user is in an area where a tsunami warning has been issued.

[0164] Step 4:

[0165] The server uses a generative AI model to generate action suggestions based on the user's location. The input is the location and a prompt, and the output is an action suggestion. Specifically, the server inputs the following prompt into the generative AI model:

[0166] "The user is currently near the coast. A tsunami warning has been issued. Please suggest what action the user should take."

[0167] Based on this prompt, the generative AI model generates the suggested action, "Evacuate to the nearest high ground."

[0168] Step 5:

[0169] The server notifies the user's device of the generated action suggestions. The input is the action suggestions, and the output is data sent to the user's device as a push notification. Specifically, the server converts the action suggestions into JSON format and sends them to the user's smartphone via a push notification service. The user's smartphone receives the notification and displays it on the screen.

[0170] (Application example 2)

[0171] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0172] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information, making it difficult to provide optimal evacuation actions for each individual user. Furthermore, they lack the functionality to present evacuation routes or notify emergency contacts, resulting in insufficient comprehensive support to ensure the user's safety.

[0173] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating actions to be taken based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for acquiring the user's location information, means for generating an optimal evacuation route based on the acquired location information and displaying it on a map, and means for automatically notifying emergency contacts. This makes it possible to provide optimal evacuation actions based on the user's current location, suggest evacuation routes, and promptly notify emergency contacts.

[0174] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0175] "Type of disaster" refers to the classification of different disasters such as earthquakes, tsunamis, floods, and fires.

[0176] "AI means" is a technology that uses artificial intelligence to analyze data and generate optimal actions based on specific conditions.

[0177] "Mobile device" refers to a portable electronic device such as a smartphone or tablet.

[0178] "Location information" is latitude and longitude data that indicates the user's current location.

[0179] An "evacuation route" refers to the optimal route for a user to evacuate to a safe location.

[0180] "Emergency contact information" is information about people to contact in the event of a disaster or emergency.

[0181] A system for carrying out this invention has the following configuration. First, a server is provided with means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The server also includes means for analyzing the content of the received emergency alert emails and determining the type of disaster. Furthermore, the server is equipped with AI means for generating actions to be taken based on the determined type of disaster.

[0182] The user's mobile device has a means for displaying the generated action. The mobile device also has a means for acquiring the user's location information, and includes a means for generating an optimal evacuation route based on the acquired location information and displaying it on a map. The mobile device also has a means for automatically notifying emergency contacts.

[0183] The following hardware and software are used to implement this system: A smartphone or tablet with GPS functionality is required for the hardware. The software uses a Python program, the Requests library (to obtain disaster information from the API), the Geopy library (to process location information), and the Folium library (to display maps).

[0184] When the server receives an emergency alert email, it analyzes its contents and determines the type of disaster. For example, if a tsunami warning is issued, the server generates an action for the user based on that information, such as "Please evacuate to the nearest high ground." The generated action is displayed on the user's mobile device.

[0185] The mobile device acquires the user's location information in real time and generates the optimal evacuation route based on that information. The evacuation route is displayed on a map, and the user can follow it to evacuate. An automatic notification function also quickly shares the user's current location and evacuation actions with emergency contacts.

[0186] For example, if a user is in Tokyo and a tsunami warning is issued, the server will notify the user, "Please evacuate to the nearest high ground," and the mobile device will display evacuation routes on a map. The following prompt sentence is used:

[0187] Example prompt sentence:

[0188] If the user's current location is Tokyo and a tsunami warning has been issued, generate an evacuation action to the nearest higher ground and display the evacuation route on a map.

[0189] In this way, a system can be realized that provides specific actions to ensure the safety of users.

[0190] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0191] Step 1:

[0192] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the emergency alert email, and the output is the content of the received email. Specifically, the server obtains emails from an external emergency alert system.

[0193] Step 2:

[0194] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the content of the received email, and the output is the type of disaster (e.g., earthquake, tsunami). Specifically, the server analyzes the text of the email, extracts keywords, and identifies the type of disaster.

[0195] Step 3:

[0196] The server generates the appropriate action to be taken based on the type of disaster it has determined. The input is the type of disaster, and the output is the appropriate action to be taken (e.g., "Please evacuate to the nearest high ground"). Specifically, the server uses a generative AI model to generate the optimal action according to the type of disaster.

[0197] Step 4:

[0198] The server sends the generated action to the user's mobile device. The input is the action to be taken, and the output is an action message to be displayed on the user's mobile device. Specifically, the server pushes the action message to the mobile device.

[0199] Step 5:

[0200] The device obtains the user's location information in real time. The input is GPS data, and the output is the user's current location (latitude and longitude). Specifically, the device obtains location information using its built-in GPS function.

[0201] Step 6:

[0202] The device generates the optimal evacuation route based on the acquired location information and displays it on a map. The input is the user's current location and evacuation location information, and the output is the evacuation route displayed on the map. Specifically, the device uses a map display library (Folium) to draw the evacuation route.

[0203] Step 7:

[0204] The device automatically notifies emergency contacts. The input is the user's current location and the generated action, and the output is a notification message sent to the emergency contacts. Specifically, the device sends the notification message to the emergency contacts via SMS or email.

[0205] In this way, a system can be realized that provides specific actions to ensure the safety of users.

[0206] Example 3

[0207] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0208] The main function of conventional disaster evacuation systems is to receive emergency alert emails and generate response actions based on their content analysis, but they lack the ability to respond individually to users' past behavioral history and location information. This means that optimal evacuation routes and instructions for each user are not provided, which can reduce the efficiency and safety of evacuation. Furthermore, there is a lack of a mechanism for using user feedback as learning data for the next system, making it difficult to improve the accuracy of the system.

[0209] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0210] In this invention, the server includes means for collecting past behavioral history data of users and storing it in a database, means for preprocessing the stored behavioral history data and training it using a generative AI model, and means for predicting the user's future behavior based on the training results and generating appropriate behavior. This makes it possible to provide optimal evacuation routes and behavioral instructions for each user, improving the efficiency and safety of evacuation. Furthermore, by using user feedback as training data for the next round, the accuracy of the system can be continuously improved.

[0211] An "emergency alert email" is an email sent to quickly notify users of emergency information such as earthquakes, tsunamis, and disaster evacuation information.

[0212] "Analysis" is the process of analyzing the contents of the received emergency alert email and determining the type of disaster and its impact.

[0213] "Artificial intelligence means" refers to machine learning algorithms and models that generate optimal response actions based on a user's behavioral history and location information.

[0214] A "mobile terminal" is an electronic device that a user carries and uses, such as a smartphone or tablet.

[0215] "Behavioral history data" refers to data about the actions a user has taken and the places they have visited in the past.

[0216] A "database" is a system for storing and managing collected behavioral history data.

[0217] "Preprocessing" refers to processes such as filling in missing values, normalizing data, and removing outliers that are carried out to improve the quality of collected data.

[0218] A "generative AI model" is a model trained using a machine learning framework and used to predict user behavior and generate appropriate countermeasures.

[0219] "Learning results" refer to user behavior patterns and predictive models obtained as a result of the generative AI model learning based on behavioral history data.

[0220] "Feedback" is the user's evaluation or opinion of the system's instructions or generated actions.

[0221] The present invention is a system that learns the user's past behavioral history and generates an action to be taken based on the learning results. Specific embodiments of this system will be described below.

[0222] Server Processing

[0223] The server collects the user's past behavioral history data and stores it in a database. Specifically, it collects GPS data, timestamps, and the type of behavior (e.g., evacuation drills, commuting, etc.) from the user's mobile device. The collected data is stored in a relational database such as MySQL (registered trademark) or PostgreSQL.

[0224] The server then preprocesses the stored data, which includes imputing missing values, normalizing the data, and removing outliers, improving the quality of the data and increasing the accuracy of the generative AI model.

[0225] The server uses the preprocessed data to train the generative AI model. This process uses machine learning frameworks such as TENSORFLOW (registered trademark) and PyTorch. Recurrent neural networks (RNN) and long short-term memory (LSTM) are used as learning algorithms. Once training is complete, the server saves the learning results. Saved models are managed in formats such as HDF5 and Pickle.

[0226] Based on the learning results, the server predicts the user's future behavior and generates appropriate actions. For example, if the user is conducting an evacuation drill, the server generates the optimal evacuation route based on past data. This information is sent to the user's mobile device.

[0227] Terminal handling

[0228] The device receives the learning results and prediction data sent from the server. The received data is used as action instructions for the user. Specifically, evacuation routes and action instructions are displayed on a map. Map services such as Google (registered trademark) Maps API are used to display the information in a visually easy-to-understand format.

[0229] The device accepts input from the user. For example, if the user instructs the device to "display evacuation routes," the device will display the appropriate information in accordance with the instruction.

[0230] User Action

[0231] The user checks the action instructions displayed on the device. For example, during an evacuation drill, the user checks the evacuation route displayed on the device. The user acts based on the information provided by the device. Specifically, the user follows the displayed evacuation route and heads to a shelter.

[0232] After taking action, the user provides feedback. For example, they can input whether the evacuation route was appropriate or not. This feedback is sent to the server and used as learning data for the next time.

[0233] Specific examples

[0234] For example, consider a case where a user preferentially instructs a route to a shelter that the user has visited in the past during evacuation drills. The server analyzes the user's past evacuation drill data and generates the optimal evacuation route. This information is sent to the user's device and is useful when the user evacuates.

[0235] Prompt Sentence Examples

[0236] "Generate optimal evacuation routes based on the user's past behavioral history. Use past evacuation drill data to prioritize routes to evacuation shelters."

[0237] By inputting this prompt sentence into the generative AI model, an optimal evacuation route is generated based on the user's behavior history. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0238] Step 1:

[0239] Data collection

[0240] The server collects GPS data, timestamps, and types of activity (e.g., evacuation drills, commuting, etc.) from the user's mobile device. The user's location information and activity history data are provided as input. The server receives this data and stores it in a database. Specifically, it periodically retrieves data from the mobile device and stores it in a relational database such as MySQL or PostgreSQL.

[0241] Step 2:

[0242] Data Preprocessing

[0243] The server preprocesses the stored behavioral history data. Raw data stored in the database is provided as input. Preprocessing includes missing value completion, data normalization, and outlier removal. This improves the quality of the data and increases the accuracy of the generative AI model. Specifically, the server uses Python's Pandas library to clean the data and perform the necessary preprocessing.

[0244] Step 3:

[0245] Model learning

[0246] The server trains a generative AI model using the preprocessed data. The preprocessed data is provided as input. Machine learning frameworks such as TensorFlow and PyTorch are used for training. Recurrent neural networks (RNNs) and long short-term memory (LSTMs) are applied as training algorithms. A trained model is obtained as output. Specifically, the data is fed into the model and the model parameters are updated every epoch.

[0247] Step 4:

[0248] Save Model

[0249] The server saves the model after training is complete. The trained model is provided as input. The model is saved in formats such as HDF5 or Pickle. The output is the saved model file. Specifically, the model is written to a file in the specified format and saved in storage.

[0250] Step 5:

[0251] Prediction and Action Generation

[0252] The server uses the stored model to predict the user's future behavior and generate appropriate actions. The input is the trained model and the user's current location information. The output is the generated action instructions. Specifically, the current data is input into the model, and the optimal action is generated based on the prediction results.

[0253] Step 6:

[0254] Data reception

[0255] The device receives the learning results and prediction data sent from the server. As input, action instruction data is provided from the server. As output, the received data is saved on the device. Specifically, the device receives data from the server and saves it in local storage.

[0256] Step 7:

[0257] Data Display

[0258] The device displays the received data to the user. The received action instruction data is provided as input. The output is information that is visually displayed to the user. Specifically, the device uses the Google Maps API to display evacuation routes on a map.

[0259] Step 8:

[0260] User Interaction

[0261] The user checks the action instructions displayed on the device and provides feedback as necessary. The information displayed on the device is provided as input. The user's feedback is obtained as output. In concrete terms, the user checks the evacuation route and inputs feedback into the device.

[0262] Step 9:

[0263] Gathering feedback

[0264] The server collects feedback from users and uses it as training data for the next model. User feedback data is provided as input. Updated training data is obtained as output. Specifically, the feedback data is saved in a database and used for the next model training.

[0265] (Application example 3)

[0266] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0267] Conventional disaster evacuation systems only provide general evacuation routes without taking into account the user's past behavioral history, making it difficult to provide the optimal evacuation route for the user. Furthermore, while quick and accurate evacuation instructions are required in emergencies, conventional systems were inadequate in this regard. This meant that effective evacuation support to ensure the safety of users could not be achieved.

[0268] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0269] In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating actions to be taken based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, and navigation means for learning the user's past behavioral history and providing the optimal evacuation route in an emergency based on the learning results. This makes it possible to provide the optimal evacuation route taking the user's past behavioral history into consideration, and to issue quick and accurate evacuation instructions in an emergency.

[0270] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0271] "Analysis" is a process of examining the details of the received emergency alert email to determine the type and situation of the disaster.

[0272] "Type of disaster" refers to the specific classification of the disaster that occurred, such as earthquake, tsunami, fire, or flood.

[0273] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[0274] "AI means" is a system that uses artificial intelligence technology to analyze received information and generate appropriate response actions.

[0275] "Mobile devices" refer to electronic devices that users carry and use, such as smartphones and tablets.

[0276] "Display" refers to visually showing the generated response action on the screen of the user's mobile device.

[0277] "Past behavioral history" refers to records of the actions a user has taken and the places they have visited in the past.

[0278] "Learning" refers to data processing to predict future behavior based on past behavioral history.

[0279] "Navigation means" is a system that guides users to the optimal route.

[0280] "Evacuation route" refers to the route that users can take to safely evacuate in the event of an emergency.

[0281] The following system configuration will be described as an embodiment of the present invention.

[0282] System Configuration

[0283] The system consists of the following major components:

[0284] 1. Server: Receives emergency alert emails, analyzes them, and determines the type of disaster.

[0285] 2. Mobile device: The user's smartphone or tablet receives instructions from the server.

[0286] 3. AI means: Installed on the server, it generates the actions to be taken based on the received information.

[0287] 4. Navigation: Learns the user's past behavior history and provides the optimal evacuation route.

[0288] Program processing explanation

[0289] The server uses an Internet connection to receive emergency alert emails. It analyzes the content of the emails and uses natural language processing technology to determine the type of disaster. Specifically, it uses a Python natural language processing library (e.g., NLTK or spaCy).

[0290] The AI ​​solution takes into account the user's location and past behavioral history to generate optimal countermeasures using machine learning models (e.g., TensorFlow and PyTorch). The generated countermeasures are displayed on the mobile device.

[0291] The navigation system learns the user's past behavior history and provides the optimal evacuation route in case of an emergency. This is done by linking with a geographic information system (GIS) and comparing the user's current location with past evacuation routes. Specifically, it uses the Geopy library for distance calculations.

[0292] Specific examples

[0293] For example, if a user previously visited a location called "shelter A" during an evacuation drill and their current location is "point B," the system will provide the optimal evacuation route from "point B" to "shelter A."

[0294] Prompt Sentence Examples

[0295] Develop a navigation system that learns from the user's past behavioral history and provides the optimal evacuation route in an emergency. Include a function that prioritizes routes to places the user has previously visited during evacuation drills.

[0296] In this way, users can receive prompt and accurate evacuation instructions in an emergency, which will enable effective evacuation support to ensure the safety of users.

[0297] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0298] Step 1:

[0299] The server receives emergency alert emails via an Internet connection.

[0300] Input: Emergency Alert Email

[0301] Output: Contents of the received emergency alert email

[0302] Specific operation: The server retrieves emergency alert emails from the mail server and saves their contents in text format.

[0303] Step 2:

[0304] The server analyzes the contents of the received emergency alert email and determines the type of disaster.

[0305] Input: Contents of the emergency alert email received

[0306] Output: Disaster type

[0307] What it does: The server uses natural language processing libraries (e.g., NLTK or spaCy) to analyze the content of the email and identify the type of disaster, such as earthquake, tsunami, or fire, based on keywords and context.

[0308] Step 3:

[0309] The server generates an action to be taken based on the determined type of disaster.

[0310] Input: Disaster Type

[0311] Output: Action to be taken

[0312] Specific operation: The server uses machine learning models (e.g., TensorFlow or PyTorch) to generate optimal response actions depending on the type of disaster. For example, in the case of an earthquake, it generates actions such as "evacuate the building," and in the case of a tsunami, it generates actions such as "evacuate to higher ground."

[0313] Step 4:

[0314] The server transmits the generated action to be taken to the user's mobile terminal.

[0315] Input: Action to be taken

[0316] Output: Actions displayed on mobile device

[0317] Specific operation: The server sends the generated response action to the user's mobile device as a push notification and displays it on the device screen.

[0318] Step 5:

[0319] The server learns the user's past behavioral history and provides the optimal evacuation route based on the results of that learning.

[0320] Input: User's past behavior history, current location

[0321] Output: Optimal evacuation route

[0322] Specific operation: The server retrieves the user's behavior history from the past behavior history database and learns it using a machine learning model. The current location is obtained using GPS data, and the Geopy library is used to compare the current location with past evacuation routes and calculate the optimal evacuation route.

[0323] Step 6:

[0324] The server sends the optimal evacuation route to the user's mobile device.

[0325] Input: Optimal evacuation route

[0326] Output: Evacuation route displayed on mobile device

[0327] Specific operation: The server sends the calculated optimal evacuation route to the user's mobile device and displays it through the navigation app. The user can then evacuate safely by following the displayed route.

[0328] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0329] "Example 1"

[0330] In one embodiment of the present invention, the AI ​​means includes an emotion engine that recognizes the user's emotions by inferring them from the user's facial expressions, tone of voice, text input, etc. For example, if the user is scared, the AI ​​uses that information to generate calmer behavioral instructions.

[0331] "Example 2"

[0332] In another embodiment of the invention, the AI ​​means takes into account the user's location and emotions to generate a response action. For example, if the user is in a public place and feels scared, the AI ​​can use that information to instruct them to evacuate to a safer, less crowded place.

[0333] "Example 3"

[0334] In yet another embodiment of the present invention, the AI ​​means learns from the user's past behavioral history and emotions and generates countermeasures based on the learning results. For example, if the user has a tendency to panic in the past, the AI ​​will use that information to generate more specific and concise instructions.

[0335] The processing flow of each embodiment will be described below.

[0336] "Example 1"

[0337] Step 1: Step 1: The AI ​​means uses an emotion engine to estimate emotions from the user's facial expressions, tone of voice, text input, etc.

[0338] Step 2: The AI ​​means generates behavioral instructions for the user taking into account the estimated emotion. For example, if the user is scared, the AI ​​uses that information to generate gentler behavioral instructions.

[0339] "Example 2"

[0340] Step 1: Step 1: The AI ​​means estimates the user's emotions using the user's location information and the emotion engine.

[0341] Step 2: The AI ​​generates instructions for the user based on the estimated emotion and location. For example, if the user is in a public place and feels scared, the AI ​​can use that information to instruct them to evacuate to a safer location with fewer people.

[0342] "Example 3"

[0343] Step 1: Step 1: The AI ​​means estimates the user's emotions using the user's past behavior history and the emotion engine.

[0344] Step 2: The AI ​​generates instructions for the user based on the estimated emotions and their past behavioral history. For example, if the user has a history of panic, the AI ​​uses that information to generate more specific and concise instructions.

[0345] Example 1

[0346] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0347] Conventional emergency alert systems can receive disaster information and provide users with action instructions, but they are unable to generate action instructions that take into account the user's emotional state. This means that if a user is feeling fear or anxiety, it becomes difficult for them to take appropriate action. Another issue is that they cannot provide individualized responses that take into account the user's location information or past behavioral history, so they can only provide generic instructions.

[0348] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0349] In this invention, the server includes a means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information; a means for analyzing the content of the received emergency alert email and determining the type of disaster; an artificial intelligence means for generating appropriate action to be taken based on the determined type of disaster; a means for displaying the generated action on the user's mobile information terminal; a means for recognizing the user's emotions and modifying action instructions based on the emotions; and a means for displaying the modified action instructions on the user's mobile information terminal. This makes it possible to provide appropriate action instructions that take into account the user's emotional state, allowing the user to take more appropriate action even when feeling fear or anxiety. It also makes it possible to provide individualized responses that take into account the user's location information and past behavioral history.

[0350] An "emergency alert email" is an email sent to quickly notify users of emergency information such as earthquakes, tsunamis, and disaster evacuation information.

[0351] "Mobile information terminal" refers to a portable information processing device such as a smartphone or tablet.

[0352] "Artificial intelligence means" is a system that includes algorithms and programs for analyzing received information and generating appropriate instructions for action.

[0353] "Means for recognizing emotions" refers to technology for inferring emotions from a user's facial expressions, tone of voice, text input, etc.

[0354] "Action instructions" are instructions that indicate specific actions that the user should take depending on the type of disaster.

[0355] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[0356] "Past behavior history" is a record of the user's past behaviors, and is data used to generate future behavioral instructions based on this record.

[0357] This invention is a system that receives emergency emails about earthquakes, tsunamis, disaster evacuation information, etc., and provides users with appropriate instructions on what to do. This system is mainly composed of three elements: a server, a terminal (mobile information terminal), and a user.

[0358] Hardware and software used

[0359] Hardware: Mobile devices (smartphones, tablets, etc.)

[0360] Software: Emergency alert email reception app, AI analysis engine, emotion engine

[0361] System program processing

[0362] Server Processing

[0363] The server analyzes the emergency alert email received from the mobile information terminal. This analysis is performed using an internal AI analysis engine. This AI analysis engine uses natural language processing technology to analyze the content of the emergency alert email and identify the type of disaster. For example, if an earthquake alert email is received, the AI ​​analysis engine will analyze the content and determine that it is an earthquake.

[0364] The server generates action instructions based on the results of the AI ​​analysis engine according to the type of disaster. The action instructions are generated based on pre-set templates. The generated action instructions are then sent to the mobile information device via an internet connection.

[0365] Terminal handling

[0366] The device receives the action instructions sent from the server and displays them on the screen. The display uses text formatted for easy viewing by the user. In addition, the device is equipped with an emotion engine that recognizes emotions from the user's facial expressions, tone of voice, and text input. This emotion engine estimates whether the user is scared or not.

[0367] User Action

[0368] The user acts according to the instructions displayed on the device. If the user's emotions affect the system, the emotion engine sends that information to the server. The server then modifies the instructions based on the results of the emotion engine. For example, if the user is scared, the AI ​​uses that information to generate gentler instructions. The modified instructions are then sent back to the device, which then displays them on the screen.

[0369] Specific examples

[0370] Example 1: Earthquake Early Warning

[0371] 1. Device: Receive emergency alert emails.

[0372] 2. Terminal: Sends the received emergency alert email to the server.

[0373] 3. Server: Analyzes the received emergency alert email using an AI analysis engine.

[0374] 4. Server: Based on the analysis results, it determines that it is an earthquake.

[0375] 5. Server: Generates an action instruction such as "An earthquake has occurred. Take cover under a desk."

[0376] 6. Server: Sends the generated action instructions to the terminal.

[0377] 7. Terminal: Displays the received action instructions on the screen.

[0378] 8. Device: Recognizes when a user is scared by their facial expression and tone of voice.

[0379] 9. Server: Based on the user's emotional information, generate a gentle instruction for action such as "Please stay calm and stay here until it is safe."

[0380] 10. Server: Sends the modified action instructions to the device.

[0381] 11. Terminal: Displays the corrected action instructions on the screen.

[0382] Example prompts to input to the generative AI model

[0383] "I received an earthquake alert email. If the user is scared, what action instructions should I generate?"

[0384] By inputting this prompt into a generative AI model, it can learn how to generate appropriate behavioral instructions based on the user's emotions.

[0385] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0386] Step 1:

[0387] The terminal receives an emergency alert email. As input, the emergency alert email arrives at the mobile information terminal. As output, the data of the received emergency alert email is obtained. In concrete terms, the terminal's email reception function is activated and the emergency alert email is received.

[0388] Step 2:

[0389] The terminal sends the received emergency alert email to the server. The input is the data of the received emergency alert email. The output is the data of the emergency alert email sent to the server. Specifically, the terminal's communication function is activated and the data is sent to the server via the Internet.

[0390] Step 3:

[0391] The server analyzes the received emergency alert email using an AI analysis engine. The input is the data from the emergency alert email. The output is the analysis result, which indicates the type of disaster. Specifically, the server's AI analysis engine analyzes the content of the email using natural language processing technology and identifies the type of disaster.

[0392] Step 4:

[0393] The server determines the type of disaster based on the analysis results. The input is the analysis results of the AI ​​analysis engine. The output is the determined type of disaster. Specifically, the server evaluates the analysis results and determines, for example, that it is an "earthquake."

[0394] Step 5:

[0395] The server generates action instructions based on the type of disaster it has determined. The input is the type of disaster it has determined. The output is the generated action instructions. As a specific operation, the server's AI generates action instructions based on a pre-set template.

[0396] Step 6:

[0397] The server sends the generated action instructions to the terminal. The generated action instructions are the input. The action instructions sent to the terminal are obtained as the output. In concrete terms, the server's communication function is activated and data is sent to the terminal via the Internet.

[0398] Step 7:

[0399] The terminal displays the received action instructions on the screen. The input is the action instructions sent from the server. The output is the action instructions displayed on the screen. Specifically, the display function of the terminal is activated and the action instructions are displayed to the user.

[0400] Step 8:

[0401] The device recognizes emotions from the user's facial expressions, tone of voice, text input, etc. Inputs include the user's facial expression data, voice data, and text data. The output is the recognized user's emotional information. Specifically, the device's emotion engine estimates the user's emotions.

[0402] Step 9:

[0403] The server modifies the action instructions based on the results of the emotion engine. The input is the recognized user's emotional information. The output is the modified action instructions. Specifically, the server's AI evaluates the user's emotional information and generates gentler action instructions.

[0404] Step 10:

[0405] The server sends the modified action instructions to the terminal. The input is the modified action instructions. The output is the modified action instructions sent to the terminal. As a specific operation, the communication function of the server is activated and data is sent to the terminal via the Internet.

[0406] Step 11:

[0407] The terminal displays the modified action instructions on the screen. The input is the modified action instructions sent from the server. The output is the modified action instructions displayed on the screen. As a specific operation, the display function of the terminal is activated to display the modified action instructions to the user.

[0408] (Application example 1)

[0409] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0410] Conventional disaster response systems can receive emergency alert emails, determine the type of disaster, and generate appropriate response actions, but they cannot respond by taking into account the user's emotional state, which can lead to users being unable to take appropriate action.There is also a need for more effective evacuation behavior by generating action instructions that correspond to the user's emotions.

[0411] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert email and determining the type of disaster, AI means including an emotion engine for recognizing the user's emotions, and means for displaying the generated actions on the user's smart device. This allows appropriate action instructions to be generated according to the user's emotional state, enabling the user to take evacuation actions more effectively.

[0412] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0413] "Analysis" is a process of examining the details of the received emergency alert email and identifying the type of disaster.

[0414] "Type of disaster" refers to the specific classification of the disaster that occurred, such as earthquake, tsunami, fire, or flood.

[0415] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[0416] "AI means" refers to means for analyzing data using artificial intelligence technology and making judgments and predictions.

[0417] An "emotion engine" is a technology that estimates emotions from a user's facial expressions, tone of voice, text input, etc.

[0418] "Smart devices" refers to portable electronic devices with internet connectivity, such as smartphones and tablets.

[0419] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[0420] "Behavioral history" refers to a record of the actions a user has taken in the past, and is the data that AI uses to learn.

[0421] A system for implementing the present invention is configured as follows.

[0422] First, the server is provided with a means for receiving emergency alert emails such as earthquake, tsunami, disaster evacuation information, etc. This allows the server to quickly receive emergency alert emails and provide information to users immediately.

[0423] The server then analyzes the content of the received emergency alert email and determines the type of disaster. This analysis is performed using an AI analysis engine (such as TensorFlow). The AI ​​analysis engine analyzes the text data in the emergency alert email and identifies the type of disaster, such as earthquake, tsunami, or fire.

[0424] Furthermore, the server is equipped with AI means, including an emotion engine that recognizes the user's emotions. The emotion engine uses technologies such as OpenCV and Google Cloud Natural Language API to analyze the user's facial expressions and tone of voice, allowing it to estimate the user's emotional state, such as whether they are scared or calm.

[0425] The server generates the appropriate action to take based on the type of disaster it has determined and the user's emotional state. The generated action instructions are displayed on the user's smart device (e.g., smartphone or tablet). React Native and other frameworks are used as user interface (UI) frameworks.

[0426] For example, when an earthquake alert is received, the AI ​​displays, "An earthquake has occurred. Please evacuate to a safe place." If the user is scared, the emotion engine recognizes this information and generates a gentler instruction to act, such as, "Please stay calm. Evacuate to a safe place."

[0427] Examples of prompts for generative AI models include:

[0428] Prompt: "An earthquake alert has been received. If the user is scared, generate calm instructions for action."

[0429] In this way, appropriate action instructions can be generated according to the user's emotional state, enabling the user to take more effective evacuation actions.

[0430] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0431] Step 1:

[0432] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the emergency alert email, and the output is the content of the received email. Specifically, the server uses the emergency alert email reception API to obtain emergency alert emails in real time.

[0433] Step 2:

[0434] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the content of the received email, and the output is the identified type of disaster. Specifically, the server uses an AI analysis engine (such as TensorFlow) to analyze the text data in the email and identify the type of disaster, such as earthquake, tsunami, or fire.

[0435] Step 3:

[0436] The server uses an emotion engine to recognize the user's emotions. The inputs are the user's facial expressions, tone of voice, and text input, and the output is the estimated user's emotional state. Specifically, the server captures the user's facial expressions and voice using the smart device's camera and microphone, and estimates the emotion using technologies such as OpenCV and Google Cloud Natural Language API.

[0437] Step 4:

[0438] The server generates the action to be taken based on the determined type of disaster and the user's emotional state. The input is the identified type of disaster and the estimated user's emotional state, and the output is the generated action instructions. Specifically, the server uses a generative AI model to generate action instructions according to the type of disaster and the user's emotions.

[0439] Step 5:

[0440] The server displays the generated action instructions on the user's smart device. The input is the generated action instructions, and the output is the action instructions displayed on the smart device screen. Specifically, the server uses a user interface (UI) framework (such as React Native) to display the action instructions in a user-friendly format.

[0441] Step 6:

[0442] The user takes evacuation action according to the action instructions displayed on the smart device. The input is the action instructions displayed on the smart device, and the output is the user's evacuation action. Specifically, the user follows the instructions displayed on the smart device screen and evacuates to a safe place.

[0443] In this way, the server, terminal, and user work together to generate appropriate action instructions according to the user's emotional state, enabling the user to take more effective evacuation actions.

[0444] Example 2

[0445] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0446] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information or emotions, making it difficult to provide optimal evacuation actions for each individual user. They also lacked the ability to learn from the user's past behavioral history and suggest appropriate evacuation actions. This led to problems such as users being unable to take appropriate evacuation actions, and safety in the event of a disaster not being ensured.

[0447] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0448] In this invention, the server includes means for receiving emergency news messages such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency news message to determine the type of disaster, means for acquiring the user's location information, means for analyzing the user's emotions, means for using a generative AI model to generate an action to be taken based on the determined type of disaster, the acquired location information, and the analyzed emotions, and means for displaying the generated action on the user's mobile device. This makes it possible to provide appropriate evacuation actions that take into account the user's location information and emotions, thereby improving safety during disasters.

[0449] An "emergency message" is a message that includes information about an emergency, such as an earthquake, tsunami, or disaster evacuation information.

[0450] "Type of disaster" refers to the classification of different disasters such as earthquake, tsunami, fire, and flood.

[0451] "Location Information" is data that indicates a user's current geographic location.

[0452] "Emotion" is data that indicates the psychological state of the user, and includes states such as fear, relief, and excitement.

[0453] A "generative AI model" is an artificial intelligence model that generates appropriate actions or answers based on input data.

[0454] "Mobile device" refers to a portable electronic device such as a smartphone or tablet.

[0455] "Behavioral history" refers to a record of actions taken by a user in the past.

[0456] "Analysis" is the process of examining received data in detail and understanding its contents.

[0457] "Learning" is the process of using past data to improve future actions and decisions.

[0458] This invention relates to a system that generates actions to be taken by taking into account the user's location information and emotions. This system operates in cooperation with a server, a terminal, and a user.

[0459] First, the server receives emergency alert messages such as earthquake, tsunami, and disaster evacuation information. These include emergency alert messages sent over general communication networks. The server analyzes the content of the received emergency alert message and determines the type of disaster. Natural language processing technology can be used for the analysis.

[0460] Next, the server uses the device's GPS function to obtain the user's location information. Specifically, it obtains the user's current location using an API that provides location information services (e.g., a map service API). The device then sends the GPS data to the server.

[0461] Furthermore, the server receives text and voice data entered by the user to analyze the user's emotions. Sentiment analysis is performed using a natural language processing library (e.g., sentiment analysis API). The user enters text and voice data into the terminal and sends it to the server.

[0462] The server uses a generative AI model to generate the appropriate action to be taken based on the acquired location information and emotion data. For example, a generative AI model is used. The server displays the generated action on the user's mobile device, allowing the user to take appropriate evacuation action.

[0463] As a concrete example, if a user is in a tsunami warning area, the server uses a location service API to obtain the user's location information and confirms that the user is within the tsunami warning area. The generative AI model then receives a prompt message: "Please tell me what to do if the user is in a tsunami warning area." The generative AI model generates the action "Evacuate to the nearest high ground," and the server notifies the user of this information.

[0464] If the user is scared in a public place, the server uses the sentiment analysis API to analyze the user's emotions and inputs the prompt "What should I do if the user is scared in a public place?" into the generative AI model. The generative AI model generates an action to "instruct the user to evacuate to a safe place with fewer people," and the server notifies the user of this information.

[0465] Examples of prompts include:

[0466] 1. "What should I do if I'm in a tsunami warning area?"

[0467] 2. "What should I do if a user is in a public place and feels scared?"

[0468] In this way, a system is realized in which the server, terminal, and user work together to generate coping actions that take into account the user's location information and emotions, and provide appropriate instructions.

[0469] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0470] Step 1:

[0471] The server receives emergency messages such as earthquake, tsunami, and disaster evacuation information.

[0472] Input: Emergency message

[0473] Specific operations: A server receives an emergency alert message transmitted over a communication network.

[0474] Output: Received emergency alert message

[0475] Step 2:

[0476] The server analyzes the contents of the received emergency message and determines the type of disaster.

[0477] Input: Received emergency alert message

[0478] Specific operation: The server uses natural language processing technology to analyze the content of the message and identify the type of disaster, such as earthquake, tsunami, or fire.

[0479] Output: Disaster type

[0480] Step 3:

[0481] The server uses the device's GPS function to obtain the user's location information.

[0482] Input: User location request

[0483] Specific operation: The device turns on the GPS function to obtain location information and sends it to the server through the location information service API.

[0484] Output: Current location of the user

[0485] Step 4:

[0486] The server receives text and voice data entered by the user in order to analyze the user's sentiment.

[0487] Input: User text or voice data

[0488] How it works: The user inputs text or voice data into the device and sends it to the server, which then uses the sentiment analysis API to analyze the data and identify the user's emotions.

[0489] Output: User emotion data

[0490] Step 5:

[0491] The server uses a generative AI model that generates appropriate actions based on the type of disaster it determines, the location information it obtains, and the emotions it analyzes.

[0492] Input: Type of disaster, user's current location, user's emotional data

[0493] Specific behavior: The server inputs a prompt sentence such as "Please tell me what to do if the user is in a tsunami warning area" into the generative AI model, and the generative AI model generates appropriate behavior.

[0494] Output: Action to be taken

[0495] Step 6:

[0496] The server displays the generated actions on the user's mobile device.

[0497] Input: Action to be taken

[0498] Specific operation: The server notifies the device of the generated action and sends the user a message such as "Please evacuate to the nearest high ground."

[0499] Output: A message to inform the user

[0500] In this way, a system is realized in which the server, terminal, and user work together to generate coping actions that take into account the user's location information and emotions, and provide appropriate instructions.

[0501] (Application example 2)

[0502] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0503] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information or emotional state, making it difficult to provide appropriate evacuation actions tailored to individual situations. Furthermore, if the user is in a state of panic or is in a specific location, the system is unable to provide appropriate evacuation instructions, reducing the effectiveness of the evacuation. This has led to the issue of evacuation actions not functioning adequately to ensure the user's safety.

[0504] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0505] In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert email and determining the type of disaster, AI means for generating an action to be taken based on the determined type of disaster, means for displaying the generated action on the user's mobile device, means for acquiring the user's location information, means for recognizing the user's emotional state, means for generating an action to be taken based on the acquired location information and the recognized emotional state, and means for notifying the user of the generated action. This makes it possible to instruct appropriate evacuation actions taking into account the user's location information and emotional state.

[0506] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0507] "Type of disaster" refers to the specific type of disaster that occurred, such as earthquake, tsunami, fire, or flood.

[0508] "AI means" refers to means that use artificial intelligence technology to analyze received information and generate appropriate response actions.

[0509] "Mobile devices" refer to electronic devices that users carry and use, such as smartphones and tablets.

[0510] "Location information" refers to the geographic coordinate information of the user's current location.

[0511] "Emotional state" refers to the user's current psychological state, such as fear, relief, panic, etc.

[0512] "Evacuation behavior" refers to specific actions that users should take in the event of a disaster, such as evacuating to higher ground or moving to a safe location.

[0513] The "notification means" refers to a means for informing the user of the generated evacuation action, and includes, for example, a push notification to a mobile device or a voice alert.

[0514] The system for implementing this invention includes a user's mobile terminal, a server, and AI means. The specific configuration and operation of the system will be described below.

[0515] System configuration

[0516] 1. Mobile devices: Mobile devices such as smartphones and tablets are equipped with a GPS sensor to acquire the user's location information, a camera and microphone to recognize the user's emotional state, and an application to receive emergency alert emails and display generated evacuation actions.

[0517] 2. Server: The server receives emergency alert emails, analyzes their contents to determine the type of disaster, and is equipped with AI means to generate appropriate evacuation actions taking into account the user's location and emotional state.

[0518] 3. AI Means: The AI ​​means uses a generative AI model to analyze the received information and generate appropriate evacuation actions. Specifically, it receives the user's location information and emotional state as input and generates prompt sentences.

[0519] System Operation

[0520] 1. Receiving and analyzing emergency alert emails: The server receives emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc. It analyzes the content of the received emails and determines the type of disaster.

[0521] 2. Location and emotional state acquisition: The mobile device acquires the user's location using the GPS sensor, and recognizes the user's emotional state using the camera and microphone.

[0522] 3. Evacuation behavior generation: The server's AI generates appropriate evacuation behavior based on the acquired location information and emotional state. For example, if the user is in a tsunami warning area, it will instruct the user to evacuate to the nearest high ground. If the user is in a panic, it will instruct the user to evacuate to a safe place with fewer people.

[0523] 4. Evacuation Action Notification: The mobile device notifies the user of the generated evacuation action via push notification and / or voice alert.

[0524] Specific examples

[0525] For example, if a user is near the coast and a tsunami warning is issued, the system operates as follows:

[0526] 1. The server receives an emergency email alerting the tsunami and determines whether a tsunami has occurred.

[0527] 2. The mobile device uses GPS sensors to obtain the user's location and cameras and microphones to recognize the user's emotional state.

[0528] 3. The server's AI means recognizes that the user is near the coast and is scared, and generates the following prompt:

[0529] The user is currently near the coast. A tsunami warning has been issued and the user is scared. Please tell them to evacuate to the nearest high ground.

[0530] 4. The mobile device notifies the user of the generated evacuation action and instructs them to evacuate to the nearest high ground.

[0531] In this way, it becomes possible to provide appropriate evacuation instructions that take into account the user's location information and emotional state.

[0532] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0533] Step 1:

[0534] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. It receives the emergency alert email as input, analyzes its content, and determines the type of disaster. It generates data that identifies the type of disaster as output. Specifically, it performs text analysis on the email content, extracts keywords and phrases, and classifies the type of disaster.

[0535] Step 2:

[0536] The device obtains the user's location information using the GPS sensor. It receives GPS data as input and determines the latitude and longitude of the user's current location. It generates the user's location information as output. Specifically, it analyzes the signal from the GPS sensor and calculates the user's precise location.

[0537] Step 3:

[0538] The device uses a camera and microphone to recognize the user's emotional state. It receives camera images and audio data as input and applies emotion recognition algorithms to identify the user's emotional state. As output, it generates data that indicates the user's emotional state. Specifically, it performs image and audio analysis to classify the user's emotions, such as fear or relief.

[0539] Step 4:

[0540] The server's AI means generates appropriate evacuation actions based on the acquired location information and emotional state. It receives location information and emotional state data as input and generates prompt sentences using a generative AI model. It generates instructions for evacuation actions as output. Specifically, it incorporates location information and emotional state into the prompt sentences to create specific evacuation instructions for the user.

[0541] Step 5:

[0542] The device notifies the user of the generated evacuation behavior. It receives evacuation behavior instructions as input and notifies the user via push notification or voice alert. It outputs the evacuation instructions to the user. Specifically, it uses the device's notification function to inform the user of the evacuation behavior via screen display or voice message.

[0543] In this way, it becomes possible to provide appropriate evacuation instructions that take into account the user's location information and emotional state.

[0544] Example 3

[0545] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0546] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's past behavioral history or emotional data, making it difficult to provide optimal evacuation instructions for each individual user. Furthermore, because evacuation instructions do not take into account the user's location information in real time, it is difficult to provide prompt and appropriate evacuation instructions in an emergency. This can easily lead to users panicking and delays in evacuation.

[0547] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0548] In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, artificial intelligence means for generating response actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for collecting data on the user's past behavioral history, means for preprocessing the collected data, means for training an artificial intelligence model using the preprocessed data, means for generating response actions using the trained artificial intelligence model, means for adjusting the generated response actions in consideration of the user's emotional data, and means for transmitting the adjusted response actions to the user's mobile device. This enables individually optimized evacuation instructions that take into consideration the user's past behavioral history and emotional data, thereby realizing quick and appropriate evacuation actions in emergencies.

[0549] "Emergency emails for earthquakes, tsunamis, disaster evacuation information, etc." are emails containing emergency information that are sent to users promptly in the event of a disaster.

[0550] "Means for receiving" refers to a device or software that has a communication function for receiving emergency alert emails.

[0551] The "means for analyzing and determining the type of disaster" refers to an algorithm or program that analyzes the contents of the received emergency alert email and identifies the type of disaster, such as earthquake, tsunami, or fire.

[0552] The "artificial intelligence means for generating appropriate actions to be taken" refers to an artificial intelligence model and its execution environment for generating appropriate evacuation actions and countermeasures based on the type of disaster.

[0553] The "means for displaying on the user's mobile device" refers to an interface and software for displaying the generated action instructions on the user's mobile device such as a smartphone or tablet.

[0554] "Means for collecting user's past behavioral history data" refers to sensors and databases that collect data on the user's past behavior and the places they have visited.

[0555] "Means for preprocessing collected data" refers to algorithms or programs for preprocessing collected data, such as filling in missing values ​​and normalizing the data.

[0556] A "means for training an artificial intelligence model using preprocessed data" is a machine learning framework and its execution environment for training an artificial intelligence model using preprocessed data.

[0557] "Means for generating response actions using a trained artificial intelligence model" refers to algorithms and programs for generating specific evacuation actions and countermeasures for users using a trained artificial intelligence model.

[0558] The "means for taking into account the user's emotional data and making adjustments" refers to an algorithm or program for adjusting the generated coping behavior in consideration of the user's emotional state.

[0559] The "means for transmitting the adjusted response action to the user's mobile device" refers to communication functions and software for transmitting the adjusted action instructions to the user's mobile device.

[0560] MODE FOR CARRYING OUT THE INVENTION

[0561] This invention is a system that provides users with prompt and appropriate evacuation instructions in the event of a disaster. The system consists of three main components: a server, a terminal, and users.

[0562] server

[0563] The server has the following functions:

[0564] 1. Data Collection

[0565] The server collects the user's past behavioral history data, including information about the evacuation shelters the user visited during evacuation drills and past behavioral patterns.

[0566] The server also collects user emotional data, including whether the user has a history of panicking.

[0567] 2. Data Preprocessing

[0568] The server preprocesses the collected data, specifically by filling in missing values ​​and normalizing the data.

[0569] The server converts the pre-processed data into a format suitable for training an AI model.

[0570] 3. Training the AI ​​model

[0571] The server uses the preprocessed data to train an AI model using frameworks such as TensorFlow and PyTorch.

[0572] The server evaluates the model's performance during training and adjusts hyperparameters as needed.

[0573] 4. Generating Action Instructions

[0574] The server uses the trained AI model to generate response actions based on the user's behavioral history, for example, giving priority to routes to evacuation shelters visited during evacuation drills.

[0575] The server takes emotion data into account and generates more specific and concise instructions for actions, for example, providing concise and clear evacuation instructions to users who tend to panic.

[0576] 5. Sending instructions

[0577] The server sends the generated action instructions to the device using the HTTPS protocol.

[0578] Terminal

[0579] The terminal has the following functions:

[0580] 1. Receiving instructions for action

[0581] The terminal receives the action instructions sent from the server. Devices such as smartphones and tablets are used for receiving the instructions.

[0582] 2. Display of action instructions

[0583] The terminal displays the received action instructions to the user using the application's user interface.

[0584] 3. Obtaining location information

[0585] The device obtains the user's current location information, which is obtained using the GPS function.

[0586] The terminal transmits the acquired location information to the server.

[0587] 4. Update of Action Instructions

[0588] The terminal receives updated action instructions from the server in real time and notifies the user.

[0589] user

[0590] The user performs the following actions:

[0591] 1. Confirmation of action instructions

[0592] The user checks the action instructions displayed on the device.

[0593] 2. Taking action

[0594] During evacuation drills and actual evacuations, users act based on instructions from the device.

[0595] 3. Sending location information

[0596] Users send their location information to the server via their devices.

[0597] 4. Check for updated instructions

[0598] The user checks the updated action instructions from the terminal and corrects the action if necessary.

[0599] Specific examples

[0600] Data collection: The server stores information about evacuation shelters that users have visited in the past during evacuation drills in a database. For example, it stores data such as "User A visited evacuation shelter B on March 15, 2023."

[0601] Data preprocessing: The server imputes missing values ​​in the collected data and normalizes the data. For example, it estimates and imputes missing dates and times of visits to evacuation centers.

[0602] Training the AI ​​model: The server uses the preprocessed data to train the AI ​​model, for example, using TensorFlow to build a model that learns user behavior patterns.

[0603] Generation of action instructions: The server uses the trained AI model to generate action instructions for user A, such as "give priority to the route to shelter B."

[0604] Sending action instructions: The server sends the generated action instructions to the terminal. For example, it sends a message via HTTPS saying, "Give priority to the route to shelter B."

[0605] Prompt Sentence Examples

[0606] "Build an AI model that learns the user's past behavioral history and prioritizes routes to evacuation shelters visited during evacuation drills."

[0607] "Please design a system that learns the user's past behavioral history and emotional data, and generates concise and clear instructions for users who tend to panic." The flow of the identification process in Example 3 will be described with reference to FIG. 21.

[0608] Step 1:

[0609] Data collection

[0610] The server collects the user's past behavioral history data. Specifically, it stores information about the evacuation shelters the user visited during evacuation drills and their past behavioral patterns in a database. The input is the user's behavioral history data, and the output is the behavioral history data stored in the database. For example, it stores data such as "User A visited evacuation shelter B on March 15, 2023."

[0611] Step 2:

[0612] Collecting Emotional Data

[0613] The server also collects user emotional data, including information on whether the user has been prone to panic in the past. The input is the user emotional data, and the output is the emotional data stored in the database. For example, the server stores data such as "User B has been prone to panic in the past."

[0614] Step 3:

[0615] Data Preprocessing

[0616] The server preprocesses the collected data. Specifically, it complements missing values ​​and normalizes the data. The input is the collected raw data, and the output is the preprocessed data. For example, it estimates and complements missing dates and times of visits to evacuation shelters.

[0617] Step 4:

[0618] Training an AI model

[0619] The server uses the preprocessed data to train an AI model. The frameworks used are TensorFlow and PyTorch. The input is the preprocessed data, and the output is the trained AI model. For example, a model that learns user behavior patterns is built.

[0620] Step 5:

[0621] Generate action instructions

[0622] The server uses the trained AI model to generate response actions based on the user's behavioral history. The input is the trained AI model and the user's behavioral history data, and the output is the generated action instructions. For example, it may give priority to routes to evacuation shelters visited during evacuation drills.

[0623] Step 6:

[0624] Adjustments based on emotional data

[0625] The server adjusts the generated response actions taking into account the emotional data. The inputs are the generated action instructions and the user's emotional data, and the output is the adjusted action instructions. For example, it provides concise and clear evacuation instructions to a user who is prone to panic.

[0626] Step 7:

[0627] Sending action instructions

[0628] The server sends the adjusted action instructions to the terminal. The HTTPS protocol is used for communication. The input is the adjusted action instructions, and the output is the action instructions sent to the terminal. For example, it sends a message saying, "Give priority to the route to shelter B."

[0629] Step 8:

[0630] Receiving action instructions

[0631] The terminal receives the action instructions sent from the server. A device such as a smartphone or tablet is used for reception. The input is the action instructions sent from the server, and the output is the action instructions displayed on the terminal.

[0632] Step 9:

[0633] Display of action instructions

[0634] The terminal displays the received action instructions to the user. The display uses the user interface of the application. The input is the received action instructions, and the output is the action instructions displayed to the user.

[0635] Step 10:

[0636] Obtaining location information

[0637] The terminal obtains the user's current location information. The GPS function is used to obtain the location information. The input is the user's current location information, and the output is the location information sent to the server.

[0638] Step 11:

[0639] Sending location information

[0640] The terminal transmits the acquired location information to the server. The input is the acquired location information, and the output is the location information transmitted to the server.

[0641] Step 12:

[0642] Updates to instructions

[0643] The terminal receives updated action instructions from the server in real time and notifies the user. The input is the updated action instructions from the server, and the output is the updated action instructions notified to the user.

[0644] Step 13:

[0645] Confirmation of action instructions

[0646] The user confirms the action instructions displayed on the terminal. The input is the action instructions displayed on the terminal, and the output is the user's confirmation action.

[0647] Step 14:

[0648] Taking action

[0649] During evacuation drills and actual evacuations, users act based on instructions from the terminal. The input is the instruction to act from the terminal, and the output is the user's evacuation behavior.

[0650] Step 15:

[0651] Sending location information

[0652] The user sends his / her location information to the server through the terminal. The input is the user's location information, and the output is the location information sent to the server.

[0653] Step 16:

[0654] Check for updated action instructions

[0655] The user checks the updated action instructions from the terminal and modifies the action if necessary. The input is the updated action instructions, and the output is the user's modified action.

[0656] (Application example 3)

[0657] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0658] Conventional emergency evacuation systems provide uniform evacuation instructions without considering the user's past behavioral history or emotional data, making it difficult to provide optimal evacuation routes and action instructions for individual users. Furthermore, for users who tend to panic, specific and concise instructions are lacking, making it difficult for them to take appropriate action in an emergency. To solve these issues, a system is needed that can learn the user's past behavioral history and emotional data and provide optimal evacuation routes and action instructions for each individual user.

[0659] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0660] In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating appropriate actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for learning the user's past behavioral history and emotional data and generating appropriate actions based on the learning results, and means for providing optimal evacuation routes and action instructions in an emergency. This makes it possible to provide individually optimal evacuation routes and action instructions that take into account the user's past behavioral history and emotional data.

[0661] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0662] "Type of disaster" refers to the classification of different disasters such as earthquakes, tsunamis, fires, and floods.

[0663] "AI means" refers to means for analyzing data using artificial intelligence technology and generating appropriate instructions for action.

[0664] "Mobile devices" refer to electronic devices that users carry and use, such as smartphones and tablets.

[0665] "Past behavioral history" refers to a record of the actions and choices a user has made in the past.

[0666] "Emotion data" refers to data relating to the user's emotional state, including, for example, panicked or calm.

[0667] An "evacuation route" refers to a route that a user can use to evacuate to a safe place in the event of an emergency.

[0668] "Action instructions" are instructions that indicate specific actions that a user should take in an emergency.

[0669] As an embodiment of the present invention, the following system can be constructed. This system is configured using a user's mobile terminal, a server, and an AI model.

[0670] First, the server has a means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The server also includes a means for analyzing the content of the received emergency alert emails and determining the type of disaster. This allows the server to generate actions to be taken based on the type of disaster.

[0671] Next, the server is equipped with an AI means that learns the user's past behavioral history and emotional data. This AI means generates optimal evacuation routes and instructions for action in an emergency based on the user's past behavioral history and emotional data. For example, it can give priority to routes to evacuation shelters that the user has previously visited during evacuation drills. It can also generate specific and concise instructions for action for users who tend to panic.

[0672] The generated action instructions are displayed on the user's mobile device, which is an electronic device carried by the user, such as a smartphone or tablet, allowing the user to take appropriate action in an emergency.

[0673] To realize this system, software such as Python and Scikit-learn is used. Python is a programming language widely used for data analysis and machine learning, and Scikit-learn is a library for machine learning. Using these software, an AI model is built that learns the user's past behavioral history and emotional data and generates optimal behavioral instructions.

[0674] For example, if a user has previously visited "Shelter A" during an evacuation drill, the route to "Shelter A" will be displayed first in the event of an emergency. Also, if the user has a tendency to panic in the past, specific and concise instructions such as "Remain calm and head to the nearest evacuation center" will be displayed in the event of an emergency.

[0675] An example of a prompt sentence might be:

[0676] "Develop an application that learns from the user's past behavioral history and emotional data and provides optimal evacuation routes and instructions in the event of an emergency. Please include a function that gives priority to routes to evacuation shelters that the user has previously visited in evacuation drills, and generates specific and concise instructions for users who tend to panic."

[0677] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0678] Step 1:

[0679] The server receives emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc. The input is the emergency alert email, and the output is the data of the received emergency alert email. This data includes information such as the type of disaster, the location of the occurrence, and the time of occurrence.

[0680] Step 2:

[0681] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the data from the emergency alert email, and the output is the type of disaster (earthquake, tsunami, fire, etc.). Specifically, it analyzes the text of the email and identifies the type of disaster using keywords and pattern matching.

[0682] Step 3:

[0683] The server generates the actions to be taken based on the type of disaster. The input is the type of disaster, and the output is instructions for the actions to be taken. Specific operations refer to predefined action instructions according to the type of disaster and generate appropriate instructions.

[0684] Step 4:

[0685] The server obtains the user's past behavioral history and emotional data. The input is the user ID, and the output is the past behavioral history and emotional data. Specifically, it queries and obtains the user's behavioral history and emotional data from the database.

[0686] Step 5:

[0687] The server trains an AI model using the acquired past behavioral history and emotional data. The inputs are past behavioral history and emotional data, and the output is a trained AI model. Specifically, it uses a machine learning library such as Scikit-learn to fit the data to the model.

[0688] Step 6:

[0689] The server uses a trained AI model to generate optimal evacuation routes and action instructions in the event of an emergency. The inputs are current situation data (location information, type of disaster, etc.) and the trained AI model, and the output is optimal evacuation routes and action instructions. Specifically, the current situation data is input into the AI ​​model, and action instructions are generated based on the prediction results.

[0690] Step 7:

[0691] The server sends the generated action instructions to the user's mobile device. The inputs are the action instructions and the user's mobile device information, and the output is the action instructions displayed on the user's mobile device. Specifically, the server formats the action instructions as a message and sends a push notification to the mobile device.

[0692] Step 8:

[0693] The user's mobile device displays the received action instructions. The input is the action instructions sent from the server, and the output is the action instructions that the user can check on the screen. The specific operation is to analyze the received message and display it on the screen.

[0694] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0695] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0696] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[0697] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0698] [Second embodiment]

[0699] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0700] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0701] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. 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. The database 24 and the communication I / F 26 are also 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).

[0702] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0703] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0704] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0705] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0706] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0707] The specific processing program 56 is an example of a "program" according to the technology of the present 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 in accordance with the specific processing program 56 executed on the RAM 30.

[0708] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0709] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0710] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0711] "Example 1"

[0712] One embodiment of the present invention is a system using a smartphone. This system has the function of receiving emergency alert emails, such as earthquake, tsunami, and disaster evacuation information. The content of the received emergency alert email is analyzed by an AI inside the system to determine the type of disaster. The AI ​​generates actions to be taken based on the determined type of disaster and displays the action instructions on the smartphone screen.

[0713] "Example 2"

[0714] Furthermore, another embodiment of the present invention is a system that takes into account the user's location information. In this system, AI acquires the user's location information and generates an action to be taken based on that location information. For example, if the user is in an area where a tsunami warning has been issued, the AI ​​generates an action such as "evacuate to the nearest high ground."

[0715] "Example 3"

[0716] Another embodiment of the present invention is a system that learns the user's past behavioral history. In this system, AI learns the user's past behavioral history and generates appropriate actions based on the learning results. For example, it is possible to prioritize routes to evacuation shelters that the user has visited in the past during evacuation drills.

[0717] The processing flow of each embodiment will be described below.

[0718] "Example 1"

[0719] Step 1: Your smartphone will receive emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc.

[0720] Step 2: The contents of the received emergency alert email are analyzed by the AI ​​within the system.

[0721] Step 3: The AI ​​determines the type of disaster and generates appropriate actions based on that type.

[0722] Step 4: Display the generated action instructions on the smartphone screen.

[0723] "Example 2"

[0724] Step 1: AI obtains the user's location information.

[0725] Step 2: Based on the acquired location information, the AI ​​generates the action to be taken.

[0726] Step 3: Display the generated action instructions on the smartphone screen.

[0727] "Example 3"

[0728] Step 1: The AI ​​learns the user's past behavioral history.

[0729] Step 2: Based on the learning results, the AI ​​generates the action to be taken.

[0730] Step 3: Display the generated action instructions on the smartphone screen.

[0731] Example 1

[0732] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0733] Conventional emergency alert systems required users to analyze the content of the received emergency alert email and determine the appropriate course of action, making it difficult to respond quickly and accurately. Furthermore, because they were unable to respond individually to users' location information or past behavioral history, they could only provide uniform instructions to all users. This made it difficult for users to take appropriate evacuation actions.

[0734] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0735] In this invention, the server includes: means for receiving emergency alert emails for earthquakes, tsunamis, disaster evacuation information, etc.; means for analyzing the content of the received emergency alert emails and determining the type of disaster; artificial intelligence means for generating response actions based on the determined type of disaster; means for displaying the generated actions on the user's mobile information terminal; means for the artificial intelligence means to analyze the content of the emergency alert emails using natural language processing technology; and means for the artificial intelligence means to display the generated action instructions on the screen of the mobile information terminal. This eliminates the need for the user to analyze the content of the received emergency alert emails themselves, allowing them to receive quick and accurate action instructions. Furthermore, it is possible to respond individually to evacuation actions taking into account the user's location information and past behavioral history, allowing for more appropriate evacuation actions.

[0736] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0737] "Mobile information terminal" refers to a portable information processing device such as a smartphone or tablet.

[0738] "Artificial intelligence means" refers to a system that uses technologies such as machine learning and deep learning to analyze data and make judgments and predictions.

[0739] "Natural language processing technology" is a technology that allows computers to understand, interpret, and generate human language.

[0740] "Action instructions" are instructions that indicate specific actions that a user should take in a particular situation.

[0741] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[0742] "Past behavior history" is a record of actions taken by a user in the past, and is used to predict future actions based on this.

[0743] This invention is a system that receives emergency emails about earthquakes, tsunamis, disaster evacuation information, etc., analyzes the contents of the emails, and generates appropriate instructions for actions. A specific embodiment of this system will be described below.

[0744] First, let's assume that a user has a mobile information terminal (smartphone). The terminal has the function to receive emergency alert emails. Emergency alert emails are sent by official organizations such as the government or the Japan Meteorological Agency. The terminal receives these emails through a built-in communication module (e.g., a 4G / 5G modem).

[0745] Next, the server analyzes the content of the received emergency alert email. Specifically, it extracts the text data from the email and analyzes it using natural language processing (NLP) technology. The generative AI model used here is, for example, BERT or GPT-3. The server obtains the content of the email in text format and inputs it into the generative AI model. For example, it analyzes the text, "A large earthquake has occurred. Please evacuate to a safe place immediately."

[0746] The server uses a generative AI model to determine the type of disaster from the email content. The model analyzes keywords and context within the text to identify disasters such as earthquakes, tsunamis, and fires. For example, if the text "A large earthquake has occurred" is input, the model will determine this as an "earthquake."

[0747] Next, the server generates actions that the user should take based on the type of disaster determined. The generative AI model generates appropriate action instructions according to the type of disaster. For example, in the case of an earthquake, it generates specific action instructions such as "hide under a desk," and in the case of a tsunami, it generates specific action instructions such as "evacuate to higher ground." The server inputs a prompt statement to the generative AI model saying, "Please generate action instructions in the case of an earthquake," and the model generates the action instruction, "Hide under a desk or evacuate outside the building."

[0748] Finally, the terminal displays the generated action instructions on the smartphone screen. The user checks the action instructions displayed on the screen and takes appropriate action. For example, a message such as "Hide under a desk or evacuate outside the building" is displayed on the smartphone screen.

[0749] As a concrete example, consider the case where a user receives an emergency alert email on their smartphone. The email message reads, "A large earthquake has occurred. Please evacuate to a safe location immediately." In this case, the server inputs the following prompt sentence into the generative AI model:

[0750] Prompt text: "An emergency alert email has been received stating, 'A large earthquake has occurred. Please evacuate to a safe place immediately.'" Analyze the contents of this email and generate the action the user should take.

[0751] Based on this prompt, the generative AI model generates the following action instructions:

[0752] Action Instructions: "Hide under a desk or leave the building."

[0753] This instruction for action is displayed on the smartphone screen, prompting the user to take appropriate action.

[0754] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0755] Step 1: Receiving emergency alert emails

[0756] The device receives emergency alert emails. Specifically, it receives emergency alert emails sent by official organizations such as the government and the Japan Meteorological Agency via the device's communications module (4G / 5G modem).

[0757] Input: Emergency alert email sent by official organization

[0758] Output: Text data of the received emergency alert email

[0759] Specific operation: If a user has a smartphone, the device will automatically receive an emergency alert email. For example, if an earthquake alert is sent, the device will receive the email.

[0760] Step 2: Analyzing the email content

[0761] The server analyzes the content of the received emergency alert email. Specifically, it extracts the text data from the email and analyzes it using natural language processing technology. The generative AI models used here are BERT and GPT-3.

[0762] Input: Text data of received emergency alert email

[0763] Output: Parsed email content text data

[0764] Specific operation: The server retrieves the contents of the email in text format and inputs it into the generative AI model. For example, it analyzes the text, "A large earthquake has occurred. Please evacuate to a safe place immediately."

[0765] Step 3: Determine the type of disaster

[0766] The server uses a generative AI model to determine the type of disaster from the email content. The model analyzes keywords and context within the text to identify disasters such as earthquakes, tsunamis, and fires.

[0767] Input: Parsed email content text data

[0768] Output: Type of disaster determined

[0769] Specific operation: The server inputs the text "A large earthquake has occurred" into the generative AI model, and the model analyzes this and determines that it is an "earthquake."

[0770] Step 4: Generate action instructions

[0771] The server generates the actions that the user should take based on the determined type of disaster, and the generative AI model generates appropriate action instructions according to the type of disaster.

[0772] Input: Type of disaster determined

[0773] Output: Generated action instructions

[0774] Specific operation: The server inputs a prompt statement to the generative AI model saying, "Generate action instructions in the event of an earthquake." The model then generates the action instruction, "Hide under a desk or evacuate outside the building."

[0775] Step 5: Displaying Action Instructions

[0776] The terminal displays the generated action instructions on the smartphone screen, and the user checks the action instructions displayed on the screen and takes appropriate action.

[0777] Input: Generated action instructions

[0778] Output: Instructions displayed on the smartphone screen

[0779] Specific operation: The device displays the action instructions received from the server on the screen. For example, a message such as "Hide under a desk or evacuate outside the building" will be displayed on the smartphone screen.

[0780] (Application example 1)

[0781] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0782] Conventional disaster response systems only received emergency alert emails and were unable to provide users with specific response actions. It was also difficult to generate appropriate action instructions in real time depending on the type of disaster, making it difficult for users to act quickly and appropriately. Furthermore, because they were unable to respond individually based on the user's location information or past behavioral history, they could only provide generic instructions.

[0783] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0784] In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating appropriate actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile information terminal, means for determining the type of disaster using a generative AI model and generating appropriate actions, and means for generating action instructions by inputting a prompt sentence into the generative AI model. This allows the user, upon receiving the emergency alert email, to quickly learn specific appropriate actions based on the type of disaster analyzed by the AI, enabling individual responses that take into account location information and past behavioral history.

[0785] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0786] "Analysis" is the process of examining the contents of the received emergency alert email in detail to determine the type of disaster and its impact.

[0787] "Type of disaster" refers to the specific type of disaster that occurred, such as earthquake, tsunami, fire, or flood.

[0788] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[0789] "AI means" refers to means that use artificial intelligence technology to analyze data, determine the type of disaster, and generate appropriate actions to respond.

[0790] The term "mobile information terminal" refers to an information processing device that is carried and used by a user, such as a smartphone or tablet.

[0791] A "generative AI model" is a generative artificial intelligence model, an AI model that has the ability to generate new data and information based on input data.

[0792] A "prompt" is text data input to a generative AI model that contains instructions or questions to obtain a specific output.

[0793] A system for implementing this invention is configured as follows: A server has means for receiving emergency alert emails, such as earthquake, tsunami, and disaster evacuation information. The content of the received emergency alert emails is analyzed in detail by analysis means within the server, and the type of disaster is determined. The analysis means determines the type of disaster using a generative AI model and generates actions to be taken in response.

[0794] The generated action instructions are displayed on the user's mobile information device, which is an information processing device such as a smartphone or tablet that the user always carries with them. The generative AI model operates by inputting prompt sentences, which contain instructions or questions to obtain a specific output.

[0795] As a specific example, suppose the server receives an emergency alert email with the content "Emergency Alert: An earthquake has occurred. The epicenter is in Tokyo." The analysis means analyzes the content of this email and inputs the following prompt sentence into the generative AI model.

[0796] Please analyze the contents of the emergency alert email below and determine the type of disaster.

[0797] Breaking News: An earthquake has occurred. The epicenter is in Tokyo.

[0798] Based on this prompt, the generative AI model determines that it is an "earthquake" and then inputs the following prompt to generate the action to be taken:

[0799] If the disaster type is an earthquake, generate the actions to be taken.

[0800] The generative AI model generates specific instructions, such as "hide under the desk," which are displayed on the user's mobile device. The user can then act quickly by following the instructions displayed on the mobile device.

[0801] The system is implemented using Python and OpenAI's API. Python is a programming language well-suited for data analysis and AI model operation, while OpenAI's API provides an interface for operating the generative AI model. This allows users to quickly learn specific response actions based on the type of disaster analyzed by the AI ​​when they receive an emergency alert email.

[0802] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0803] Step 1:

[0804] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the content of the emergency alert email, and the output is the received email data. Specifically, the server retrieves the emergency alert email from the mail server and stores its content in memory.

[0805] Step 2:

[0806] The server analyzes the content of the received emergency alert email. The input is the received email data, and the output is the analyzed disaster type. Specifically, the server inputs the following prompt sentence into the generative AI model:

[0807] Please analyze the contents of the emergency alert email below and determine the type of disaster.

[0808] Breaking News: An earthquake has occurred. The epicenter is in Tokyo.

[0809] Based on this prompt, the generative AI model determines the type of disaster to be "earthquake."

[0810] Step 3:

[0811] The server generates the action to be taken based on the type of disaster determined. The input is the type of disaster, and the output is the action to be taken. Specifically, the server inputs the following prompt sentence into the generation AI model:

[0812] If the disaster type is an earthquake, generate the actions to be taken.

[0813] Based on this prompt, the generative AI model generates specific action instructions, such as "hide under the desk."

[0814] Step 4:

[0815] The server sends the generated action instructions to the user's mobile information device. The input is the action instructions to be handled, and the output is the action instructions to be displayed on the user's mobile information device. Specifically, the server formats the action instructions as a text message and sends it to the user's mobile information device.

[0816] Step 5:

[0817] The user acts according to the instructions displayed on the mobile information terminal. The input is the instructions displayed on the mobile information terminal, and the output is the user's specific actions. As a specific action, the user follows instructions such as "hide under a desk" to evacuate to a safe place.

[0818] Example 2

[0819] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0820] Conventional disaster evacuation systems provide uniform evacuation information without considering the user's location information, making it difficult to suggest optimal evacuation actions for each individual user. Furthermore, because they do not consider the user's past behavioral history, they are unable to suggest appropriate actions based on the user's characteristics. This has led to the issue of not being able to adequately ensure the safety of users in emergencies.

[0821] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0822] In this invention, the server includes a means for acquiring user location information, a means for transmitting the acquired location information to the server, and a means for the server to analyze the location information received and generate an action to be taken using a generative AI model. This makes it possible to propose appropriate evacuation actions based on the user's current location in real time. In addition, the generative AI model learns the user's past behavioral history and generates an action to be taken based on the learning results, making it possible to propose more appropriate actions according to the user's characteristics.

[0823] An "emergency message" is a message that includes information about an emergency, such as an earthquake, tsunami, or disaster evacuation information.

[0824] "Location information" means data that indicates a user's current geographic location, including latitude and longitude information.

[0825] The "server" is a computer system that receives and analyzes the user's location information and uses a generative AI model to generate an action to be taken.

[0826] A "generative AI model" is an artificial intelligence model that generates appropriate behavior based on a user's location information and past behavioral history.

[0827] A "prompt" is a textual instruction that is input into a generative AI model to generate appropriate behavior based on the user's location and situation.

[0828] "Terminal" refers to a device used by a user, including a smartphone or tablet.

[0829] "Analysis" is the process of interpreting information and deriving meaning from received data.

[0830] "Action suggestions" are specific instructions for actions that a user should take, generated by a generative AI model based on the user's location information and past behavioral history.

[0831] "Notification" is the process of sending the action suggestions generated by the server to the user's terminal to inform the user.

[0832] This invention is a system that suggests appropriate actions based on the user's location information. This system is realized using the user's device, a server, and a generative AI model.

[0833] First, the user's device is a device with GPS functionality, such as a smartphone or tablet. The device uses the GPS sensor to obtain the user's current location information. The obtained location information is expressed as latitude and longitude data.

[0834] The device then sends the acquired location information to the server using the HTTPS protocol to ensure data security. The device converts the location information into JSON format and sends it to the server via an HTTPS request.

[0835] The server analyzes the received location information. A generative AI model, such as a natural language processing model, is used for the analysis. The server identifies the area the user is in based on the location information and obtains risk information related to that area.

[0836] The server then uses the generative AI model to generate suggested actions based on the user's location. Specifically, the server inputs prompts into the generative AI model to generate appropriate actions. Examples of prompts include:

[0837] "The user is currently near the coast. A tsunami warning has been issued. Please suggest what action the user should take."

[0838] Based on this prompt, the generative AI model generates the suggested action, "Evacuate to the nearest high ground."

[0839] Finally, the server notifies the user's device of the generated action suggestions using a push notification service. The server converts the action suggestions into JSON format and sends them to the user's smartphone via the push notification service. The user's smartphone receives the notification and displays it on the screen.

[0840] In this way, the system can take into account the user's location and suggest appropriate actions in real time, allowing users to take prompt and appropriate action even in emergencies.

[0841] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0842] Step 1:

[0843] The user's device uses the GPS sensor to obtain current location information. The input is data from the GPS sensor, and the output is latitude and longitude information. Specifically, when a user enables location services on their smartphone, the device periodically obtains GPS data.

[0844] Step 2:

[0845] The device sends the acquired location information to the server. The input is latitude and longitude information, and the output is data sent to the server as an HTTPS request. Specifically, the device converts the location information into JSON format and sends it to the server using the HTTPS protocol.

[0846] Step 3:

[0847] The server analyzes the received location information. The input is the location information sent from the device, and the output is the result of identifying the area the user is in. Specifically, the server checks the location information against a database to see if the user is in an area where a tsunami warning has been issued.

[0848] Step 4:

[0849] The server uses a generative AI model to generate action suggestions based on the user's location. The input is the location and a prompt, and the output is an action suggestion. Specifically, the server inputs the following prompt into the generative AI model:

[0850] "The user is currently near the coast. A tsunami warning has been issued. Please suggest what action the user should take."

[0851] Based on this prompt, the generative AI model generates the suggested action, "Evacuate to the nearest high ground."

[0852] Step 5:

[0853] The server notifies the user's device of the generated action suggestions. The input is the action suggestions, and the output is data sent to the user's device as a push notification. Specifically, the server converts the action suggestions into JSON format and sends them to the user's smartphone via a push notification service. The user's smartphone receives the notification and displays it on the screen.

[0854] (Application example 2)

[0855] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0856] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information, making it difficult to provide optimal evacuation actions for each individual user. Furthermore, they lack the functionality to present evacuation routes or notify emergency contacts, resulting in insufficient comprehensive support to ensure the user's safety.

[0857] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating actions to be taken based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for acquiring the user's location information, means for generating an optimal evacuation route based on the acquired location information and displaying it on a map, and means for automatically notifying emergency contacts. This makes it possible to provide optimal evacuation actions based on the user's current location, suggest evacuation routes, and promptly notify emergency contacts.

[0858] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0859] "Type of disaster" refers to the classification of different disasters such as earthquakes, tsunamis, floods, and fires.

[0860] "AI means" is a technology that uses artificial intelligence to analyze data and generate optimal actions based on specific conditions.

[0861] "Mobile device" refers to a portable electronic device such as a smartphone or tablet.

[0862] "Location information" is latitude and longitude data that indicates the user's current location.

[0863] An "evacuation route" refers to the optimal route for a user to evacuate to a safe location.

[0864] "Emergency contact information" is information about people to contact in the event of a disaster or emergency.

[0865] A system for carrying out this invention has the following configuration. First, a server is provided with means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The server also includes means for analyzing the content of the received emergency alert emails and determining the type of disaster. Furthermore, the server is equipped with AI means for generating actions to be taken based on the determined type of disaster.

[0866] The user's mobile device has a means for displaying the generated action. The mobile device also has a means for acquiring the user's location information, and includes a means for generating an optimal evacuation route based on the acquired location information and displaying it on a map. The mobile device also has a means for automatically notifying emergency contacts.

[0867] The following hardware and software are used to implement this system: A smartphone or tablet with GPS functionality is required for the hardware. The software uses a Python program, the Requests library (to obtain disaster information from the API), the Geopy library (to process location information), and the Folium library (to display maps).

[0868] When the server receives an emergency alert email, it analyzes its contents and determines the type of disaster. For example, if a tsunami warning is issued, the server generates an action for the user based on that information, such as "Please evacuate to the nearest high ground." The generated action is displayed on the user's mobile device.

[0869] The mobile device acquires the user's location information in real time and generates the optimal evacuation route based on that information. The evacuation route is displayed on a map, and the user can follow it to evacuate. An automatic notification function also quickly shares the user's current location and evacuation actions with emergency contacts.

[0870] For example, if a user is in Tokyo and a tsunami warning is issued, the server will notify the user, "Please evacuate to the nearest high ground," and the mobile device will display evacuation routes on a map. The following prompt sentence is used:

[0871] Example prompt sentence:

[0872] If the user's current location is Tokyo and a tsunami warning has been issued, generate an evacuation action to the nearest higher ground and display the evacuation route on a map.

[0873] In this way, a system can be realized that provides specific actions to ensure the safety of users.

[0874] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0875] Step 1:

[0876] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the emergency alert email, and the output is the content of the received email. Specifically, the server obtains emails from an external emergency alert system.

[0877] Step 2:

[0878] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the content of the received email, and the output is the type of disaster (e.g., earthquake, tsunami). Specifically, the server analyzes the text of the email, extracts keywords, and identifies the type of disaster.

[0879] Step 3:

[0880] The server generates the appropriate action to be taken based on the type of disaster it has determined. The input is the type of disaster, and the output is the appropriate action to be taken (e.g., "Please evacuate to the nearest high ground"). Specifically, the server uses a generative AI model to generate the optimal action according to the type of disaster.

[0881] Step 4:

[0882] The server sends the generated action to the user's mobile device. The input is the action to be taken, and the output is an action message to be displayed on the user's mobile device. Specifically, the server pushes the action message to the mobile device.

[0883] Step 5:

[0884] The device obtains the user's location information in real time. The input is GPS data, and the output is the user's current location (latitude and longitude). Specifically, the device obtains location information using its built-in GPS function.

[0885] Step 6:

[0886] The device generates the optimal evacuation route based on the acquired location information and displays it on a map. The input is the user's current location and evacuation location information, and the output is the evacuation route displayed on the map. Specifically, the device uses a map display library (Folium) to draw the evacuation route.

[0887] Step 7:

[0888] The device automatically notifies emergency contacts. The input is the user's current location and the generated action, and the output is a notification message sent to the emergency contacts. Specifically, the device sends the notification message to the emergency contacts via SMS or email.

[0889] In this way, a system can be realized that provides specific actions to ensure the safety of users.

[0890] Example 3

[0891] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0892] The main function of conventional disaster evacuation systems is to receive emergency alert emails and generate response actions based on their content analysis, but they lack the ability to respond individually to users' past behavioral history and location information. This means that optimal evacuation routes and instructions for each user are not provided, which can reduce the efficiency and safety of evacuation. Furthermore, there is a lack of a mechanism for using user feedback as learning data for the next system, making it difficult to improve the accuracy of the system.

[0893] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0894] In this invention, the server includes means for collecting past behavioral history data of users and storing it in a database, means for preprocessing the stored behavioral history data and training it using a generative AI model, and means for predicting the user's future behavior based on the training results and generating appropriate behavior. This makes it possible to provide optimal evacuation routes and behavioral instructions for each user, improving the efficiency and safety of evacuation. Furthermore, by using user feedback as training data for the next round, the accuracy of the system can be continuously improved.

[0895] An "emergency alert email" is an email sent to quickly notify users of emergency information such as earthquakes, tsunamis, and disaster evacuation information.

[0896] "Analysis" is the process of analyzing the contents of the received emergency alert email and determining the type of disaster and its impact.

[0897] "Artificial intelligence means" refers to machine learning algorithms and models that generate optimal response actions based on a user's behavioral history and location information.

[0898] A "mobile terminal" is an electronic device that a user carries and uses, such as a smartphone or tablet.

[0899] "Behavioral history data" refers to data about the actions a user has taken and the places they have visited in the past.

[0900] A "database" is a system for storing and managing collected behavioral history data.

[0901] "Preprocessing" refers to processes such as filling in missing values, normalizing data, and removing outliers that are carried out to improve the quality of collected data.

[0902] A "generative AI model" is a model trained using a machine learning framework and used to predict user behavior and generate appropriate countermeasures.

[0903] "Learning results" refer to user behavior patterns and predictive models obtained as a result of the generative AI model learning based on behavioral history data.

[0904] "Feedback" is the user's evaluation or opinion of the system's instructions or generated actions.

[0905] The present invention is a system that learns the user's past behavioral history and generates an action to be taken based on the learning results. Specific embodiments of this system will be described below.

[0906] Server Processing

[0907] The server collects the user's past behavioral history data and stores it in a database. Specifically, it collects GPS data, timestamps, and the type of behavior (e.g., evacuation drills, commuting, etc.) from the user's mobile device. The collected data is stored in a relational database such as MySQL or PostgreSQL.

[0908] The server then preprocesses the stored data, which includes imputing missing values, normalizing the data, and removing outliers, improving the quality of the data and increasing the accuracy of the generative AI model.

[0909] Using the preprocessed data, the server trains a generative AI model. This process uses machine learning frameworks such as TensorFlow and PyTorch. Recurrent neural networks (RNNs) and long short-term memory (LSTMs) are used as learning algorithms. Once training is complete, the server saves the learning results. Saved models are managed in formats such as HDF5 and Pickle.

[0910] Based on the learning results, the server predicts the user's future behavior and generates appropriate actions. For example, if the user is conducting an evacuation drill, the server generates the optimal evacuation route based on past data. This information is sent to the user's mobile device.

[0911] Terminal handling

[0912] The device receives the learning results and prediction data sent from the server. The received data is used as action instructions for the user. Specifically, evacuation routes and action instructions are displayed on a map. Map services such as Google Maps API are used to display the information in a visually easy-to-understand format.

[0913] The device accepts input from the user. For example, if the user instructs the device to "display evacuation routes," the device will display the appropriate information in accordance with the instruction.

[0914] User Action

[0915] The user checks the action instructions displayed on the device. For example, during an evacuation drill, the user checks the evacuation route displayed on the device. The user acts based on the information provided by the device. Specifically, the user follows the displayed evacuation route and heads to a shelter.

[0916] After taking action, the user provides feedback. For example, they can input whether the evacuation route was appropriate or not. This feedback is sent to the server and used as learning data for the next time.

[0917] Specific examples

[0918] For example, consider a case where a user preferentially instructs a route to a shelter that the user has visited in the past during evacuation drills. The server analyzes the user's past evacuation drill data and generates the optimal evacuation route. This information is sent to the user's device and is useful when the user evacuates.

[0919] Prompt Sentence Examples

[0920] "Generate optimal evacuation routes based on the user's past behavioral history. Use past evacuation drill data to prioritize routes to evacuation shelters."

[0921] By inputting this prompt sentence into the generative AI model, an optimal evacuation route is generated based on the user's behavior history. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0922] Step 1:

[0923] Data collection

[0924] The server collects GPS data, timestamps, and types of activity (e.g., evacuation drills, commuting, etc.) from the user's mobile device. The user's location information and activity history data are provided as input. The server receives this data and stores it in a database. Specifically, it periodically retrieves data from the mobile device and stores it in a relational database such as MySQL or PostgreSQL.

[0925] Step 2:

[0926] Data Preprocessing

[0927] The server preprocesses the stored behavioral history data. Raw data stored in the database is provided as input. Preprocessing includes missing value completion, data normalization, and outlier removal. This improves the quality of the data and increases the accuracy of the generative AI model. Specifically, the server uses Python's Pandas library to clean the data and perform the necessary preprocessing.

[0928] Step 3:

[0929] Model learning

[0930] The server trains a generative AI model using the preprocessed data. The preprocessed data is provided as input. Machine learning frameworks such as TensorFlow and PyTorch are used for training. Recurrent neural networks (RNNs) and long short-term memory (LSTMs) are applied as training algorithms. A trained model is obtained as output. Specifically, the data is fed into the model and the model parameters are updated every epoch.

[0931] Step 4:

[0932] Save Model

[0933] The server saves the model after training is complete. The trained model is provided as input. The model is saved in formats such as HDF5 or Pickle. The output is the saved model file. Specifically, the model is written to a file in the specified format and saved in storage.

[0934] Step 5:

[0935] Prediction and Action Generation

[0936] The server uses the stored model to predict the user's future behavior and generate appropriate actions. The input is the trained model and the user's current location information. The output is the generated action instructions. Specifically, the current data is input into the model, and the optimal action is generated based on the prediction results.

[0937] Step 6:

[0938] Data reception

[0939] The device receives the learning results and prediction data sent from the server. As input, action instruction data is provided from the server. As output, the received data is saved on the device. Specifically, the device receives data from the server and saves it in local storage.

[0940] Step 7:

[0941] Data Display

[0942] The device displays the received data to the user. The received action instruction data is provided as input. The output is information that is visually displayed to the user. Specifically, the device uses the Google Maps API to display evacuation routes on a map.

[0943] Step 8:

[0944] User Interaction

[0945] The user checks the action instructions displayed on the device and provides feedback as necessary. The information displayed on the device is provided as input. The user's feedback is obtained as output. In concrete terms, the user checks the evacuation route and inputs feedback into the device.

[0946] Step 9:

[0947] Gathering feedback

[0948] The server collects feedback from users and uses it as training data for the next model. User feedback data is provided as input. Updated training data is obtained as output. Specifically, the feedback data is saved in a database and used for the next model training.

[0949] (Application example 3)

[0950] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0951] Conventional disaster evacuation systems only provide general evacuation routes without taking into account the user's past behavioral history, making it difficult to provide the optimal evacuation route for the user. Furthermore, while quick and accurate evacuation instructions are required in emergencies, conventional systems were inadequate in this regard. This meant that effective evacuation support to ensure the safety of users could not be achieved.

[0952] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0953] In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating actions to be taken based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, and navigation means for learning the user's past behavioral history and providing the optimal evacuation route in an emergency based on the learning results. This makes it possible to provide the optimal evacuation route taking the user's past behavioral history into consideration, and to issue quick and accurate evacuation instructions in an emergency.

[0954] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[0955] "Analysis" is a process of examining the details of the received emergency alert email to determine the type and situation of the disaster.

[0956] "Type of disaster" refers to the specific classification of the disaster that occurred, such as earthquake, tsunami, fire, or flood.

[0957] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[0958] "AI means" is a system that uses artificial intelligence technology to analyze received information and generate appropriate response actions.

[0959] "Mobile devices" refer to electronic devices that users carry and use, such as smartphones and tablets.

[0960] "Display" refers to visually showing the generated response action on the screen of the user's mobile device.

[0961] "Past behavioral history" refers to records of the actions a user has taken and the places they have visited in the past.

[0962] "Learning" refers to data processing to predict future behavior based on past behavioral history.

[0963] "Navigation means" is a system that guides users to the optimal route.

[0964] "Evacuation route" refers to the route that users can take to safely evacuate in the event of an emergency.

[0965] The following system configuration will be described as an embodiment of the present invention.

[0966] System Configuration

[0967] The system consists of the following major components:

[0968] 1. Server: Receives emergency alert emails, analyzes them, and determines the type of disaster.

[0969] 2. Mobile device: The user's smartphone or tablet receives instructions from the server.

[0970] 3. AI means: Installed on the server, it generates the actions to be taken based on the received information.

[0971] 4. Navigation: Learns the user's past behavior history and provides the optimal evacuation route.

[0972] Program processing explanation

[0973] The server uses an Internet connection to receive emergency alert emails. It analyzes the content of the emails and uses natural language processing technology to determine the type of disaster. Specifically, it uses a Python natural language processing library (e.g., NLTK or spaCy).

[0974] The AI ​​solution takes into account the user's location and past behavioral history to generate optimal countermeasures using machine learning models (e.g., TensorFlow and PyTorch). The generated countermeasures are displayed on the mobile device.

[0975] The navigation system learns the user's past behavior history and provides the optimal evacuation route in case of an emergency. This is done by linking with a geographic information system (GIS) and comparing the user's current location with past evacuation routes. Specifically, it uses the Geopy library for distance calculations.

[0976] Specific examples

[0977] For example, if a user previously visited a location called "shelter A" during an evacuation drill and their current location is "point B," the system will provide the optimal evacuation route from "point B" to "shelter A."

[0978] Prompt Sentence Examples

[0979] Develop a navigation system that learns from the user's past behavioral history and provides the optimal evacuation route in an emergency. Include a function that prioritizes routes to places the user has previously visited during evacuation drills.

[0980] In this way, users can receive prompt and accurate evacuation instructions in an emergency, which will enable effective evacuation support to ensure the safety of users.

[0981] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0982] Step 1:

[0983] The server receives emergency alert emails via an Internet connection.

[0984] Input: Emergency Alert Email

[0985] Output: Contents of the received emergency alert email

[0986] Specific operation: The server retrieves emergency alert emails from the mail server and saves their contents in text format.

[0987] Step 2:

[0988] The server analyzes the contents of the received emergency alert email and determines the type of disaster.

[0989] Input: Contents of the emergency alert email received

[0990] Output: Disaster type

[0991] What it does: The server uses natural language processing libraries (e.g., NLTK or spaCy) to analyze the content of the email and identify the type of disaster, such as earthquake, tsunami, or fire, based on keywords and context.

[0992] Step 3:

[0993] The server generates an action to be taken based on the determined type of disaster.

[0994] Input: Disaster Type

[0995] Output: Action to be taken

[0996] Specific operation: The server uses machine learning models (e.g., TensorFlow or PyTorch) to generate optimal response actions depending on the type of disaster. For example, in the case of an earthquake, it generates actions such as "evacuate the building," and in the case of a tsunami, it generates actions such as "evacuate to higher ground."

[0997] Step 4:

[0998] The server transmits the generated action to be taken to the user's mobile terminal.

[0999] Input: Action to be taken

[1000] Output: Actions displayed on mobile device

[1001] Specific operation: The server sends the generated response action to the user's mobile device as a push notification and displays it on the device screen.

[1002] Step 5:

[1003] The server learns the user's past behavioral history and provides the optimal evacuation route based on the results of that learning.

[1004] Input: User's past behavior history, current location

[1005] Output: Optimal evacuation route

[1006] Specific operation: The server retrieves the user's behavior history from the past behavior history database and learns it using a machine learning model. The current location is obtained using GPS data, and the Geopy library is used to compare the current location with past evacuation routes and calculate the optimal evacuation route.

[1007] Step 6:

[1008] The server sends the optimal evacuation route to the user's mobile device.

[1009] Input: Optimal evacuation route

[1010] Output: Evacuation route displayed on mobile device

[1011] Specific operation: The server sends the calculated optimal evacuation route to the user's mobile device and displays it through the navigation app. The user can then evacuate safely by following the displayed route.

[1012] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1013] "Example 1"

[1014] In one embodiment of the present invention, the AI ​​means includes an emotion engine that recognizes the user's emotions by inferring them from the user's facial expressions, tone of voice, text input, etc. For example, if the user is scared, the AI ​​uses that information to generate calmer behavioral instructions.

[1015] "Example 2"

[1016] In another embodiment of the invention, the AI ​​means takes into account the user's location and emotions to generate a response action. For example, if the user is in a public place and feels scared, the AI ​​can use that information to instruct them to evacuate to a safer, less crowded place.

[1017] "Example 3"

[1018] In yet another embodiment of the present invention, the AI ​​means learns from the user's past behavioral history and emotions and generates countermeasures based on the learning results. For example, if the user has a tendency to panic in the past, the AI ​​will use that information to generate more specific and concise instructions.

[1019] The processing flow of each embodiment will be described below.

[1020] "Example 1"

[1021] Step 1: Step 1: The AI ​​means uses an emotion engine to estimate emotions from the user's facial expressions, tone of voice, text input, etc.

[1022] Step 2: The AI ​​means generates behavioral instructions for the user taking into account the estimated emotion. For example, if the user is scared, the AI ​​uses that information to generate gentler behavioral instructions.

[1023] "Example 2"

[1024] Step 1: Step 1: The AI ​​means estimates the user's emotions using the user's location information and the emotion engine.

[1025] Step 2: The AI ​​generates instructions for the user based on the estimated emotion and location. For example, if the user is in a public place and feels scared, the AI ​​can use that information to instruct them to evacuate to a safer location with fewer people.

[1026] "Example 3"

[1027] Step 1: Step 1: The AI ​​means estimates the user's emotions using the user's past behavior history and the emotion engine.

[1028] Step 2: The AI ​​generates instructions for the user based on the estimated emotions and their past behavioral history. For example, if the user has a history of panic, the AI ​​uses that information to generate more specific and concise instructions.

[1029] Example 1

[1030] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1031] Conventional emergency alert systems can receive disaster information and provide users with action instructions, but they are unable to generate action instructions that take into account the user's emotional state. This means that if a user is feeling fear or anxiety, it becomes difficult for them to take appropriate action. Another issue is that they cannot provide individualized responses that take into account the user's location information or past behavioral history, so they can only provide generic instructions.

[1032] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1033] In this invention, the server includes a means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information; a means for analyzing the content of the received emergency alert email and determining the type of disaster; an artificial intelligence means for generating appropriate action to be taken based on the determined type of disaster; a means for displaying the generated action on the user's mobile information terminal; a means for recognizing the user's emotions and modifying action instructions based on the emotions; and a means for displaying the modified action instructions on the user's mobile information terminal. This makes it possible to provide appropriate action instructions that take into account the user's emotional state, allowing the user to take more appropriate action even when feeling fear or anxiety. It also makes it possible to provide individualized responses that take into account the user's location information and past behavioral history.

[1034] An "emergency alert email" is an email sent to quickly notify users of emergency information such as earthquakes, tsunamis, and disaster evacuation information.

[1035] "Mobile information terminal" refers to a portable information processing device such as a smartphone or tablet.

[1036] "Artificial intelligence means" is a system that includes algorithms and programs for analyzing received information and generating appropriate instructions for action.

[1037] "Means for recognizing emotions" refers to technology for inferring emotions from a user's facial expressions, tone of voice, text input, etc.

[1038] "Action instructions" are instructions that indicate specific actions that the user should take depending on the type of disaster.

[1039] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[1040] "Past behavior history" is a record of the user's past behaviors, and is data used to generate future behavioral instructions based on this record.

[1041] This invention is a system that receives emergency emails about earthquakes, tsunamis, disaster evacuation information, etc., and provides users with appropriate instructions on what to do. This system is mainly composed of three elements: a server, a terminal (mobile information terminal), and a user.

[1042] Hardware and software used

[1043] Hardware: Mobile devices (smartphones, tablets, etc.)

[1044] Software: Emergency alert email reception app, AI analysis engine, emotion engine

[1045] System program processing

[1046] Server Processing

[1047] The server analyzes the emergency alert email received from the mobile information terminal. This analysis is performed using an internal AI analysis engine. This AI analysis engine uses natural language processing technology to analyze the content of the emergency alert email and identify the type of disaster. For example, if an earthquake alert email is received, the AI ​​analysis engine will analyze the content and determine that it is an earthquake.

[1048] The server generates action instructions based on the results of the AI ​​analysis engine according to the type of disaster. The action instructions are generated based on pre-set templates. The generated action instructions are then sent to the mobile information device via an internet connection.

[1049] Terminal handling

[1050] The device receives the action instructions sent from the server and displays them on the screen. The display uses text formatted for easy viewing by the user. In addition, the device is equipped with an emotion engine that recognizes emotions from the user's facial expressions, tone of voice, and text input. This emotion engine estimates whether the user is scared or not.

[1051] User Action

[1052] The user acts according to the instructions displayed on the device. If the user's emotions affect the system, the emotion engine sends that information to the server. The server then modifies the instructions based on the results of the emotion engine. For example, if the user is scared, the AI ​​uses that information to generate gentler instructions. The modified instructions are then sent back to the device, which then displays them on the screen.

[1053] Specific examples

[1054] Example 1: Earthquake Early Warning

[1055] 1. Device: Receive emergency alert emails.

[1056] 2. Terminal: Sends the received emergency alert email to the server.

[1057] 3. Server: Analyzes the received emergency alert email using an AI analysis engine.

[1058] 4. Server: Based on the analysis results, it determines that it is an earthquake.

[1059] 5. Server: Generates an action instruction such as "An earthquake has occurred. Take cover under a desk."

[1060] 6. Server: Sends the generated action instructions to the terminal.

[1061] 7. Terminal: Displays the received action instructions on the screen.

[1062] 8. Device: Recognizes when a user is scared by their facial expression and tone of voice.

[1063] 9. Server: Based on the user's emotional information, generate a gentle instruction for action such as "Please stay calm and stay here until it is safe."

[1064] 10. Server: Sends the modified action instructions to the device.

[1065] 11. Terminal: Displays the corrected action instructions on the screen.

[1066] Example prompts to input to the generative AI model

[1067] "I received an earthquake alert email. If the user is scared, what action instructions should I generate?"

[1068] By inputting this prompt into a generative AI model, it can learn how to generate appropriate behavioral instructions based on the user's emotions.

[1069] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1070] Step 1:

[1071] The terminal receives an emergency alert email. As input, the emergency alert email arrives at the mobile information terminal. As output, the data of the received emergency alert email is obtained. In concrete terms, the terminal's email reception function is activated and the emergency alert email is received.

[1072] Step 2:

[1073] The terminal sends the received emergency alert email to the server. The input is the data of the received emergency alert email. The output is the data of the emergency alert email sent to the server. Specifically, the terminal's communication function is activated and the data is sent to the server via the Internet.

[1074] Step 3:

[1075] The server analyzes the received emergency alert email using an AI analysis engine. The input is the data from the emergency alert email. The output is the analysis result, which indicates the type of disaster. Specifically, the server's AI analysis engine analyzes the content of the email using natural language processing technology and identifies the type of disaster.

[1076] Step 4:

[1077] The server determines the type of disaster based on the analysis results. The input is the analysis results of the AI ​​analysis engine. The output is the determined type of disaster. Specifically, the server evaluates the analysis results and determines, for example, that it is an "earthquake."

[1078] Step 5:

[1079] The server generates action instructions based on the type of disaster it has determined. The input is the type of disaster it has determined. The output is the generated action instructions. As a specific operation, the server's AI generates action instructions based on a pre-set template.

[1080] Step 6:

[1081] The server sends the generated action instructions to the terminal. The generated action instructions are the input. The action instructions sent to the terminal are obtained as the output. In concrete terms, the server's communication function is activated and data is sent to the terminal via the Internet.

[1082] Step 7:

[1083] The terminal displays the received action instructions on the screen. The input is the action instructions sent from the server. The output is the action instructions displayed on the screen. Specifically, the display function of the terminal is activated and the action instructions are displayed to the user.

[1084] Step 8:

[1085] The device recognizes emotions from the user's facial expressions, tone of voice, text input, etc. Inputs include the user's facial expression data, voice data, and text data. The output is the recognized user's emotional information. Specifically, the device's emotion engine estimates the user's emotions.

[1086] Step 9:

[1087] The server modifies the action instructions based on the results of the emotion engine. The input is the recognized user's emotional information. The output is the modified action instructions. Specifically, the server's AI evaluates the user's emotional information and generates gentler action instructions.

[1088] Step 10:

[1089] The server sends the modified action instructions to the terminal. The input is the modified action instructions. The output is the modified action instructions sent to the terminal. As a specific operation, the communication function of the server is activated and data is sent to the terminal via the Internet.

[1090] Step 11:

[1091] The terminal displays the modified action instructions on the screen. The input is the modified action instructions sent from the server. The output is the modified action instructions displayed on the screen. As a specific operation, the display function of the terminal is activated to display the modified action instructions to the user.

[1092] (Application example 1)

[1093] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1094] Conventional disaster response systems can receive emergency alert emails, determine the type of disaster, and generate appropriate response actions, but they cannot respond by taking into account the user's emotional state, which can lead to users being unable to take appropriate action.There is also a need for more effective evacuation behavior by generating action instructions that correspond to the user's emotions.

[1095] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert email and determining the type of disaster, AI means including an emotion engine for recognizing the user's emotions, and means for displaying the generated actions on the user's smart device. This allows appropriate action instructions to be generated according to the user's emotional state, enabling the user to take evacuation actions more effectively.

[1096] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[1097] "Analysis" is a process of examining the details of the received emergency alert email and identifying the type of disaster.

[1098] "Type of disaster" refers to the specific classification of the disaster that occurred, such as earthquake, tsunami, fire, or flood.

[1099] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[1100] "AI means" refers to means for analyzing data using artificial intelligence technology and making judgments and predictions.

[1101] An "emotion engine" is a technology that estimates emotions from a user's facial expressions, tone of voice, text input, etc.

[1102] "Smart devices" refers to portable electronic devices with internet connectivity, such as smartphones and tablets.

[1103] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[1104] "Behavioral history" refers to a record of the actions a user has taken in the past, and is the data that AI uses to learn.

[1105] A system for implementing the present invention is configured as follows.

[1106] First, the server is provided with a means for receiving emergency alert emails such as earthquake, tsunami, disaster evacuation information, etc. This allows the server to quickly receive emergency alert emails and provide information to users immediately.

[1107] The server then analyzes the content of the received emergency alert email and determines the type of disaster. This analysis is performed using an AI analysis engine (such as TensorFlow). The AI ​​analysis engine analyzes the text data in the emergency alert email and identifies the type of disaster, such as earthquake, tsunami, or fire.

[1108] Furthermore, the server is equipped with AI means, including an emotion engine that recognizes the user's emotions. The emotion engine uses technologies such as OpenCV and Google Cloud Natural Language API to analyze the user's facial expressions and tone of voice, allowing it to estimate the user's emotional state, such as whether they are scared or calm.

[1109] The server generates the appropriate action to take based on the type of disaster it has determined and the user's emotional state. The generated action instructions are displayed on the user's smart device (e.g., smartphone or tablet). React Native and other frameworks are used as user interface (UI) frameworks.

[1110] For example, when an earthquake alert is received, the AI ​​displays, "An earthquake has occurred. Please evacuate to a safe place." If the user is scared, the emotion engine recognizes this information and generates a gentler instruction to act, such as, "Please stay calm. Evacuate to a safe place."

[1111] Examples of prompts for generative AI models include:

[1112] Prompt: "An earthquake alert has been received. If the user is scared, generate calm instructions for action."

[1113] In this way, appropriate action instructions can be generated according to the user's emotional state, enabling the user to take more effective evacuation actions.

[1114] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1115] Step 1:

[1116] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the emergency alert email, and the output is the content of the received email. Specifically, the server uses the emergency alert email reception API to obtain emergency alert emails in real time.

[1117] Step 2:

[1118] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the content of the received email, and the output is the identified type of disaster. Specifically, the server uses an AI analysis engine (such as TensorFlow) to analyze the text data in the email and identify the type of disaster, such as earthquake, tsunami, or fire.

[1119] Step 3:

[1120] The server uses an emotion engine to recognize the user's emotions. The inputs are the user's facial expressions, tone of voice, and text input, and the output is the estimated user's emotional state. Specifically, the server captures the user's facial expressions and voice using the smart device's camera and microphone, and estimates the emotion using technologies such as OpenCV and Google Cloud Natural Language API.

[1121] Step 4:

[1122] The server generates the action to be taken based on the determined type of disaster and the user's emotional state. The input is the identified type of disaster and the estimated user's emotional state, and the output is the generated action instructions. Specifically, the server uses a generative AI model to generate action instructions according to the type of disaster and the user's emotions.

[1123] Step 5:

[1124] The server displays the generated action instructions on the user's smart device. The input is the generated action instructions, and the output is the action instructions displayed on the smart device screen. Specifically, the server uses a user interface (UI) framework (such as React Native) to display the action instructions in a user-friendly format.

[1125] Step 6:

[1126] The user takes evacuation action according to the action instructions displayed on the smart device. The input is the action instructions displayed on the smart device, and the output is the user's evacuation action. Specifically, the user follows the instructions displayed on the smart device screen and evacuates to a safe place.

[1127] In this way, the server, terminal, and user work together to generate appropriate action instructions according to the user's emotional state, enabling the user to take more effective evacuation actions.

[1128] Example 2

[1129] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1130] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information or emotions, making it difficult to provide optimal evacuation actions for each individual user. They also lacked the ability to learn from the user's past behavioral history and suggest appropriate evacuation actions. This led to problems such as users being unable to take appropriate evacuation actions, and safety in the event of a disaster not being ensured.

[1131] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1132] In this invention, the server includes means for receiving emergency news messages such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency news message to determine the type of disaster, means for acquiring the user's location information, means for analyzing the user's emotions, means for using a generative AI model to generate an action to be taken based on the determined type of disaster, the acquired location information, and the analyzed emotions, and means for displaying the generated action on the user's mobile device. This makes it possible to provide appropriate evacuation actions that take into account the user's location information and emotions, thereby improving safety during disasters.

[1133] An "emergency message" is a message that includes information about an emergency, such as an earthquake, tsunami, or disaster evacuation information.

[1134] "Type of disaster" refers to the classification of different disasters such as earthquake, tsunami, fire, and flood.

[1135] "Location Information" is data that indicates a user's current geographic location.

[1136] "Emotion" is data that indicates the psychological state of the user, and includes states such as fear, relief, and excitement.

[1137] A "generative AI model" is an artificial intelligence model that generates appropriate actions or answers based on input data.

[1138] "Mobile device" refers to a portable electronic device such as a smartphone or tablet.

[1139] "Behavioral history" refers to a record of actions taken by a user in the past.

[1140] "Analysis" is the process of examining received data in detail and understanding its contents.

[1141] "Learning" is the process of using past data to improve future actions and decisions.

[1142] This invention relates to a system that generates actions to be taken by taking into account the user's location information and emotions. This system operates in cooperation with a server, a terminal, and a user.

[1143] First, the server receives emergency alert messages such as earthquake, tsunami, and disaster evacuation information. These include emergency alert messages sent over general communication networks. The server analyzes the content of the received emergency alert message and determines the type of disaster. Natural language processing technology can be used for the analysis.

[1144] Next, the server uses the device's GPS function to obtain the user's location information. Specifically, it obtains the user's current location using an API that provides location information services (e.g., a map service API). The device then sends the GPS data to the server.

[1145] Furthermore, the server receives text and voice data entered by the user to analyze the user's emotions. Sentiment analysis is performed using a natural language processing library (e.g., sentiment analysis API). The user enters text and voice data into the terminal and sends it to the server.

[1146] The server uses a generative AI model to generate the appropriate action to be taken based on the acquired location information and emotion data. For example, a generative AI model is used. The server displays the generated action on the user's mobile device, allowing the user to take appropriate evacuation action.

[1147] As a concrete example, if a user is in a tsunami warning area, the server uses a location service API to obtain the user's location information and confirms that the user is within the tsunami warning area. The generative AI model then receives a prompt message: "Please tell me what to do if the user is in a tsunami warning area." The generative AI model generates the action "Evacuate to the nearest high ground," and the server notifies the user of this information.

[1148] If the user is scared in a public place, the server uses the sentiment analysis API to analyze the user's emotions and inputs the prompt "What should I do if the user is scared in a public place?" into the generative AI model. The generative AI model generates an action to "instruct the user to evacuate to a safe place with fewer people," and the server notifies the user of this information.

[1149] Examples of prompts include:

[1150] 1. "What should I do if I'm in a tsunami warning area?"

[1151] 2. "What should I do if a user is in a public place and feels scared?"

[1152] In this way, a system is realized in which the server, terminal, and user work together to generate coping actions that take into account the user's location information and emotions, and provide appropriate instructions.

[1153] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1154] Step 1:

[1155] The server receives emergency messages such as earthquake, tsunami, and disaster evacuation information.

[1156] Input: Emergency message

[1157] Specific operations: A server receives an emergency alert message transmitted over a communication network.

[1158] Output: Received emergency alert message

[1159] Step 2:

[1160] The server analyzes the contents of the received emergency message and determines the type of disaster.

[1161] Input: Received emergency alert message

[1162] Specific operation: The server uses natural language processing technology to analyze the content of the message and identify the type of disaster, such as earthquake, tsunami, or fire.

[1163] Output: Disaster type

[1164] Step 3:

[1165] The server uses the device's GPS function to obtain the user's location information.

[1166] Input: User location request

[1167] Specific operation: The device turns on the GPS function to obtain location information and sends it to the server through the location information service API.

[1168] Output: Current location of the user

[1169] Step 4:

[1170] The server receives text and voice data entered by the user in order to analyze the user's sentiment.

[1171] Input: User text or voice data

[1172] How it works: The user inputs text or voice data into the device and sends it to the server, which then uses the sentiment analysis API to analyze the data and identify the user's emotions.

[1173] Output: User emotion data

[1174] Step 5:

[1175] The server uses a generative AI model that generates appropriate actions based on the type of disaster it determines, the location information it obtains, and the emotions it analyzes.

[1176] Input: Type of disaster, user's current location, user's emotional data

[1177] Specific behavior: The server inputs a prompt sentence such as "Please tell me what to do if the user is in a tsunami warning area" into the generative AI model, and the generative AI model generates appropriate behavior.

[1178] Output: Action to be taken

[1179] Step 6:

[1180] The server displays the generated actions on the user's mobile device.

[1181] Input: Action to be taken

[1182] Specific operation: The server notifies the device of the generated action and sends the user a message such as "Please evacuate to the nearest high ground."

[1183] Output: A message to inform the user

[1184] In this way, a system is realized in which the server, terminal, and user work together to generate coping actions that take into account the user's location information and emotions, and provide appropriate instructions.

[1185] (Application example 2)

[1186] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1187] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information or emotional state, making it difficult to provide appropriate evacuation actions tailored to individual situations. Furthermore, if the user is in a state of panic or is in a specific location, the system is unable to provide appropriate evacuation instructions, reducing the effectiveness of the evacuation. This has led to the issue of evacuation actions not functioning adequately to ensure the user's safety.

[1188] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1189] In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert email and determining the type of disaster, AI means for generating an action to be taken based on the determined type of disaster, means for displaying the generated action on the user's mobile device, means for acquiring the user's location information, means for recognizing the user's emotional state, means for generating an action to be taken based on the acquired location information and the recognized emotional state, and means for notifying the user of the generated action. This makes it possible to instruct appropriate evacuation actions taking into account the user's location information and emotional state.

[1190] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[1191] "Type of disaster" refers to the specific type of disaster that occurred, such as earthquake, tsunami, fire, or flood.

[1192] "AI means" refers to means that use artificial intelligence technology to analyze received information and generate appropriate response actions.

[1193] "Mobile devices" refer to electronic devices that users carry and use, such as smartphones and tablets.

[1194] "Location information" refers to the geographic coordinate information of the user's current location.

[1195] "Emotional state" refers to the user's current psychological state, such as fear, relief, panic, etc.

[1196] "Evacuation behavior" refers to specific actions that users should take in the event of a disaster, such as evacuating to higher ground or moving to a safe location.

[1197] The "notification means" refers to a means for informing the user of the generated evacuation action, and includes, for example, a push notification to a mobile device or a voice alert.

[1198] The system for implementing this invention includes a user's mobile terminal, a server, and AI means. The specific configuration and operation of the system will be described below.

[1199] System configuration

[1200] 1. Mobile devices: Mobile devices such as smartphones and tablets are equipped with a GPS sensor to acquire the user's location information, a camera and microphone to recognize the user's emotional state, and an application to receive emergency alert emails and display generated evacuation actions.

[1201] 2. Server: The server receives emergency alert emails, analyzes their contents to determine the type of disaster, and is equipped with AI means to generate appropriate evacuation actions taking into account the user's location and emotional state.

[1202] 3. AI Means: The AI ​​means uses a generative AI model to analyze the received information and generate appropriate evacuation actions. Specifically, it receives the user's location information and emotional state as input and generates prompt sentences.

[1203] System Operation

[1204] 1. Receiving and analyzing emergency alert emails: The server receives emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc. It analyzes the content of the received emails and determines the type of disaster.

[1205] 2. Location and emotional state acquisition: The mobile device acquires the user's location using the GPS sensor, and recognizes the user's emotional state using the camera and microphone.

[1206] 3. Evacuation behavior generation: The server's AI generates appropriate evacuation behavior based on the acquired location information and emotional state. For example, if the user is in a tsunami warning area, it will instruct the user to evacuate to the nearest high ground. If the user is in a panic, it will instruct the user to evacuate to a safe place with fewer people.

[1207] 4. Evacuation Action Notification: The mobile device notifies the user of the generated evacuation action via push notification and / or voice alert.

[1208] Specific examples

[1209] For example, if a user is near the coast and a tsunami warning is issued, the system operates as follows:

[1210] 1. The server receives an emergency email alerting the tsunami and determines whether a tsunami has occurred.

[1211] 2. The mobile device uses GPS sensors to obtain the user's location and cameras and microphones to recognize the user's emotional state.

[1212] 3. The server's AI means recognizes that the user is near the coast and is scared, and generates the following prompt:

[1213] The user is currently near the coast. A tsunami warning has been issued and the user is scared. Please tell them to evacuate to the nearest high ground.

[1214] 4. The mobile device notifies the user of the generated evacuation action and instructs them to evacuate to the nearest high ground.

[1215] In this way, it becomes possible to provide appropriate evacuation instructions that take into account the user's location information and emotional state.

[1216] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1217] Step 1:

[1218] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. It receives the emergency alert email as input, analyzes its content, and determines the type of disaster. It generates data that identifies the type of disaster as output. Specifically, it performs text analysis on the email content, extracts keywords and phrases, and classifies the type of disaster.

[1219] Step 2:

[1220] The device obtains the user's location information using the GPS sensor. It receives GPS data as input and determines the latitude and longitude of the user's current location. It generates the user's location information as output. Specifically, it analyzes the signal from the GPS sensor and calculates the user's precise location.

[1221] Step 3:

[1222] The device uses a camera and microphone to recognize the user's emotional state. It receives camera images and audio data as input and applies emotion recognition algorithms to identify the user's emotional state. As output, it generates data that indicates the user's emotional state. Specifically, it performs image and audio analysis to classify the user's emotions, such as fear or relief.

[1223] Step 4:

[1224] The server's AI means generates appropriate evacuation actions based on the acquired location information and emotional state. It receives location information and emotional state data as input and generates prompt sentences using a generative AI model. It generates instructions for evacuation actions as output. Specifically, it incorporates location information and emotional state into the prompt sentences to create specific evacuation instructions for the user.

[1225] Step 5:

[1226] The device notifies the user of the generated evacuation behavior. It receives evacuation behavior instructions as input and notifies the user via push notification or voice alert. It outputs the evacuation instructions to the user. Specifically, it uses the device's notification function to inform the user of the evacuation behavior via screen display or voice message.

[1227] In this way, it becomes possible to provide appropriate evacuation instructions that take into account the user's location information and emotional state.

[1228] Example 3

[1229] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1230] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's past behavioral history or emotional data, making it difficult to provide optimal evacuation instructions for each individual user. Furthermore, because evacuation instructions do not take into account the user's location information in real time, it is difficult to provide prompt and appropriate evacuation instructions in an emergency. This can easily lead to users panicking and delays in evacuation.

[1231] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1232] In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, artificial intelligence means for generating response actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for collecting data on the user's past behavioral history, means for preprocessing the collected data, means for training an artificial intelligence model using the preprocessed data, means for generating response actions using the trained artificial intelligence model, means for adjusting the generated response actions in consideration of the user's emotional data, and means for transmitting the adjusted response actions to the user's mobile device. This enables individually optimized evacuation instructions that take into consideration the user's past behavioral history and emotional data, thereby realizing quick and appropriate evacuation actions in emergencies.

[1233] "Emergency emails for earthquakes, tsunamis, disaster evacuation information, etc." are emails containing emergency information that are sent to users promptly in the event of a disaster.

[1234] "Means for receiving" refers to a device or software that has a communication function for receiving emergency alert emails.

[1235] The "means for analyzing and determining the type of disaster" refers to an algorithm or program that analyzes the contents of the received emergency alert email and identifies the type of disaster, such as earthquake, tsunami, or fire.

[1236] The "artificial intelligence means for generating appropriate actions to be taken" refers to an artificial intelligence model and its execution environment for generating appropriate evacuation actions and countermeasures based on the type of disaster.

[1237] The "means for displaying on the user's mobile device" refers to an interface and software for displaying the generated action instructions on the user's mobile device such as a smartphone or tablet.

[1238] "Means for collecting user's past behavioral history data" refers to sensors and databases that collect data on the user's past behavior and the places they have visited.

[1239] "Means for preprocessing collected data" refers to algorithms or programs for preprocessing collected data, such as filling in missing values ​​and normalizing the data.

[1240] A "means for training an artificial intelligence model using preprocessed data" is a machine learning framework and its execution environment for training an artificial intelligence model using preprocessed data.

[1241] "Means for generating response actions using a trained artificial intelligence model" refers to algorithms and programs for generating specific evacuation actions and countermeasures for users using a trained artificial intelligence model.

[1242] The "means for taking into account the user's emotional data and making adjustments" refers to an algorithm or program for adjusting the generated coping behavior in consideration of the user's emotional state.

[1243] The "means for transmitting the adjusted response action to the user's mobile device" refers to communication functions and software for transmitting the adjusted action instructions to the user's mobile device.

[1244] MODE FOR CARRYING OUT THE INVENTION

[1245] This invention is a system that provides users with prompt and appropriate evacuation instructions in the event of a disaster. The system consists of three main components: a server, a terminal, and users.

[1246] server

[1247] The server has the following functions:

[1248] 1. Data Collection

[1249] The server collects the user's past behavioral history data, including information about the evacuation shelters the user visited during evacuation drills and past behavioral patterns.

[1250] The server also collects user emotional data, including whether the user has a history of panicking.

[1251] 2. Data Preprocessing

[1252] The server preprocesses the collected data, specifically by filling in missing values ​​and normalizing the data.

[1253] The server converts the pre-processed data into a format suitable for training an AI model.

[1254] 3. Training the AI ​​model

[1255] The server uses the preprocessed data to train an AI model using frameworks such as TensorFlow and PyTorch.

[1256] The server evaluates the model's performance during training and adjusts hyperparameters as needed.

[1257] 4. Generating Action Instructions

[1258] The server uses the trained AI model to generate response actions based on the user's behavioral history, for example, giving priority to routes to evacuation shelters visited during evacuation drills.

[1259] The server takes emotion data into account and generates more specific and concise instructions for actions, for example, providing concise and clear evacuation instructions to users who tend to panic.

[1260] 5. Sending instructions

[1261] The server sends the generated action instructions to the device using the HTTPS protocol.

[1262] Terminal

[1263] The terminal has the following functions:

[1264] 1. Receiving instructions for action

[1265] The terminal receives the action instructions sent from the server. Devices such as smartphones and tablets are used for receiving the instructions.

[1266] 2. Display of action instructions

[1267] The terminal displays the received action instructions to the user using the application's user interface.

[1268] 3. Obtaining location information

[1269] The device obtains the user's current location information, which is obtained using the GPS function.

[1270] The terminal transmits the acquired location information to the server.

[1271] 4. Update of Action Instructions

[1272] The terminal receives updated action instructions from the server in real time and notifies the user.

[1273] user

[1274] The user performs the following actions:

[1275] 1. Confirmation of action instructions

[1276] The user checks the action instructions displayed on the device.

[1277] 2. Taking action

[1278] During evacuation drills and actual evacuations, users act based on instructions from the device.

[1279] 3. Sending location information

[1280] Users send their location information to the server via their devices.

[1281] 4. Check for updated instructions

[1282] The user checks the updated action instructions from the terminal and corrects the action if necessary.

[1283] Specific examples

[1284] Data collection: The server stores information about evacuation shelters that users have visited in the past during evacuation drills in a database. For example, it stores data such as "User A visited evacuation shelter B on March 15, 2023."

[1285] Data preprocessing: The server imputes missing values ​​in the collected data and normalizes the data. For example, it estimates and imputes missing dates and times of visits to evacuation centers.

[1286] Training the AI ​​model: The server uses the preprocessed data to train the AI ​​model, for example, using TensorFlow to build a model that learns user behavior patterns.

[1287] Generation of action instructions: The server uses the trained AI model to generate action instructions for user A, such as "give priority to the route to shelter B."

[1288] Sending action instructions: The server sends the generated action instructions to the terminal. For example, it sends a message via HTTPS saying, "Give priority to the route to shelter B."

[1289] Prompt Sentence Examples

[1290] "Build an AI model that learns the user's past behavioral history and prioritizes routes to evacuation shelters visited during evacuation drills."

[1291] "Please design a system that learns the user's past behavioral history and emotional data, and generates concise and clear instructions for users who tend to panic." The flow of the identification process in Example 3 will be described with reference to FIG. 21.

[1292] Step 1:

[1293] Data collection

[1294] The server collects the user's past behavioral history data. Specifically, it stores information about the evacuation shelters the user visited during evacuation drills and their past behavioral patterns in a database. The input is the user's behavioral history data, and the output is the behavioral history data stored in the database. For example, it stores data such as "User A visited evacuation shelter B on March 15, 2023."

[1295] Step 2:

[1296] Collecting Emotional Data

[1297] The server also collects user emotional data, including information on whether the user has been prone to panic in the past. The input is the user emotional data, and the output is the emotional data stored in the database. For example, the server stores data such as "User B has been prone to panic in the past."

[1298] Step 3:

[1299] Data Preprocessing

[1300] The server preprocesses the collected data. Specifically, it complements missing values ​​and normalizes the data. The input is the collected raw data, and the output is the preprocessed data. For example, it estimates and complements missing dates and times of visits to evacuation shelters.

[1301] Step 4:

[1302] Training an AI model

[1303] The server uses the preprocessed data to train an AI model. The frameworks used are TensorFlow and PyTorch. The input is the preprocessed data, and the output is the trained AI model. For example, a model that learns user behavior patterns is built.

[1304] Step 5:

[1305] Generate action instructions

[1306] The server uses the trained AI model to generate response actions based on the user's behavioral history. The input is the trained AI model and the user's behavioral history data, and the output is the generated action instructions. For example, it may give priority to routes to evacuation shelters visited during evacuation drills.

[1307] Step 6:

[1308] Adjustments based on emotional data

[1309] The server adjusts the generated response actions taking into account the emotional data. The inputs are the generated action instructions and the user's emotional data, and the output is the adjusted action instructions. For example, it provides concise and clear evacuation instructions to a user who is prone to panic.

[1310] Step 7:

[1311] Sending action instructions

[1312] The server sends the adjusted action instructions to the terminal. The HTTPS protocol is used for communication. The input is the adjusted action instructions, and the output is the action instructions sent to the terminal. For example, it sends a message saying, "Give priority to the route to shelter B."

[1313] Step 8:

[1314] Receiving action instructions

[1315] The terminal receives the action instructions sent from the server. A device such as a smartphone or tablet is used for reception. The input is the action instructions sent from the server, and the output is the action instructions displayed on the terminal.

[1316] Step 9:

[1317] Display of action instructions

[1318] The terminal displays the received action instructions to the user. The display uses the user interface of the application. The input is the received action instructions, and the output is the action instructions displayed to the user.

[1319] Step 10:

[1320] Obtaining location information

[1321] The terminal obtains the user's current location information. The GPS function is used to obtain the location information. The input is the user's current location information, and the output is the location information sent to the server.

[1322] Step 11:

[1323] Sending location information

[1324] The terminal transmits the acquired location information to the server. The input is the acquired location information, and the output is the location information transmitted to the server.

[1325] Step 12:

[1326] Updates to instructions

[1327] The terminal receives updated action instructions from the server in real time and notifies the user. The input is the updated action instructions from the server, and the output is the updated action instructions notified to the user.

[1328] Step 13:

[1329] Confirmation of action instructions

[1330] The user confirms the action instructions displayed on the terminal. The input is the action instructions displayed on the terminal, and the output is the user's confirmation action.

[1331] Step 14:

[1332] Taking action

[1333] During evacuation drills and actual evacuations, users act based on instructions from the terminal. The input is the instruction to act from the terminal, and the output is the user's evacuation behavior.

[1334] Step 15:

[1335] Sending location information

[1336] The user sends his / her location information to the server through the terminal. The input is the user's location information, and the output is the location information sent to the server.

[1337] Step 16:

[1338] Check for updated action instructions

[1339] The user checks the updated action instructions from the terminal and modifies the action if necessary. The input is the updated action instructions, and the output is the user's modified action.

[1340] (Application example 3)

[1341] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1342] Conventional emergency evacuation systems provide uniform evacuation instructions without considering the user's past behavioral history or emotional data, making it difficult to provide optimal evacuation routes and action instructions for individual users. Furthermore, for users who tend to panic, specific and concise instructions are lacking, making it difficult for them to take appropriate action in an emergency. To solve these issues, a system is needed that can learn the user's past behavioral history and emotional data and provide optimal evacuation routes and action instructions for each individual user.

[1343] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1344] In this invention, the server includes means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating appropriate actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for learning the user's past behavioral history and emotional data and generating appropriate actions based on the learning results, and means for providing optimal evacuation routes and action instructions in an emergency. This makes it possible to provide individually optimal evacuation routes and action instructions that take into account the user's past behavioral history and emotional data.

[1345] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[1346] "Type of disaster" refers to the classification of different disasters such as earthquakes, tsunamis, fires, and floods.

[1347] "AI means" refers to means for analyzing data using artificial intelligence technology and generating appropriate instructions for action.

[1348] "Mobile devices" refer to electronic devices that users carry and use, such as smartphones and tablets.

[1349] "Past behavioral history" refers to a record of the actions and choices a user has made in the past.

[1350] "Emotion data" refers to data relating to the user's emotional state, including, for example, panicked or calm.

[1351] An "evacuation route" refers to a route that a user can use to evacuate to a safe place in the event of an emergency.

[1352] "Action instructions" are instructions that indicate specific actions that a user should take in an emergency.

[1353] As an embodiment of the present invention, the following system can be constructed. This system is configured using a user's mobile terminal, a server, and an AI model.

[1354] First, the server has a means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The server also includes a means for analyzing the content of the received emergency alert emails and determining the type of disaster. This allows the server to generate actions to be taken based on the type of disaster.

[1355] Next, the server is equipped with an AI means that learns the user's past behavioral history and emotional data. This AI means generates optimal evacuation routes and instructions for action in an emergency based on the user's past behavioral history and emotional data. For example, it can give priority to routes to evacuation shelters that the user has previously visited during evacuation drills. It can also generate specific and concise instructions for action for users who tend to panic.

[1356] The generated action instructions are displayed on the user's mobile device, which is an electronic device carried by the user, such as a smartphone or tablet, allowing the user to take appropriate action in an emergency.

[1357] To realize this system, software such as Python and Scikit-learn is used. Python is a programming language widely used for data analysis and machine learning, and Scikit-learn is a library for machine learning. Using these software, an AI model is built that learns the user's past behavioral history and emotional data and generates optimal behavioral instructions.

[1358] For example, if a user has previously visited "Shelter A" during an evacuation drill, the route to "Shelter A" will be displayed first in the event of an emergency. Also, if the user has a tendency to panic in the past, specific and concise instructions such as "Remain calm and head to the nearest evacuation center" will be displayed in the event of an emergency.

[1359] An example of a prompt sentence might be:

[1360] "Develop an application that learns from the user's past behavioral history and emotional data and provides optimal evacuation routes and instructions in the event of an emergency. Please include a function that gives priority to routes to evacuation shelters that the user has previously visited in evacuation drills, and generates specific and concise instructions for users who tend to panic."

[1361] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1362] Step 1:

[1363] The server receives emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc. The input is the emergency alert email, and the output is the data of the received emergency alert email. This data includes information such as the type of disaster, the location of the occurrence, and the time of occurrence.

[1364] Step 2:

[1365] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the data from the emergency alert email, and the output is the type of disaster (earthquake, tsunami, fire, etc.). Specifically, it analyzes the text of the email and identifies the type of disaster using keywords and pattern matching.

[1366] Step 3:

[1367] The server generates the actions to be taken based on the type of disaster. The input is the type of disaster, and the output is instructions for the actions to be taken. Specific operations refer to predefined action instructions according to the type of disaster and generate appropriate instructions.

[1368] Step 4:

[1369] The server obtains the user's past behavioral history and emotional data. The input is the user ID, and the output is the past behavioral history and emotional data. Specifically, it queries and obtains the user's behavioral history and emotional data from the database.

[1370] Step 5:

[1371] The server trains an AI model using the acquired past behavioral history and emotional data. The inputs are past behavioral history and emotional data, and the output is a trained AI model. Specifically, it uses a machine learning library such as Scikit-learn to fit the data to the model.

[1372] Step 6:

[1373] The server uses a trained AI model to generate optimal evacuation routes and action instructions in the event of an emergency. The inputs are current situation data (location information, type of disaster, etc.) and the trained AI model, and the output is optimal evacuation routes and action instructions. Specifically, the current situation data is input into the AI ​​model, and action instructions are generated based on the prediction results.

[1374] Step 7:

[1375] The server sends the generated action instructions to the user's mobile device. The inputs are the action instructions and the user's mobile device information, and the output is the action instructions displayed on the user's mobile device. Specifically, the server formats the action instructions as a message and sends a push notification to the mobile device.

[1376] Step 8:

[1377] The user's mobile device displays the received action instructions. The input is the action instructions sent from the server, and the output is the action instructions that the user can check on the screen. The specific operation is to analyze the received message and display it on the screen.

[1378] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1379] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1380] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[1381] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1382] [Third embodiment]

[1383] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1384] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1385] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. 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. The database 24 and the communication I / F 26 are also 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).

[1386] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1387] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1388] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1389] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1390] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1391] The specific processing program 56 is an example of a "program" according to the technology of the present 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 in accordance with the specific processing program 56 executed on the RAM 30.

[1392] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1393] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1394] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1395] "Example 1"

[1396] One embodiment of the present invention is a system using a smartphone. This system has the function of receiving emergency alert emails, such as earthquake, tsunami, and disaster evacuation information. The content of the received emergency alert email is analyzed by an AI inside the system to determine the type of disaster. The AI ​​generates actions to be taken based on the determined type of disaster and displays the action instructions on the smartphone screen.

[1397] "Example 2"

[1398] Furthermore, another embodiment of the present invention is a system that takes into account the user's location information. In this system, AI acquires the user's location information and generates an action to be taken based on that location information. For example, if the user is in an area where a tsunami warning has been issued, the AI ​​generates an action such as "evacuate to the nearest high ground."

[1399] "Example 3"

[1400] Another embodiment of the present invention is a system that learns the user's past behavioral history. In this system, AI learns the user's past behavioral history and generates appropriate actions based on the learning results. For example, it is possible to prioritize routes to evacuation shelters that the user has visited in the past during evacuation drills.

[1401] The processing flow of each embodiment will be described below.

[1402] "Example 1"

[1403] Step 1: Your smartphone will receive emergency alert emails about earthquakes, tsunamis, disaster evacuation information, etc.

[1404] Step 2: The contents of the received emergency alert email are analyzed by the AI ​​within the system.

[1405] Step 3: The AI ​​determines the type of disaster and generates appropriate actions based on that type.

[1406] Step 4: Display the generated action instructions on the smartphone screen.

[1407] "Example 2"

[1408] Step 1: AI obtains the user's location information.

[1409] Step 2: Based on the acquired location information, the AI ​​generates the action to be taken.

[1410] Step 3: Display the generated action instructions on the smartphone screen.

[1411] "Example 3"

[1412] Step 1: The AI ​​learns the user's past behavioral history.

[1413] Step 2: Based on the learning results, the AI ​​generates the action to be taken.

[1414] Step 3: Display the generated action instructions on the smartphone screen.

[1415] Example 1

[1416] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1417] Conventional emergency alert systems required users to analyze the content of the received emergency alert email and determine the appropriate course of action, making it difficult to respond quickly and accurately. Furthermore, because they were unable to provide individualized responses that took into account the user's location information or past behavioral history, they could only provide uniform instructions to all users. This made it difficult for users to take appropriate evacuation actions.

[1418] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1419] In this invention, the server includes: means for receiving emergency alert emails for earthquakes, tsunamis, disaster evacuation information, etc.; means for analyzing the content of the received emergency alert emails and determining the type of disaster; artificial intelligence means for generating response actions based on the determined type of disaster; means for displaying the generated actions on the user's mobile information terminal; means for the artificial intelligence means to analyze the content of the emergency alert emails using natural language processing technology; and means for the artificial intelligence means to display the generated action instructions on the screen of the mobile information terminal. This eliminates the need for the user to analyze the content of the received emergency alert emails themselves, allowing them to receive quick and accurate action instructions. Furthermore, it is possible to respond individually to evacuation actions taking into account the user's location information and past behavioral history, allowing for more appropriate evacuation actions.

[1420] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[1421] "Mobile information terminal" refers to a portable information processing device such as a smartphone or tablet.

[1422] "Artificial intelligence means" refers to a system that uses technologies such as machine learning and deep learning to analyze data and make judgments and predictions.

[1423] "Natural language processing technology" is a technology that allows computers to understand, interpret, and generate human language.

[1424] "Action instructions" are instructions that indicate specific actions that a user should take in a particular situation.

[1425] "Location information" is data that indicates a user's current location and is obtained using technologies such as GPS.

[1426] "Past behavior history" is a record of actions taken by a user in the past, and is used to predict future actions based on this.

[1427] This invention is a system that receives emergency emails about earthquakes, tsunamis, disaster evacuation information, etc., analyzes the contents of the emails, and generates appropriate instructions for actions. A specific embodiment of this system will be described below.

[1428] First, let's assume that a user has a mobile information terminal (smartphone). The terminal has the function to receive emergency alert emails. Emergency alert emails are sent by official organizations such as the government or the Japan Meteorological Agency. The terminal receives these emails through a built-in communication module (e.g., a 4G / 5G modem).

[1429] Next, the server analyzes the content of the received emergency alert email. Specifically, it extracts the text data from the email and analyzes it using natural language processing (NLP) technology. The generative AI model used here is, for example, BERT or GPT-3. The server obtains the content of the email in text format and inputs it into the generative AI model. For example, it analyzes the text, "A large earthquake has occurred. Please evacuate to a safe place immediately."

[1430] The server uses a generative AI model to determine the type of disaster from the email content. The model analyzes keywords and context within the text to identify disasters such as earthquakes, tsunamis, and fires. For example, if the text "A large earthquake has occurred" is input, the model will determine this as an "earthquake."

[1431] Next, the server generates actions that the user should take based on the type of disaster determined. The generative AI model generates appropriate action instructions according to the type of disaster. For example, in the case of an earthquake, it generates specific action instructions such as "hide under a desk," and in the case of a tsunami, it generates specific action instructions such as "evacuate to higher ground." The server inputs a prompt statement to the generative AI model saying, "Please generate action instructions in the case of an earthquake," and the model generates the action instruction, "Hide under a desk or evacuate outside the building."

[1432] Finally, the terminal displays the generated action instructions on the smartphone screen. The user checks the action instructions displayed on the screen and takes appropriate action. For example, a message such as "Hide under a desk or evacuate outside the building" is displayed on the smartphone screen.

[1433] As a concrete example, consider the case where a user receives an emergency alert email on their smartphone. The email message reads, "A large earthquake has occurred. Please evacuate to a safe location immediately." In this case, the server inputs the following prompt sentence into the generative AI model:

[1434] Prompt text: "An emergency alert email has been received stating, 'A large earthquake has occurred. Please evacuate to a safe place immediately.'" Analyze the contents of this email and generate the action the user should take.

[1435] Based on this prompt, the generative AI model generates the following action instructions:

[1436] Action Instructions: "Hide under a desk or leave the building."

[1437] This instruction for action is displayed on the smartphone screen, prompting the user to take appropriate action.

[1438] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1439] Step 1: Receiving emergency alert emails

[1440] The device receives emergency alert emails. Specifically, it receives emergency alert emails sent by official organizations such as the government and the Japan Meteorological Agency via the device's communications module (4G / 5G modem).

[1441] Input: Emergency alert email sent by official organization

[1442] Output: Text data of the received emergency alert email

[1443] Specific operation: If a user has a smartphone, the device will automatically receive an emergency alert email. For example, if an earthquake alert is sent, the device will receive the email.

[1444] Step 2: Analyzing the email content

[1445] The server analyzes the content of the received emergency alert email. Specifically, it extracts the text data from the email and analyzes it using natural language processing technology. The generative AI models used here are BERT and GPT-3.

[1446] Input: Text data of received emergency alert email

[1447] Output: Parsed email content text data

[1448] Specific operation: The server retrieves the contents of the email in text format and inputs it into the generative AI model. For example, it analyzes the text, "A large earthquake has occurred. Please evacuate to a safe place immediately."

[1449] Step 3: Determine the type of disaster

[1450] The server uses a generative AI model to determine the type of disaster from the email content. The model analyzes keywords and context within the text to identify disasters such as earthquakes, tsunamis, and fires.

[1451] Input: Parsed email content text data

[1452] Output: Type of disaster determined

[1453] Specific operation: The server inputs the text "A large earthquake has occurred" into the generative AI model, and the model analyzes this and determines that it is an "earthquake."

[1454] Step 4: Generate action instructions

[1455] The server generates the actions that the user should take based on the determined type of disaster, and the generative AI model generates appropriate action instructions according to the type of disaster.

[1456] Input: Type of disaster determined

[1457] Output: Generated action instructions

[1458] Specific operation: The server inputs a prompt statement to the generative AI model saying, "Generate action instructions in the event of an earthquake." The model then generates the action instruction, "Hide under a desk or evacuate outside the building."

[1459] Step 5: Displaying Action Instructions

[1460] The terminal displays the generated action instructions on the smartphone screen, and the user checks the action instructions displayed on the screen and takes appropriate action.

[1461] Input: Generated action instructions

[1462] Output: Instructions displayed on the smartphone screen

[1463] Specific operation: The device displays the action instructions received from the server on the screen. For example, a message such as "Hide under a desk or evacuate outside the building" will be displayed on the smartphone screen.

[1464] (Application example 1)

[1465] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1466] Conventional disaster response systems only received emergency alert emails and were unable to provide users with specific response actions. It was also difficult to generate appropriate action instructions in real time depending on the type of disaster, making it difficult for users to act quickly and appropriately. Furthermore, because they were unable to respond individually based on the user's location information or past behavioral history, they could only provide generic instructions.

[1467] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1468] In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating appropriate actions based on the determined type of disaster, means for displaying the generated actions on the user's mobile information terminal, means for determining the type of disaster using a generative AI model and generating appropriate actions, and means for generating action instructions by inputting a prompt sentence into the generative AI model. This allows the user, upon receiving the emergency alert email, to quickly learn specific appropriate actions based on the type of disaster analyzed by the AI, enabling individual responses that take into account location information and past behavioral history.

[1469] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[1470] "Analysis" is the process of examining the contents of the received emergency alert email in detail to determine the type of disaster and its impact.

[1471] "Type of disaster" refers to the specific type of disaster that occurred, such as earthquake, tsunami, fire, or flood.

[1472] "Actions to be taken" refers to specific instructions for actions that users should take depending on the type of disaster.

[1473] "AI means" refers to means that use artificial intelligence technology to analyze data, determine the type of disaster, and generate appropriate actions to respond.

[1474] The term "mobile information terminal" refers to an information processing device that is carried and used by a user, such as a smartphone or tablet.

[1475] A "generative AI model" is a generative artificial intelligence model, an AI model that has the ability to generate new data and information based on input data.

[1476] A "prompt" is text data input to a generative AI model that contains instructions or questions to obtain a specific output.

[1477] A system for implementing this invention is configured as follows: A server has means for receiving emergency alert emails, such as earthquake, tsunami, and disaster evacuation information. The content of the received emergency alert emails is analyzed in detail by analysis means within the server, and the type of disaster is determined. The analysis means determines the type of disaster using a generative AI model and generates actions to be taken in response.

[1478] The generated action instructions are displayed on the user's mobile information device, which is an information processing device such as a smartphone or tablet that the user always carries with them. The generative AI model operates by inputting prompt sentences, which contain instructions or questions to obtain a specific output.

[1479] As a specific example, suppose the server receives an emergency alert email with the content "Emergency Alert: An earthquake has occurred. The epicenter is in Tokyo." The analysis means analyzes the content of this email and inputs the following prompt sentence into the generative AI model.

[1480] Please analyze the contents of the emergency alert email below and determine the type of disaster.

[1481] Breaking News: An earthquake has occurred. The epicenter is in Tokyo.

[1482] Based on this prompt, the generative AI model determines that it is an "earthquake" and then inputs the following prompt to generate the action to be taken:

[1483] If the disaster type is an earthquake, generate the actions to be taken.

[1484] The generative AI model generates specific instructions, such as "hide under the desk," which are displayed on the user's mobile device. The user can then act quickly by following the instructions displayed on the mobile device.

[1485] The system is implemented using Python and OpenAI's API. Python is a programming language well-suited for data analysis and AI model operation, while OpenAI's API provides an interface for operating the generative AI model. This allows users to quickly learn specific response actions based on the type of disaster analyzed by the AI ​​when they receive an emergency alert email.

[1486] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1487] Step 1:

[1488] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the content of the emergency alert email, and the output is the received email data. Specifically, the server retrieves the emergency alert email from the mail server and stores its content in memory.

[1489] Step 2:

[1490] The server analyzes the content of the received emergency alert email. The input is the received email data, and the output is the analyzed disaster type. Specifically, the server inputs the following prompt sentence into the generative AI model:

[1491] Please analyze the contents of the emergency alert email below and determine the type of disaster.

[1492] Breaking News: An earthquake has occurred. The epicenter is in Tokyo.

[1493] Based on this prompt, the generative AI model determines the type of disaster to be "earthquake."

[1494] Step 3:

[1495] The server generates the action to be taken based on the type of disaster determined. The input is the type of disaster, and the output is the action to be taken. Specifically, the server inputs the following prompt sentence into the generation AI model:

[1496] If the disaster type is an earthquake, generate the actions to be taken.

[1497] Based on this prompt, the generative AI model generates specific action instructions, such as "hide under the desk."

[1498] Step 4:

[1499] The server sends the generated action instructions to the user's mobile information device. The input is the action instructions to be handled, and the output is the action instructions to be displayed on the user's mobile information device. Specifically, the server formats the action instructions as a text message and sends it to the user's mobile information device.

[1500] Step 5:

[1501] The user acts according to the instructions displayed on the mobile information terminal. The input is the instructions displayed on the mobile information terminal, and the output is the user's specific actions. As a specific action, the user follows instructions such as "hide under a desk" to evacuate to a safe place.

[1502] Example 2

[1503] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1504] Conventional disaster evacuation systems provide uniform evacuation information without considering the user's location information, making it difficult to suggest optimal evacuation actions for each individual user. Furthermore, because they do not consider the user's past behavioral history, they are unable to suggest appropriate actions based on the user's characteristics. This has led to the issue of not being able to adequately ensure the safety of users in emergencies.

[1505] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1506] In this invention, the server includes a means for acquiring user location information, a means for transmitting the acquired location information to the server, and a means for the server to analyze the location information received and generate an action to be taken using a generative AI model. This makes it possible to propose appropriate evacuation actions based on the user's current location in real time. In addition, the generative AI model learns the user's past behavioral history and generates an action to be taken based on the learning results, making it possible to propose more appropriate actions according to the user's characteristics.

[1507] An "emergency message" is a message that includes information about an emergency, such as an earthquake, tsunami, or disaster evacuation information.

[1508] "Location information" means data that indicates a user's current geographic location, including latitude and longitude information.

[1509] The "server" is a computer system that receives and analyzes the user's location information and uses a generative AI model to generate an action to be taken.

[1510] A "generative AI model" is an artificial intelligence model that generates appropriate behavior based on a user's location information and past behavioral history.

[1511] A "prompt" is a textual instruction that is input into a generative AI model to generate appropriate behavior based on the user's location and situation.

[1512] "Terminal" refers to a device used by a user, including a smartphone or tablet.

[1513] "Analysis" is the process of interpreting information and deriving meaning from received data.

[1514] "Action suggestions" are specific instructions for actions that a user should take, generated by a generative AI model based on the user's location information and past behavioral history.

[1515] "Notification" is the process of sending the action suggestions generated by the server to the user's terminal to inform the user.

[1516] This invention is a system that suggests appropriate actions based on the user's location information. This system is realized using the user's device, a server, and a generative AI model.

[1517] First, the user's device is a device with GPS functionality, such as a smartphone or tablet. The device uses the GPS sensor to obtain the user's current location information. The obtained location information is expressed as latitude and longitude data.

[1518] The device then sends the acquired location information to the server using the HTTPS protocol to ensure data security. The device converts the location information into JSON format and sends it to the server via an HTTPS request.

[1519] The server analyzes the received location information. A generative AI model, such as a natural language processing model, is used for the analysis. The server identifies the area the user is in based on the location information and obtains risk information related to that area.

[1520] The server then uses the generative AI model to generate suggested actions based on the user's location. Specifically, the server inputs prompts into the generative AI model to generate appropriate actions. Examples of prompts include:

[1521] "The user is currently near the coast. A tsunami warning has been issued. Please suggest what action the user should take."

[1522] Based on this prompt, the generative AI model generates the suggested action, "Evacuate to the nearest high ground."

[1523] Finally, the server notifies the user's device of the generated action suggestions using a push notification service. The server converts the action suggestions into JSON format and sends them to the user's smartphone via the push notification service. The user's smartphone receives the notification and displays it on the screen.

[1524] In this way, the system can take into account the user's location and suggest appropriate actions in real time, allowing users to take prompt and appropriate action even in emergencies.

[1525] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1526] Step 1:

[1527] The user's device uses the GPS sensor to obtain current location information. The input is data from the GPS sensor, and the output is latitude and longitude information. Specifically, when a user enables location services on their smartphone, the device periodically obtains GPS data.

[1528] Step 2:

[1529] The device sends the acquired location information to the server. The input is latitude and longitude information, and the output is data sent to the server as an HTTPS request. Specifically, the device converts the location information into JSON format and sends it to the server using the HTTPS protocol.

[1530] Step 3:

[1531] The server analyzes the received location information. The input is the location information sent from the device, and the output is the result of identifying the area the user is in. Specifically, the server checks the location information against a database to see if the user is in an area where a tsunami warning has been issued.

[1532] Step 4:

[1533] The server uses a generative AI model to generate action suggestions based on the user's location. The input is the location and a prompt, and the output is an action suggestion. Specifically, the server inputs the following prompt into the generative AI model:

[1534] "The user is currently near the coast. A tsunami warning has been issued. Please suggest what action the user should take."

[1535] Based on this prompt, the generative AI model generates the suggested action, "Evacuate to the nearest high ground."

[1536] Step 5:

[1537] The server notifies the user's device of the generated action suggestions. The input is the action suggestions, and the output is data sent to the user's device as a push notification. Specifically, the server converts the action suggestions into JSON format and sends them to the user's smartphone via a push notification service. The user's smartphone receives the notification and displays it on the screen.

[1538] (Application example 2)

[1539] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1540] Conventional disaster evacuation systems provide uniform evacuation instructions without considering the user's location information, making it difficult to provide optimal evacuation actions for each individual user. Furthermore, they lack the functionality to present evacuation routes or notify emergency contacts, resulting in insufficient comprehensive support to ensure the user's safety.

[1541] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving emergency alert emails such as those for earthquakes, tsunamis, and disaster evacuation information, means for analyzing the content of the received emergency alert emails and determining the type of disaster, AI means for generating actions to be taken based on the determined type of disaster, means for displaying the generated actions on the user's mobile device, means for acquiring the user's location information, means for generating an optimal evacuation route based on the acquired location information and displaying it on a map, and means for automatically notifying emergency contacts. This makes it possible to provide optimal evacuation actions based on the user's current location, suggest evacuation routes, and promptly notify emergency contacts.

[1542] An "emergency alert email" is an email that quickly notifies users of information about emergencies such as earthquakes, tsunamis, and disaster evacuation information.

[1543] "Type of disaster" refers to the classification of different disasters such as earthquakes, tsunamis, floods, and fires.

[1544] "AI means" is a technology that uses artificial intelligence to analyze data and generate optimal actions based on specific conditions.

[1545] "Mobile device" refers to a portable electronic device such as a smartphone or tablet.

[1546] "Location information" is latitude and longitude data that indicates the user's current location.

[1547] An "evacuation route" refers to the optimal route for a user to evacuate to a safe location.

[1548] "Emergency contact information" is information about people to contact in the event of a disaster or emergency.

[1549] A system for carrying out this invention has the following configuration. First, a server is provided with means for receiving emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The server also includes means for analyzing the content of the received emergency alert emails and determining the type of disaster. Furthermore, the server is equipped with AI means for generating actions to be taken based on the determined type of disaster.

[1550] The user's mobile device has a means for displaying the generated action. The mobile device also has a means for acquiring the user's location information, and includes a means for generating an optimal evacuation route based on the acquired location information and displaying it on a map. The mobile device also has a means for automatically notifying emergency contacts.

[1551] The following hardware and software are used to implement this system: A smartphone or tablet with GPS functionality is required for the hardware. The software uses a Python program, the Requests library (to obtain disaster information from the API), the Geopy library (to process location information), and the Folium library (to display maps).

[1552] When the server receives an emergency alert email, it analyzes its contents and determines the type of disaster. For example, if a tsunami warning is issued, the server generates an action for the user based on that information, such as "Please evacuate to the nearest high ground." The generated action is displayed on the user's mobile device.

[1553] The mobile device acquires the user's location information in real time and generates the optimal evacuation route based on that information. The evacuation route is displayed on a map, and the user can follow it to evacuate. An automatic notification function also quickly shares the user's current location and evacuation actions with emergency contacts.

[1554] For example, if a user is in Tokyo and a tsunami warning is issued, the server will notify the user, "Please evacuate to the nearest high ground," and the mobile device will display evacuation routes on a map. The following prompt sentence is used:

[1555] Example prompt sentence:

[1556] If the user's current location is Tokyo and a tsunami warning has been issued, generate an evacuation action to the nearest higher ground and display the evacuation route on a map.

[1557] In this way, a system can be realized that provides specific actions to ensure the safety of users.

[1558] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1559] Step 1:

[1560] The server receives emergency alert emails such as earthquake, tsunami, and disaster evacuation information. The input is the emergency alert email, and the output is the content of the received email. Specifically, the server obtains emails from an external emergency alert system.

[1561] Step 2:

[1562] The server analyzes the content of the received emergency alert email and determines the type of disaster. The input is the content of the received email, and the output is the type of disaster (e.g., earthquake, tsunami). Specifically, the server analyzes the text of the email, extracts keywords, and identifies the type of disaster.

[1563] Step 3:

[1564] The server generates the appropriate action to be taken based on the type of disaster it has determined. The input is the type of disaster, and the output is the appropriate action to be taken (e.g., "Please evacuate to the nearest high ground"). Specifically, the server uses a generative AI model to generate the optimal action according to the type of disaster.

[1565] Step 4:

[1566] The server sends the generated action to the user's mobile device. The input is the action to be taken, and the output is an action message to be displayed on the user's mobile device. Specifically, the server pushes the action message to the mobile device.

[1567] Step 5:

[1568] The device obtains the user's location information in real time. The input is GPS data, and the output is the user's current location (latitude and longitude). Specifically, the device obtains location information using its built-in GPS function.

[1569] Step 6:

[1570] The device generates the optimal evacuation route based on the acquired location information and displays it on a map. The input is the user's current location and evacuation location information, and the output is the evacuation route displayed on the map. Specifically, the device uses a map display library (Folium) to draw the evacuation route.

[1571] Step 7:

[1572] The device automatically notifies emergency contacts. The input is the user's current location and the generated action, and the output is a notification message sent to the emergency contacts. Specifically, the device sends the notification message to the emergency contacts via SMS or email.

[1573] In this way, a system can be realized that provides specific actions to ensure the safety of users.

[1574] Example 3

[1575] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1576] The main function of conventional disaster evacuation systems is to receive emergency alert emails and generate response actions based on their content analysis, but they lack the ability to respond individually to users' past behavioral history and location information. This means that optimal evacuation routes and instructions for each user are not provided, which can reduce the efficiency and safety of evacuation. Furthermore, there is a lack of a mechanism for using user feedback as learning data for the next system, making it difficult to improve the accuracy of the system.

[1577] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1578] In this invention, the server includes means for collecting past behavioral history data of users and storing it in a database, means for preprocessing the stored behavioral history data and training it using a generative AI model, and means for predicting the user's future behavior based on the training results and generating appropriate behavior. This makes it possible to provide optimal evacuation routes and behavioral instructions for each user, improving the efficiency and safety of evacuation. Furthermore, by using user feedback as training data for the next round, the accuracy of the system can be continuously improved.

[1579] An "emergency alert email" is an email sent to quickly notify users of emergency information such as earthquakes, tsunamis, and disaster evacuation information.

[1580] "Analysis" is the process of analyzing the contents of the received emergency alert email and determining the type of disaster and its impact.

[1581] "Artificial intelligence means" refers to machine learning algorithms and models that generate optimal response actions based on a user's behavioral history and location information.

[1582] A "mobile terminal" is an electronic device that a user carries and uses, such as a smartphone or tablet.

[1583] "Behavioral history data" refers to data about the actions a user has taken and the places they have visited in the past.

[1584] A "database" is a system for storing and managing collected behavioral history data.

[1585] "Preprocessing" refers to processes such as filling in missing values, normalizing data, and removing outliers that are carried out to improve the quality of collected data.

[1586] A "generative AI model" is a model trained using a machine learning framework and used to predict user behavior and generate appropriate countermeasures.

[1587] "Learning results" refer to user behavior patterns and predictive models obtained as a result of the generative AI model learning based on behavioral history data.

[1588] "Feedback" is the user's evaluation or opinion of the system's instructions or generated actions.

[1589] The present invention is a system that learns the user's past behavioral history and generates an action to be taken based on the learning results. Specific embodiments of this system will be described below.

[1590] Server Processing

[1591] The server collects the user's past behavioral history data and stores it in a database. Specifically, it collects GPS data, timestamps, and the type of behavior (e.g., evacuation drills, commuting, etc.) from the user's mobile device. The collected data is stored in a relational database such as MySQL or PostgreSQL.

[1592] The server then preprocesses the stored data, which includes imputing missing values, normalizing the data, and removing outliers, improving the quality of the data and increasing the accuracy of the generative AI model.

[1593] Using the preprocessed data, the server trains a generative AI model. This process uses machine learning frameworks such as TensorFlow and PyTorch. Recurrent neural networks (RNNs) and long short-term memory (LSTMs) are used as learning algorithms. Once training is complete, the server saves the learning results. Saved models are managed in formats such as HDF5 and Pickle.

[1594] Based on the learning results, the server predicts the user's future behavior and generates appropriate actions. For example, if the user is conducting an evacuation drill, the server generates the optimal evacuation route based on past data. This information is sent to the user's mobile device.

[1595] Terminal handling

[1596] The device receives the learning results and prediction data sent from the server. The received data is used as action instructions for the user. Specifically, evacuation routes and action instructions are displayed on a map. Map services such as Google Maps API are used to display the information in a visually easy-to-understand format.

[1597] The device accepts input from the user. For example, if the user instructs the device to "display evacuation routes," the device will display the appropriate information in accordance with the instruction.

[1598] User Action

[1599] The user checks the action instructions displayed on the device. For example, during an evacuation drill, the user checks the evacuation route displayed on the device. The user acts based on the information provided by the device. Specifically, the user follows the displayed evacuation route and heads to a shelter.

[1600] After taking action, the user provides feedback. For example, they can input whether the evacuation route was appropriate or not. This feedback is sent to the server and used as learning data for the next time.

[1601] Specific examples

[1602] For example, consider a case where a user preferentially instructs a route to a shelter that the user has visited in the past during evacuation drills. The server analyzes the user's past evacuation drill data and generates the optimal evacuation route. This information is sent to the user's device and is useful when the user evacuates.

[1603] Prompt Sentence Examples

[1604] "Generate optimal evacuation routes based on the user's past behavioral history. Use past evacuation drill data to prioritize routes to evacuation shelters."

[1605] By inputting this prompt sentence into the generative AI model, an optimal evacuation route is generated based on the user's behavior history. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1606] Step 1:

[1607] Data collection

[1608] The server collects GPS data, timestamps, and types of activity (e.g., evacuation drills, commuting, etc.) from the user's mobile device. The user's location information and activity history data are provided as input. The server receives this data and stores it in a database. Specifically, it periodically retrieves data from the mobile device and stores it in a relational database such as MySQL or PostgreSQL.

[1609] Step 2:

[1610] Data Preprocessing

[1611] The server preprocesses the stored behavioral history data. Raw data stored in the database is provided as input. Preprocessing includes missing value completion, data normalization, and outlier removal. This improves the quality of the data and increases the accuracy of the generative AI model. Specifically, the server uses Python's Pandas library to clean the data and perform the necessary preprocessing.

[1612] Step 3:

[1613] Model learning

[1614] The server trains a generative AI model using the preprocessed data. The preprocessed data is provided as input. Machine learning frameworks such as TensorFlow and PyTorch are used for training. Recurrent neural networks (RNNs) and long short-term memory (LSTMs) are applied as training algorithms. A trained model is obtained as output. Specifically, the data is fed into the model and the model parameters are updated every epoch.

[1615] Step 4:

[1616] Save Model

[1617] The server saves the model after training is complete. The trained model is provided as input. The model is saved in formats such as HDF5 or Pickle. The output is the saved model file. Specifically, the model is written to a file in the specified format and saved in storage.

[1618] Step 5:

[1619] Prediction and Action Generation

[1620] The server uses the stored model to predict the user's future behavior and generate appropriate actions. The input is the trained model and the user's current location information. The output is the generated action instructions. Specifically, the current data is input into the model, and the optimal action is generated based on the prediction results.

[1621] Step 6:

[1622] Data reception

[1623] The device receives the learning results and prediction data sent from the server. As input, action instruction data is provided from the server. As output, the received data is saved on the device. Specifically, the device receives data from the server and saves it in local storage.

[1624] Step 7:

[1625] Data Display

[1626] The device displays the received data to the user. The received action instruction data is provided as input. The output is information that is visually displayed to the user. Specifically, the device uses the Google Maps API to display evacuation routes on a map.

[1627] ...

Claims

[Claim 1] A means to receive emergency alert emails including disaster evacuation information for earthquakes, tsunamis, fires, floods, etc. A means for analyzing the content of the received emergency alert email using a generating AI model and determining the type of disaster; A means for obtaining user location information; A means for collecting user's past behavioral history data; a means for generating a prompt sentence for instructing the user to generate an action instruction based on the determined type of disaster and the acquired past action history data of the user; A means for generating an action instruction for the user to take using the generated prompt sentence and the generative AI model; means for displaying the generated action instructions on the user's mobile device; means for identifying an emotional state of a user using an emotion engine; a means for generating a prompt sentence instructing the user to correct an action instruction to be taken based on the identified emotional state of the user, the acquired location information of the user, and the acquired past action history data of the user; A means for modifying the instruction to be acted upon by the user using the generated prompt sentence and the generative AI model; means for displaying the generated modified action instructions on the user's mobile device; A system including:

Images