System

Abstract

To provide a system.SOLUTION: A system includes: means for receiving questions including personal information and anonymizing the personal information; means for transmitting an anonymized question to generative AI; means for receiving an answer from the generative AI and decoding the anonymous information; means for providing an interface that receives an input from a user; means including an emotion engine that analyzes emotions from text information of the user's question; means for transmitting to the generative AI the emotion information analyzed by the emotion engine; means with which the generative AI generates answers on the basis of the emotion information; and means for displaying to the user the generated answer.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] When using generative AI, it is important not to transmit personal or confidential information, but this is the first and biggest challenge for enterprise use. [Means for solving the problem]

[0005] To solve this problem, we developed an AI that operates in a local environment and anonymizes personal information in advance. Specifically, the system includes a means for receiving questions containing personal information and anonymizing the relevant personal information, a means for sending the anonymized questions to a generation AI, and a means for receiving answers from the generation AI and decrypting the relevant anonymous information. [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] In one embodiment of the present invention, an interface for accepting input from a user is provided as a means for receiving questions including personal information. This interface may be, for example, an input form on a web page or a text input field in a smartphone application.

[0029] "Example 2"

[0030] Next, as a means of anonymizing the received personal information, a function is provided to convert personal information based on specific rules, such as converting names to initials or addresses to area-level information.

[0031] "Example 3"

[0032] A communication module is provided as a means for transmitting the anonymized questions to the generation AI. This communication module converts the anonymized questions into a format that can be analyzed by the generation AI and transmits them to the generation AI.

[0033] "Example 4"

[0034] Finally, we will provide a mechanism to reverse the specific rules described above as a means of receiving the answers from the Generator AI and decrypting the relevant anonymous information, thereby returning the answers from the Generator AI in a form that can be understood by the user.

[0035] The processing flow of each embodiment will be described below.

[0036] "Example 1"

[0037] Step 1: A user enters a question into a form on a webpage, which may contain personal information.

[0038] Step 2: The system receives the question and anonymizes the personal information based on certain rules, for example, converting names to initials and addresses to locality level information.

[0039] Step 3: The anonymized question is sent to the generation AI, where a communication module converts the question into a format that the generation AI can parse.

[0040] Step 4: Receive the answer from the Generator AI and decrypt the corresponding anonymous information, specifically by applying the specific rules mentioned above in reverse to return the answer to a form that the user can understand.

[0041] Example 1

[0042] 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."

[0043] Conventional systems have had difficulty safely processing questions containing personal information and generating appropriate answers. Furthermore, the process of anonymizing and decrypting personal information is complex, resulting in delays in providing answers to users. Furthermore, there is a lack of efficient means for generating prompts suitable for generative AI models.

[0044] 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.

[0045] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for analyzing the anonymized question and generating a prompt sentence suitable for the generative AI model, means for sending the generated prompt sentence to the generative AI model, means for receiving an answer from the generative AI model and decrypting the relevant anonymous information, and means for displaying the decrypted answer to the user. This makes it possible to process personal information safely and provide the user with a prompt and appropriate answer.

[0046] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0047] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0048] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers based on input prompt sentences.

[0049] A "prompt sentence" is a sentence input to a generative AI model that appropriately expresses the user's question.

[0050] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0051] A "server" is a computer system that processes and stores data over a network.

[0052] A "user" is an individual or organization that uses the system to enter questions and receive answers.

[0053] This invention is a system that allows a user to input a question that includes personal information and provides an appropriate answer to the question using a generative AI model. Specific embodiments of this system are described below.

[0054] A user enters a question using a form on a web page or a text input field in a smartphone application, for example, "My name is Taro Tanaka. My address is Shinjuku-ku, Tokyo. How can I protect my personal information?"

[0055] The device sends the entered question to the server using the HTTPS protocol, with the data encrypted.

[0056] The server analyzes the received questions and anonymizes the personal information. Anonymization involves converting personal information based on specific rules. For example, "Taro Tanaka" could be "User A" and "Shinjuku Ward, Tokyo" could be "Area X."

[0057] The server then analyzes the anonymized question and generates a prompt suitable for the generative AI model, such as:

[0058] The user's name is User A and they live in Region X. How do you protect their personal information?

[0059] The server sends the generated prompt to a generative AI model, such as OpenAI's GPT-3 or GPT-4.

[0060] The generative AI model generates an answer based on the prompt it receives, for example, the following answer:

[0061] You can help protect your personal information by:

[0062] 1. Use strong passwords.

[0063] 2. Only share personal information with people you trust.

[0064] 3. Install security software and keep it up to date.

[0065] 4. Be careful when using public Wi-Fi.

[0066] The server decrypts the answer received from the generative AI model and restores the original personal information. For example, it restores "User A" to "Taro Tanaka" and "Area X" to "Shinjuku Ward, Tokyo."

[0067] Finally, the server displays the decrypted answers to the user, who can then check them on a web page or smartphone application.

[0068] This system allows users to safely process their personal information and receive prompt and appropriate answers.

[0069] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0070] Step 1:

[0071] The user enters a question.

[0072] A user types a question into a text field on a web page or smartphone application, for example, "My name is Taro Tanaka. My address is Shinjuku-ku, Tokyo. How can I protect my personal information?"

[0073] Input: User question

[0074] Output: Question data entered

[0075] Step 2:

[0076] The terminal sends a question to the server.

[0077] The device sends the entered question to a server using the HTTPS protocol, with the data encrypted.

[0078] Input: The entered question data

[0079] Output: The query data sent to the server

[0080] Step 3:

[0081] The server analyzes the question and anonymizes personal information.

[0082] The server analyzes the received questions and anonymizes personal information based on specific rules. For example, "Taro Tanaka" is "User A" and "Shinjuku Ward, Tokyo" is "Area X."

[0083] Input: Query data sent to the server

[0084] Output: Anonymized question data

[0085] Step 4:

[0086] The server analyzes the anonymized questions and generates prompts appropriate for the generative AI model.

[0087] The server analyzes the anonymized question and generates a prompt appropriate for the generative AI model, such as: "The user's name is User A and their address is in Region X. Please tell me how to protect their personal information."

[0088] Input: Anonymized question data

[0089] Output: Generated prompt statement

[0090] Step 5:

[0091] The server sends the generated prompt to the generative AI model.

[0092] The server sends the generated prompt to a generative AI model, which uses natural language processing technology.

[0093] Input: Generated prompt text

[0094] Output: The prompt sent to the generative AI model

[0095] Step 6:

[0096] A generative AI model generates the answer.

[0097] The generative AI model generates answers based on the prompt it receives, such as: "To protect your personal information, you can: 1. Use strong passwords. 2. Only share personal information with people you trust. 3. Install security software and keep it up to date. 4. Exercise caution when using public Wi-Fi."

[0098] Input: The prompt sent to the generative AI model

[0099] Output: The generated answer

[0100] Step 7:

[0101] The server receives the answer from the generative AI model and decrypts the anonymous information.

[0102] The server decrypts the answer received from the generative AI model and restores the original personal information. For example, it restores "User A" to "Taro Tanaka" and "Area X" to "Shinjuku Ward, Tokyo."

[0103] Input: Generated Answer

[0104] Output: Decoded answer

[0105] Step 8:

[0106] The server displays the decrypted answer to the user.

[0107] The server then displays the decrypted answers to the user, who can then check them on a web page or smartphone application.

[0108] Input: Decrypted answer

[0109] Output: The answer displayed to the user

[0110] (Application example 1)

[0111] 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."

[0112] There is a need to provide an environment where questions containing personal information can be handled safely and where users can enter questions with peace of mind. However, conventional systems have a high risk of personal information leakage and unauthorized access, making it difficult to adequately protect user privacy. Furthermore, even when providing answers using generative AI models, there is a lack of means to properly anonymize and decrypt users' personal information.

[0113] 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.

[0114] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means for decrypting the encrypted data and inputting it as a prompt sentence to the generation AI model, and means for obtaining an answer from the generation AI model and providing it to the user. This makes it possible to provide an appropriate answer using the generation AI model while safely handling the user's personal information.

[0115] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0116] A "question" is a sentence or phrase entered by a user to request information.

[0117] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0118] "Generative AI" is a system that uses artificial intelligence technology to generate appropriate answers to users' questions.

[0119] "Decryption" is the process of restoring anonymized information to its original form.

[0120] "Encryption" is the process of converting data using a specific algorithm to make it unreadable to third parties.

[0121] A "prompt sentence" is a sentence input to a generative AI model, and is an instruction sentence that enables the AI ​​to generate an appropriate answer.

[0122] An "answer" is information generated by the generative AI in response to a user's question.

[0123] "User" means an individual or entity that utilizes the System to enter questions and receive answers.

[0124] As an embodiment of the present invention, the following system is constructed.

[0125] System Configuration

[0126] The system includes a user terminal, a server, and a generative AI model. The user terminal is a device such as a smartphone or PC, and provides an interface for users to input questions. The server receives questions from users, anonymizes them, encrypts them, decrypts them, generates prompts for the generative AI model, and provides answers.

[0127] Hardware and software used

[0128] Hardware: smartphones, PCs, servers

[0129] Software: Python, OpenAI API, cryptography library

[0130] Data processing and calculation

[0131] 1. User Device:

[0132] The user enters a question through an interface on the device, which is an application that includes a text entry field.

[0133] 2. Server:

[0134] When the server receives a question from a user, it first anonymizes the personal information. The anonymized question is then encrypted and sent to a generative AI model. The generative AI model generates an appropriate answer based on the prompt. The generated answer is then decrypted by the server and provided to the user.

[0135] Specific examples

[0136] If a user types the question "How do I manage my passwords securely?", the system will:

[0137] 1. Enter the question "How do I manage my passwords securely?" on the user device.

[0138] 2. The server receives this question and anonymizes your personal information.

[0139] 3. The anonymized question is encrypted and sent as a prompt to the generative AI model.

[0140] 4. The generative AI model generates an answer based on the prompt.

[0141] 5. The server decrypts the generated answer and provides it to the user.

[0142] Prompt Sentence Examples

[0143] Users ask: How do I keep my passwords safe?

[0144] Please provide security advice.

[0145] This system allows users to receive appropriate answers using generative AI models while keeping their personal information safe.

[0146] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0147] Step 1:

[0148] Enter the question on the user's device.

[0149] Users enter their questions into a text entry field on a smartphone or PC application, and the entered questions are sent to the server.

[0150] Input: A question entered by the user (e.g., "How do I keep my passwords safe?")

[0151] Output: The question data sent to the server

[0152] Step 2:

[0153] The server receives the question and anonymizes the personal information.

[0154] The server anonymizes personal information from the received questions based on specific rules, for example, masking personal information such as name and address.

[0155] Input: Question data submitted by the user

[0156] Output: Anonymized question data

[0157] Step 3:

[0158] The server encrypts the anonymized question.

[0159] The server encrypts the anonymized question data using an encryption algorithm, ensuring the security of the data.

[0160] Input: Anonymized question data

[0161] Output: Encrypted question data

[0162] Step 4:

[0163] The server sends the encrypted question to the generative AI model.

[0164] The server decrypts the encrypted question data and sends it to the generative AI model as a prompt, which is an instruction for the generative AI model to generate an appropriate answer.

[0165] Input: Encrypted question data

[0166] Output: The prompt sent to the generative AI model

[0167] Step 5:

[0168] A generative AI model generates an answer based on the prompt.

[0169] The generative AI model analyzes the prompt and generates an appropriate answer, which is then sent to the server.

[0170] Input: The prompt sent to the generative AI model

[0171] Output: Answer data from the generative AI model

[0172] Step 6:

[0173] The server receives and decodes the answer from the generative AI model.

[0174] The server decrypts the answer data received from the generative AI model and returns it to its original format.

[0175] Input: Answer data from the generative AI model

[0176] Output: Decoded answer data

[0177] Step 7:

[0178] The server provides the decrypted answer to the user.

[0179] The server sends the decrypted answer data to the user's terminal and provides it to the user, who can then check the answer on the terminal.

[0180] Input: Decrypted answer data

[0181] Output: Answer data displayed on the user's device

[0182] Example 2

[0183] 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."

[0184] When inputting data including personal information into a generative AI model, it is necessary to obtain useful responses from the generative AI model while reducing the risk of personal information leakage. However, conventional systems have not been able to efficiently anonymize personal information and input it into the generative AI model, making it difficult to achieve both security and convenience.

[0185] 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.

[0186] In this invention, the server includes means for receiving data including personal information and anonymizing the personal information, means for transmitting the anonymized data to the generative AI model, means for receiving a response from the generative AI model and generating a prompt sentence, means for inputting the generated prompt sentence into the generative AI model, and means for receiving a response from the generative AI model. This makes it possible to efficiently obtain useful responses from the generative AI model while reducing the risk of personal information leaks.

[0187] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0188] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0189] A "generative AI model" is an artificial intelligence model that performs natural language processing and generates a response based on an input prompt.

[0190] A "prompt" is a document that serves as input to a generative AI model and contains instructions and information that enable the model to generate an appropriate response.

[0191] A "server" is a computer system that processes and stores data over a network.

[0192] "Data" means a collection of information, including personal information and other related information.

[0193] A "response" is the output result that a generative AI model generates based on a prompt sentence.

[0194] "Transmitting" is the act of moving data from one place to another.

[0195] "Receiving" is the act of receiving transmitted data.

[0196] "Conversion" is the process of changing the format or content of data into a different format or content.

[0197] MODE FOR CARRYING OUT THE INVENTION

[0198] This invention relates to a system that anonymizes data including personal information, generates prompt sentences to be input into a generative AI model, and receives responses from the generative AI model. Specific embodiments of this system are described below.

[0199] Hardware and software used

[0200] Hardware: High-performance server (e.g. general-purpose server)

[0201] Software: Python, Pandas, Numpy, generative AI models (e.g., natural language processing models)

[0202] System configuration

[0203] 1. Receipt of personal information

[0204] The server receives personal information submitted by the user, such as name and address information entered by the user into a web form.

[0205] Specific operation: The server receives an HTTP request and extracts personal information from the request body.

[0206] 2. Application of anonymization rules

[0207] The server applies anonymization rules to the personal information it receives, for example, converting names to initials and addresses to locality level information.

[0208] Specific operation: The server runs a Python script to convert the name "Taro Tanaka" to "TT" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[0209] 3. Save the converted data

[0210] The server stores the anonymized data in a database, which allows it to be retrieved later if needed.

[0211] Specific operation: The server saves the converted data in the SQL database using an INSERT statement.

[0212] 4. Prompt generation

[0213] The server generates prompts to input to the generative AI model, for example, prompts containing anonymized names and addresses.

[0214] What happens: The server uses Python's string manipulation functions to generate a prompt like this:

[0215] Name:TT

[0216] Address: Shinjuku Ward, Tokyo

[0217] Use this information to generate a user profile.

[0218] 5. Input to the generative AI model

[0219] The server inputs the generated prompt sentence into a generative AI model and receives a response from the model. For example, it inputs a prompt sentence using a natural language processing model and obtains the generated text.

[0220] Specific operation: The server calls the API of the generative AI model, sends a prompt sentence, and receives a response.

[0221] In this way, the server can anonymize personal information, generate prompts to input into the generative AI model, and receive responses from the model. This system makes it possible to efficiently obtain useful responses from the generative AI model while reducing the risk of personal information leaks.

[0222] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0223] Step 1:

[0224] The server receives the personal information sent by the user.

[0225] Input: Personal information such as name and address that a user enters into a web form.

[0226] Specific operation: The server receives an HTTP request and extracts personal information from the request body.

[0227] Output: Extracted personal information (e.g., name "Taro Tanaka", address "Nishi-Shinjuku, Shinjuku-ku, Tokyo").

[0228] Step 2:

[0229] The server applies anonymization rules to the received personal information.

[0230] Input: Personal information extracted in step 1.

[0231] Specific operation: The server runs a Python script to convert the name "Taro Tanaka" to "TT" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[0232] Output: Anonymized data (e.g., name "TT", address "Shinjuku-ku, Tokyo").

[0233] Step 3:

[0234] The server stores the anonymized data in a database.

[0235] Input: Anonymized data from step 2.

[0236] Specific operation: The server saves the converted data in the SQL database using an INSERT statement.

[0237] Output: Anonymized data stored in a database.

[0238] Step 4:

[0239] The server generates prompt sentences to input into the generative AI model.

[0240] Input: Anonymized data from step 2.

[0241] Specific operation: The server generates a prompt sentence using Python's string manipulation functions.

[0242] Output: Generated prompt text (e.g. "Name: TT Address: Shinjuku-ku, Tokyo Please generate a user profile based on this information.").

[0243] Step 5:

[0244] The server inputs the generated prompt sentence into the generative AI model and receives a response from the model.

[0245] Input: The prompt statement generated in step 4.

[0246] Specific operation: The server calls the API of the generative AI model, sends a prompt sentence, and receives a response.

[0247] Output: The response from the generative AI model (e.g., a generated user profile).

[0248] (Application example 2)

[0249] 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."

[0250] Safely handling data that includes personal information is extremely important from the perspective of privacy protection. However, transmitting personal information as is carries the risk of information leakage. In particular, when using generative AI models, anonymization is necessary because there is a possibility that personal information may be passed on to third parties. Furthermore, the accuracy of restoring anonymized information to its original form is also important. To solve these issues, a system is needed that anonymizes personal information based on specific rules and allows the safe use of generative AI models.

[0251] 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.

[0252] In this invention, the server includes means for receiving questions containing personal information and anonymizing the relevant personal information, means for sending the anonymized questions to the generation AI, means for receiving answers from the generation AI and decrypting the relevant anonymous information, means for converting names to initials, means for converting addresses to area-level information, and means for anonymizing other personal information based on specific rules. This makes it possible to safely anonymize personal information and reduce the risk of information leakage when using a generative AI model.

[0253] "Personal information" is information that can identify a specific individual.

[0254] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0255] "Generative AI" is a system that uses artificial intelligence technology to generate new data based on input data.

[0256] "Initials" is an abbreviation that refers to the first letter of each part of a name.

[0257] "Area-level information" is information that indicates a wide area without including detailed address information.

[0258] "Decryption" is the process of restoring anonymized information to its original form.

[0259] "Specific Rules" are the set of rules used to anonymize or decrypt personal information.

[0260] A "question" is part of the data input to the generating AI and is the information used to obtain an answer.

[0261] An "answer" is the output data that the generative AI generates based on the question.

[0262] A system for implementing this invention includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to a generation AI, means for receiving a response from the generation AI and decrypting the relevant anonymous information, means for converting names to initials, means for converting addresses to area-level information, and means for anonymizing other personal information based on specific rules.

[0263] System configuration

[0264] The server receives a question containing personal information entered by the user. The received personal information is anonymized based on specific rules. For example, names are converted to initials, and addresses are converted to area-level information. The anonymized question is sent to the generation AI. The generation AI generates an answer based on the question and sends the answer back to the server. The server receives the answer from the generation AI and decrypts the anonymized information back to the original information.

[0265] Hardware and software used

[0266] The system uses mobile devices such as smartphones and tablets, and the software is implemented using Python, including functions for anonymizing personal information and an API for communicating with the generative AI model.

[0267] Specific examples

[0268] For example, if a user enters the name "Yamada Taro" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo," the server converts this to "YT" and "Tokyo." The converted information is sent to the generation AI, which generates an answer to the question. The generated answer is sent back to the server, which decrypts the anonymized information to its original form.

[0269] Prompt Sentence Examples

[0270] An example of a prompt to input to a generative AI model is as follows:

[0271] Create an application that anonymizes personal information entered by users, converting names to initials and addresses to locality level information. Implement it using Python.

[0272] In this way, it is possible to safely anonymize personal information and reduce the risk of information leakage when using generative AI models.

[0273] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0274] Step 1:

[0275] The user inputs a question containing personal information into the terminal.

[0276] Input: The name, address, and question entered by the user.

[0277] Output: Question data including personal information entered into the terminal.

[0278] Specific operation: The user enters the name "Yamada Taro," the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo," and the question into the input form on the terminal.

[0279] Step 2:

[0280] The device anonymizes your personal information.

[0281] Input: Question data, including personal information entered by the user.

[0282] Output: Anonymized question data.

[0283] Specific operations: The device will convert your name to "YT" and your address to "Tokyo." Other personal information will also be anonymized based on specific rules.

[0284] Step 3:

[0285] The device sends the anonymized question to the server.

[0286] Input: Anonymized question data.

[0287] Output: Anonymized question data sent to the server.

[0288] Specific operation: The terminal transmits data over the network to send the anonymized question data to the server.

[0289] Step 4:

[0290] The server sends the anonymized questions to the generation AI.

[0291] Input: Anonymized question data received by the server.

[0292] Output: Anonymized question data sent to the generative AI.

[0293] Specific operation: The server calls the API to send the anonymized question data to the generative AI model.

[0294] Step 5:

[0295] Generative AI generates answers to questions.

[0296] Input: Anonymized question data sent to the generative AI.

[0297] Output: Answer data generated by the generation AI.

[0298] How it works: The generative AI model analyzes anonymized question data and generates appropriate answers.

[0299] Step 6:

[0300] The server receives the answer from the generation AI.

[0301] Input: Answer data generated by the generation AI.

[0302] Output: The response data received by the server.

[0303] Specific operation: The server obtains data through an API to receive answer data from the generative AI model.

[0304] Step 7:

[0305] The server decrypts the anonymized information.

[0306] Input: The answer data received from the generation AI.

[0307] Output: The decoded answer data.

[0308] Specific operation: The server restores the anonymized information to its original form based on specific rules. For example, it decrypts "YT" into "Yamada Taro" and "Tokyo" into "Tokyo, Shinjuku Ward, Nishi-Shinjuku."

[0309] Step 8:

[0310] The server sends the decrypted response to the terminal.

[0311] Input: Decoded answer data.

[0312] Output: The decrypted answer data sent to the device.

[0313] Specific operation: The server transmits the decrypted response data to the terminal through the network.

[0314] Step 9:

[0315] The terminal displays the decrypted answer to the user.

[0316] Input: The decrypted response data received by the terminal.

[0317] Output: The decoded answer displayed to the user.

[0318] Specific operation: The terminal displays the decrypted answer data on the screen to show it to the user.

[0319] Example 3

[0320] Next, a description will be given of a third embodiment of the third embodiment. 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."

[0321] Conventional systems have the problem of insufficient privacy protection because they send queries containing personal information directly to the generative AI model. Furthermore, there is also the issue that the responses from the generative AI model are difficult for users to understand if they remain anonymized. Furthermore, the process of converting the data into a format that the generative AI model can analyze is complex, requiring efficient data processing.

[0322] 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.

[0323] In this invention, the server includes means for receiving an inquiry including personal information and anonymizing the relevant personal information, means for sending the anonymized inquiry to the generative AI model, means for receiving a response from the generative AI model and decrypting the relevant anonymous information, a communication module for converting the response into a format that can be analyzed by the generative AI model, and means for providing the decrypted response to the user. This makes it possible to provide a response in a format that is easy for the user to understand while ensuring the protection of personal information.

[0324] "Personal information" is information that can identify a specific individual.

[0325] An "inquiry" is a question or request made by a user to the system.

[0326] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0327] A "generative AI model" is an algorithm or system that uses artificial intelligence to analyze input data and generate an appropriate response.

[0328] A "communications module" is hardware or software for transmitting and receiving data.

[0329] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0330] "User" means any individual or entity that uses the System.

[0331] A "server" is a computer system that processes and stores data on a network.

[0332] A "response" is the answer or information provided by a generative AI model as a result of its analysis.

[0333] MODE FOR CARRYING OUT THE INVENTION

[0334] This invention is a system that anonymizes inquiries that include personal information, sends them to a generative AI model, and decrypts the responses from the generative AI model and provides them to the user. Specific embodiments of this system are described below.

[0335] Hardware and software used

[0336] Hardware: Servers, communication modules, user devices (PCs, smartphones, etc.)

[0337] Software: Generative AI models (e.g., OpenAI GPT-4), programs implementing specific rules for anonymization and decryption

[0338] System Operation

[0339] 1. The user enters a question

[0340] A user uses a terminal to input a question into the system's interface, for example, "My name is Taro Tanaka. How should I learn programming?"

[0341] 2. The server receives the query

[0342] The server receives the questions sent by the user and stores them in their original format.

[0343] 3. The server anonymizes the question

[0344] The server anonymizes personal information and specific identifying information contained in the question based on specific rules, for example, converting "My name is Tanaka Taro" into "My name is [anonymous]."

[0345] 4. The server sends the anonymized questions to the generating AI via the communication module.

[0346] The server sends the anonymized questions to the Generator AI using a communication module, which converts the questions into a format that the Generator AI can parse.

[0347] 5. Generative AI analyzes the question and generates an answer

[0348] The generative AI analyzes the question it receives and generates an appropriate answer. For example, in response to the question, "How should I learn programming?", it generates the answer, "To learn programming, you should first choose a basic language and then proceed with your studies using online courses and books."

[0349] 6. The server receives the answer from the generation AI

[0350] The server receives the answers sent by the generation AI and stores them in their original format.

[0351] 7. The server decrypts the response

[0352] The server decrypts the anonymous information contained in the answer from the AI ​​generator based on specific rules. For example, it converts the answer "For [anonymous], in order to learn programming..." into "For Taro Tanaka, in order to learn programming..."

[0353] 8. The server provides the decrypted answer to the user

[0354] The server sends the decrypted answer to the user's terminal, where the user can check the answer through the terminal interface.

[0355] Specific examples

[0356] Example 1: A user submits the question "My name is Taro Tanaka. How can I learn programming?"

[0357] 1. The user uses the terminal to enter a question and clicks the submit button.

[0358] 2. The server receives the question, anonymizes it, and converts it into "My name is [Anonymous]. How can I learn programming?"

[0359] 3. Send the anonymized question to the generation AI via the communication module.

[0360] 4. In response to the question, "How should I learn programming?", the generative AI generates the answer, "To learn programming, it's a good idea to first choose a basic language and then progress through online courses and books."

[0361] 5. The server receives the answer from the generating AI and provides it to the user as is, since there is no need to decrypt the anonymous information.

[0362] Example prompt sentence:

[0363] User asks: "My name is Taro Tanaka. How can I learn programming?"

[0364] Anonymized question: "My name is [Anonymous]. How do I learn to code?"

[0365] Prompt for generative AI: "How do I learn to code?"

[0366] Generative AI's answer: "To learn programming, start by choosing a basic language and then work your way up using online courses and books."

[0367] In this way, the system anonymizes the user's question, sends it to the generation AI, and decrypts the answer from the generation AI and provides it to the user. The flow of the identification process in Example 3 will be described with reference to FIG.

[0368] Step 1:

[0369] The user enters a question

[0370] Input: The user uses the terminal to type a question, for example, "My name is Taro Tanaka. How should I learn programming?"

[0371] Specific actions: The user enters a question into the device interface and clicks the send button.

[0372] Output: The entered question is sent from the terminal to the server.

[0373] Step 2:

[0374] The server receives the query

[0375] Input: The question sent from the terminal.

[0376] Specific operation: The server receives the question sent by the user. The received question is saved in its original format.

[0377] Output: The received questions are saved on the server.

[0378] Step 3:

[0379] The server anonymizes the question

[0380] Input: A question stored on the server.

[0381] Specific operation: The server anonymizes personal information and specific identifying information included in the question based on specific rules. For example, it converts "My name is Tanaka Taro" into "My name is [anonymous]."

[0382] Output: Anonymized questions are generated.

[0383] Step 4:

[0384] The server sends the anonymized questions to the generating AI via a communication module.

[0385] Input: Anonymized question.

[0386] Specific operation: The server sends the anonymized question to the generation AI using a communication module, which converts the question into a format that the generation AI can analyze.

[0387] Output: The question is sent to the Generator AI in a format that the Generator AI can parse.

[0388] Step 5:

[0389] Generative AI analyzes the question and generates an answer

[0390] Input: Anonymized question sent to the generative AI.

[0391] How it works: Generative AI analyzes the questions it receives and generates appropriate answers. For example, in response to the question, "How should I learn programming?", it generates the answer, "To learn programming, first choose a basic language and then study using online courses and books."

[0392] Output: The generated answer is sent to the server.

[0393] Step 6:

[0394] The server receives the answer from the generation AI.

[0395] Input: The answer sent by the generation AI.

[0396] Specific operation: The server receives the answer sent from the generation AI. The received answer is saved in its original format.

[0397] Output: The received answer is saved on the server.

[0398] Step 7:

[0399] The server decrypts the response

[0400] Input: The answer sent by the generation AI.

[0401] Specific operation: The server decrypts the anonymous information included in the answer from the generation AI based on specific rules. For example, the answer "For [anonymous], in order to learn programming..." is converted into "For Taro Tanaka, in order to learn programming..."

[0402] Output: The decoded answer is generated.

[0403] Step 8:

[0404] The server provides the decrypted answer to the user

[0405] Input: The decrypted answer.

[0406] Specific operation: The server sends the decrypted answer to the user's device, where the user can check the answer through the device interface.

[0407] Output: The decrypted answer is displayed on the user's terminal.

[0408] (Application example 3)

[0409] 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."

[0410] In conventional systems, when sending questions containing personal information to the AI ​​generator, there was a high risk of personal information being leaked, and user privacy was not protected. Another issue was the difficulty of accurately decrypting anonymized information when returning the AI ​​generator's answers to a format that users could understand. Furthermore, there was a lack of efficient ways to perform these processes on smartphones.

[0411] 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.

[0412] In this invention, the server includes means for receiving questions that include personal information and anonymizing the relevant personal information, means for sending the anonymized questions to the generation AI, means for receiving answers from the generation AI and decrypting the relevant anonymous information, and means, as an application installed on a smartphone, for sending the anonymized questions to the generation AI and decrypting the answers from the generation AI and providing them to the user. This makes it possible to efficiently realize a question-answering system that uses the generation AI on a smartphone while protecting user privacy.

[0413] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0414] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0415] "Generative AI" is an artificial intelligence model that generates natural language answers based on given input data.

[0416] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0417] A "smartphone" is a type of mobile phone, a multi-function device that can connect to the Internet and install applications.

[0418] An "application" is software installed on a smartphone to provide specific functions or services.

[0419] A "question" is text data that a user enters to request information or advice from the generated AI.

[0420] An "answer" is text data generated by the generation AI in response to a question.

[0421] A "user" is an individual who uses the system to submit questions to the generating AI and receive answers.

[0422] As an embodiment of the present invention, the application "SecureQuery" installed on a smartphone is used. A specific implementation method thereof will be described below.

[0423] First, a user launches the SecureQuery application on their smartphone. The application provides an interface for entering a question, including personal information. Once the user enters the question, the application anonymizes it.

[0424] The anonymization process involves converting personal information based on specific rules. For example, personal information such as usernames and addresses are replaced with generic terms like "anonymous" or "private." These anonymized questions are then converted into a format that can be analyzed by the generative AI.

[0425] The application then sends the anonymized question to a Generative AI, which generates a natural language answer based on the given question. This Generative AI is located on a server in the cloud and accessed via the internet.

[0426] The answer from the AI ​​generator is sent back to the application, which receives the answer and performs a decryption process to restore the anonymized information to its original form based on specific rules. This allows the user to receive the answer from the AI ​​generator in a format that they can understand.

[0427] The system is implemented using the following hardware and software:

[0428] Hardware: Smartphone

[0429] Software: Python, requests library, JSON

[0430] The specific data processing and calculation flow is as follows:

[0431] 1. Anonymization processing: Personal information from questions entered by the user is anonymized within the application based on specific rules.

[0432] 2. Send to Generator: Send the anonymized question to the Generator endpoint, which analyzes the question and generates an answer.

[0433] 3. Decryption process: Receives the answer from the generation AI and restores the anonymized information to its original form based on specific rules.

[0434] For example, if a user enters the question, "What should I do if my username password is leaked?", this question is anonymized as, "What should I do if my anonymous password is leaked?" The Generative AI generates an appropriate answer to this question, and the application decrypts the answer and provides it to the user.

[0435] Example prompt sentence:

[0436] "What should I do if my username password is compromised?"

[0437] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0438] Step 1:

[0439] The user launches the "SecureQuery" application on their smartphone and enters a question.

[0440] Input: The question text entered by the user (e.g., "What should I do if my username's password is leaked?")

[0441] Output: The question text entered

[0442] Specific behavior: The user enters a question into the application's input field and presses the "Submit" button.

[0443] Step 2:

[0444] The device anonymizes the entered question.

[0445] Input: The question text entered by the user

[0446] Output: Anonymized question text (e.g., "What should I do if my anonymous password is leaked?")

[0447] What it does: An anonymization module within the application replaces personal information with generic terms based on specific rules.

[0448] Step 3:

[0449] The device sends anonymized questions to the generation AI.

[0450] Input: Anonymized question text

[0451] Output: Question text sent to the generation AI

[0452] What it does: Your application makes an HTTP request and sends an anonymized question to the Generative AI endpoint.

[0453] Step 4:

[0454] The server uses generative AI to generate answers to questions.

[0455] Input: Anonymized question text

[0456] Output: Generated answer text

[0457] How it works: The generation AI analyzes the question and generates an appropriate answer, which is then sent back to the device from the server.

[0458] Step 5:

[0459] The device receives the answer from the generating AI and decodes it.

[0460] Input: Answer text returned by the generation AI (e.g., "If your anonymous password is leaked, please change it immediately.")

[0461] Output: Decrypted answer text (e.g., "If your username password has been leaked, please change it immediately.")

[0462] What it does: A decryption module within the application converts the anonymized information back into the original information based on specific rules.

[0463] Step 6:

[0464] The terminal displays the decrypted answer to the user.

[0465] Input: Decrypted answer text

[0466] Output: The answer text that is displayed to the user

[0467] What it does: The application displays the decrypted answer on the screen for the user to review.

[0468] 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.

[0469] "Example 1"

[0470] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[0471] "Example 2"

[0472] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[0473] "Example 3"

[0474] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[0475] The processing flow of each embodiment will be described below.

[0476] "Example 1"

[0477] Step 1: The user enters a question, which is sent to the system as text information.

[0478] Step 2: The emotion engine analyzes the emotion from this text information. For example, it recognizes the emotion "sadness" from the question "I am very sad."

[0479] Step 3: The emotion engine sends the analyzed emotion information to the generation AI.

[0480] Step 4: The generative AI generates an answer based on the emotional information received from the emotion engine. For example, if the generative AI recognizes the emotion "sadness," it generates an answer that comforts the user.

[0481] Step 5: The generation AI provides the generated answer to the user, which allows it to provide a more appropriate answer that reflects the user's emotions.

[0482] Example 1

[0483] 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."

[0484] Conventional systems have had difficulty providing appropriate answers that reflect the user's emotions. Furthermore, they lacked the means to safely process questions containing personal information and to anonymize and decrypt them. This resulted in low user satisfaction and created privacy issues.

[0485] 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.

[0486] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means for providing an interface for accepting input from a user, means for including an emotion engine for analyzing emotions from text information of the user's question, means for sending the emotion information analyzed by the emotion engine to the generation AI, means for the generation AI to generate an answer based on the emotion information, and means for displaying the generated answer to the user. This makes it possible to provide an appropriate answer corresponding to the user's emotion, and to enable safe processing of personal information and privacy protection.

[0487] "Personal information" is information that can identify a specific individual.

[0488] A "question" is an inquiry entered by a user.

[0489] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0490] "Generative AI" is a system that uses artificial intelligence technology to generate answers to users' questions.

[0491] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0492] An "interface" is a means by which a user inputs information into a system, and includes input forms on web pages and text input fields in smartphone applications.

[0493] The "emotion engine" is a system that analyzes emotions from the text information of users' questions.

[0494] "Emotion information" is information about the user's emotions obtained as a result of analysis by the emotion engine.

[0495] An "answer" is the response that the generation AI generates in response to a user's question.

[0496] "Display" refers to the process of visually presenting the generated answer to the user.

[0497] MODE FOR CARRYING OUT THE INVENTION

[0498] The present invention is a system for receiving a question from a user and generating an appropriate answer based on emotions to the question. A specific embodiment of this system will be described below.

[0499] Hardware and Software Configuration

[0500] 1. User Device

[0501] Users enter questions using input forms on web pages or text input fields in smartphone applications. These interfaces are the means by which users provide input to the system.

[0502] 2. Server

[0503] The server plays a central role in processing queries received from users. It has the following functions:

[0504] Anonymization function: Personal information is converted based on specific rules so that it cannot be used to identify a specific individual.

[0505] Emotion engine: Analyzes emotions from the text information of the user's question. The emotion engine uses natural language processing (NLP) technology to analyze text data and recognize the user's emotions.

[0506] Generative AI: Generates answers to user questions based on the emotional information analyzed by the emotion engine. Generative AI uses advanced language models such as GPT-4.

[0507] Decryption function: Receives the answer from the generating AI and decrypts the corresponding anonymous information.

[0508] Display function: displays the generated answer to the user.

[0509] Data processing and calculation

[0510] 1. User Input

[0511] The user enters "I am very sad" into the text input field of the smartphone application. After completing the input, the user presses the send button.

[0512] 2. Sending input data

[0513] The terminal sends the text data entered by the user to the server using the HTTPS protocol. Since the data is encrypted, there is a low risk of it being intercepted by a third party.

[0514] 3. Emotion analysis using an emotion engine

[0515] The server passes the received text data to the emotion engine, which uses NLP technology to recognize the emotion "sadness" from the text "I am very sad." The emotion engine uses an emotion analysis library implemented in Python.

[0516] 4. Sending emotional information to the AI

[0517] The server sends the emotion information recognized by the emotion engine, "sadness," to the generation AI in JSON format, for example, {"emotion": "sadness"}.

[0518] 5. Generating emotion-aware answers

[0519] The generative AI generates an answer based on the emotional information it receives. For example, if it receives the emotional information "sadness," the generative AI will generate the answer "I understand how you feel. Let me know if there's anything I can help you with." The generative AI uses OpenAI's GPT-4 model.

[0520] 6. Submitting and Viewing Your Answers

[0521] The server sends the answer generated by the AI ​​to the device. The device displays the received answer to the user. The user sees a message on their smartphone screen saying, "We understand how you feel. Please let us know if there's anything we can do to help."

[0522] Specific examples

[0523] For example, if a user enters "I am very sad," the emotion engine will recognize the emotion "sadness" from this text, and the generation AI will generate a response that will comfort the user. An example of a specific prompt sentence is, "If a user enters 'I am very sad,' the emotion engine will recognize the emotion "sadness" from this text, and the generation AI will generate a response that will comfort the user."

[0524] In this way, a system is realized that provides appropriate answers that correspond to the user's emotions.

[0525] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0526] Step 1:

[0527] A user inputs a question into a text input field of a smartphone application. For example, the user inputs "I am very sad" and presses the send button. The input data is in text format.

[0528] Step 2:

[0529] The terminal sends the text data entered by the user to the server using the HTTPS protocol. Since the input data is encrypted, there is a low risk of it being intercepted by a third party. The output is encrypted text data.

[0530] Step 3:

[0531] The server passes the received text data to the emotion engine. The input is encrypted text data, which is decrypted to obtain text information. The emotion engine uses natural language processing (NLP) technology to analyze the text data and recognize the user's emotion. For example, it recognizes the emotion "sadness" from the text "I am very sad." The output is emotional information.

[0532] Step 4:

[0533] The server sends the emotion information recognized by the emotion engine in JSON format to the generation AI. The input is emotion information, which is converted into JSON format. For example, it is sent in the format {"emotion": "sadness"}. The output is emotion information in JSON format.

[0534] Step 5:

[0535] The generative AI generates an answer based on the emotional information it receives. The input is emotional information in JSON format, which is analyzed to generate an appropriate answer. For example, if the generative AI receives the emotional information "sadness," it generates the answer "I understand how you feel. Please let me know if there's anything I can do to help." The output is a text answer.

[0536] Step 6:

[0537] The server sends the answer generated by the generation AI to the terminal. The input is a text answer, which is encrypted and sent to the terminal. The output is an encrypted text answer.

[0538] Step 7:

[0539] The device decrypts the received answer and displays it to the user. The input is the encrypted text answer, which is decrypted and displayed to the user. The user sees a message on their smartphone screen saying, "We understand how you feel. Let us know if there's anything we can do to help." The output is the text answer displayed to the user.

[0540] (Application example 1)

[0541] 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."

[0542] Conventional systems often struggle to provide appropriate answers to user questions. In particular, they generate answers without considering the user's emotions, making it difficult to alleviate the user's anxiety or sadness. Furthermore, they lack sufficient anonymization and decryption methods to safely process questions containing personal information, which poses a privacy problem.

[0543] The identification process by the identification 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 a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing emotions from the user's input text, means for sending the analyzed emotion information to the generation AI and generating an answer based on the emotion, and means for displaying the generated answer to the user. This makes it possible to provide an appropriate answer that corresponds to the user's emotions and process the question safely while protecting privacy.

[0544] "Personal information" is information that can identify a specific individual.

[0545] A "question" is an inquiry that a user inputs to the system.

[0546] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0547] "Generative AI" is artificial intelligence that generates appropriate answers to questions from users.

[0548] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0549] An "emotion engine" is software or hardware for analyzing emotions from user input text.

[0550] "Emotion information" is data on the user's emotions analyzed by the emotion engine.

[0551] "Display" refers to the process of visually presenting the generated answer to the user.

[0552] A system for implementing this invention is configured as follows: First, a user inputs a question using a device such as a smartphone. Since the input question may contain personal information, the server anonymizes it. The anonymized question is sent to a generation AI, which then generates an appropriate answer.

[0553] Next, an emotion engine is used to analyze emotions from the user's input text. The emotion engine analyzes the user's text information and generates emotion information. This emotion information is sent to the generation AI, which then generates an answer based on the emotion. The generated answer is displayed to the user by the server.

[0554] Specifically, the following hardware and software are used:

[0555] Hardware: Smartphones, servers

[0556] Software: Python, OpenAI API, TextBlob

[0557] The server receives input from the user and analyzes the emotion using TextBlob. The analyzed emotion information is sent to the generation AI using OpenAI's API. The generation AI generates an appropriate answer based on the emotion information and returns the answer to the server. The server displays the generated answer to the user.

[0558] For example, if a user types, "There may be a suspicious person in my house," the emotion engine recognizes the emotion of "anxiety." Based on this emotional information, the generative AI generates the answer, "Please contact the police immediately. Please evacuate to a safe place and securely close the doors and windows."

[0559] An example of a prompt is as follows:

[0560] The user is feeling anxious. Please generate appropriate answers to the following questions: There might be a suspicious person in my house.

[0561] In this way, it is possible to provide appropriate answers that correspond to the user's feelings and process questions safely while protecting privacy.

[0562] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0563] Step 1:

[0564] A user inputs a question using a terminal. The input question is sent to the server in text format. The input data is the user's question text.

[0565] Step 2:

[0566] The server anonymizes personal information from the received question text. For anonymization, the server converts the personal information based on specific rules so that a specific individual cannot be identified. The input data is the user's question text, and the output data is the anonymized question text.

[0567] Step 3:

[0568] The server sends the anonymized question text to the generation AI, which generates an appropriate answer based on the received question text. The input data is the anonymized question text, and the output data is the generated answer text.

[0569] Step 4:

[0570] The server uses an emotion engine to analyze emotions from the user's input text. The emotion engine uses software such as TextBlob to analyze the text information and generate emotion information. The input data is the user's question text, and the output data is the analyzed emotion information.

[0571] Step 5:

[0572] The server sends the analyzed emotional information to the generation AI, which then generates an answer based on the emotion. The generation AI then generates an appropriate answer taking the emotional information into account. The input data is the analyzed emotional information and anonymized question text, and the output data is an answer text based on the emotion.

[0573] Step 6:

[0574] The server displays the generated answer text to the user, who can visually check the generated answer through the terminal. The input data is the generated answer text, and the output data is the answer displayed to the user.

[0575] In this way, it is possible to provide an appropriate answer that corresponds to the user's feelings and process questions safely while protecting privacy.

[0576] Example 2

[0577] 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."

[0578] Conventional systems do not adequately anonymize and decrypt data, including personal information, resulting in incomplete privacy protection. They also have difficulty generating appropriate answers that take user sentiment into account. This leads to problems such as a poor user experience and a loss of system reliability.

[0579] 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.

[0580] In this invention, the server includes means for receiving data including personal information and anonymizing the personal information, means for transmitting the anonymized data to the generative AI model, means for receiving a question from a user and analyzing the corresponding emotion, means for transmitting the analyzed emotion information to the generative AI model, means for receiving an answer from the generative AI model and decrypting the corresponding anonymous information, and means for transmitting the generated answer to the user. This makes it possible to provide an appropriate answer corresponding to the user's emotion while ensuring the protection of personal information.

[0581] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0582] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0583] A "generative AI model" is an algorithm or system that uses artificial intelligence technology to analyze data and generate appropriate answers or information.

[0584] "Emotion analysis" is a process of identifying emotions from text information entered by a user and extracting that emotion information.

[0585] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0586] A "user" is an individual or group that uses the system and enters questions and personal information.

[0587] A "server" is a computer that is the center of a system and is a device that receives, processes, and transmits data.

[0588] "Data" is a collection of information, expressed in the form of text, numbers, images, etc.

[0589] A "question" is an inquiry or request that a user inputs to the system.

[0590] An "answer" is a response or piece of information that a generative AI model generates in response to a user's question.

[0591] The present invention relates to a system that anonymizes data including personal information, analyzes user emotions, and generates appropriate answers. A specific embodiment of this system is described below.

[0592] The server receives data including personal information. For example, a user inputs personal information such as "Yamada Taro, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo" using a terminal. The terminal then transmits this information to the server.

[0593] The server anonymizes the received personal information. Specifically, it converts names into initials and addresses into area-level information. For example, it converts "Yamada Taro" into "YT" and "1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo" into "Shinjuku-ku, Tokyo." This anonymization process is performed based on specific rules.

[0594] Next, the user uses the terminal to input a question, for example, "I am very sad," which is then sent to the server.

[0595] The server uses an emotion engine to analyze emotions from the text information of the user's question. The emotion engine recognizes the emotion "sadness" from the input text. This emotion information is sent to the generative AI model.

[0596] The generative AI model generates a response based on the emotional information it receives. For example, if it recognizes the emotion "sadness," it will generate the response "That's very painful. Please let me know if there's anything I can do to help."

[0597] The server generates a response and sends it to the user, who receives it at his terminal.

[0598] As a concrete example, the following prompt sentence is shown.

[0599] Input: The user types "Taro Yamada, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo."

[0600] Anonymization: The server converts it to "YT, Shinjuku-ku, Tokyo."

[0601] Question input: User types "I am very sad."

[0602] Emotion analysis: The server recognizes it as "sadness."

[0603] Emotion information transmission: The server sends the emotion information of "sadness" to the generative AI model.

[0604] Answer generation: The generative AI model generates the answer, "That's very painful. Please let me know if there's anything I can do to help."

[0605] Send Answer: The server sends the generated answer to the user.

[0606] In this way, the server can provide appropriate responses that correspond to the user's emotions while protecting the user's personal information.

[0607] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0608] Step 1:

[0609] The user enters personal information.

[0610] The user uses the terminal to input personal information such as name and address. For example, they input "Yamada Taro, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo." The terminal then sends this information to the server.

[0611] Input: "Yamada Taro, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo"

[0612] Output: Personal information sent to the server

[0613] Step 2:

[0614] The server receives the personal information.

[0615] The server receives the personal information sent by the device and temporarily stores it for further processing.

[0616] Input: Personal information sent from the device

[0617] Output: Temporarily stored personal information

[0618] Step 3:

[0619] The server anonymizes your personal information.

[0620] The server anonymizes personal information based on specific rules. Specifically, it converts names to initials and addresses to area-level information. For example, it converts "Yamada Taro" to "YT" and "1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[0621] Input: Temporarily stored personal information

[0622] Output: Anonymized personal information (e.g., "YT, Shinjuku-ku, Tokyo")

[0623] Step 4:

[0624] The user enters a question.

[0625] The user uses the terminal to input a question containing an emotion, for example, "I am very sad." The terminal then sends this question to the server.

[0626] Enter: "I'm so sad."

[0627] Output: The question sent to the server

[0628] Step 5:

[0629] The server analyzes emotions using an emotion engine.

[0630] The server uses an emotion engine to analyze emotions from the text information of the user's question. For example, it recognizes the emotion "sadness" from the text "I am very sad."

[0631] Input: The question sent to the server

[0632] Output: Parsed emotion information (e.g. "sadness")

[0633] Step 6:

[0634] The server sends the emotion information to the generative AI model.

[0635] The server sends the analyzed emotion information to the generative AI model, for example, emotion information of "sadness."

[0636] Input: Parsed emotion information

[0637] Output: Emotion information sent to the generative AI model

[0638] Step 7:

[0639] A generative AI model generates answers based on emotional information.

[0640] The generative AI model generates a response based on the emotional information it receives. For example, if it recognizes the emotion "sadness," it will generate the response "That's very painful. Please let me know if there's anything I can do to help."

[0641] Input: Emotion information sent to the generative AI model

[0642] Output: The generated answer (e.g., "That's very painful. Let me know if there's anything I can do to help you.")

[0643] Step 8:

[0644] The server sends the generated answer to the user.

[0645] The server sends the answer generated by the generative AI model to the user, who receives the answer on their device.

[0646] Input: Generated answer

[0647] Output: The answer sent to the user

[0648] In this way, the server can provide appropriate responses that correspond to the user's emotions while protecting the user's personal information.

[0649] (Application example 2)

[0650] 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."

[0651] Conventional systems have difficulty not only anonymizing and decrypting personal information, but also responding appropriately based on the user's emotions. Furthermore, responses that do not take the user's emotions into consideration can potentially reduce user satisfaction. Furthermore, there was no system that could simultaneously anonymize personal information and analyze emotions, making it difficult to balance security and user experience.

[0652] 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.

[0653] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing the user's emotions, and means for the generation AI to generate an answer based on the emotion information from the emotion engine. This makes it possible to protect personal information and respond appropriately according to the user's emotions.

[0654] "Personal information" is information that can identify a specific individual, including name, address, telephone number, etc.

[0655] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0656] "Generative AI" is an artificial intelligence system that generates appropriate answers or information based on input data.

[0657] The "emotion engine" is a system that analyzes emotions from the user's text information and sends that emotional information to the generation AI.

[0658] "Emotion information" is data that represents the user's emotions, and indicates an emotional state such as "sadness" or "happiness," for example.

[0659] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0660] A "question" is an inquiry or request that a user enters into the system.

[0661] An "answer" is a response or information that a generative AI generates in response to a user's question.

[0662] A system for implementing this invention includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to a generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing the user's emotions, and means for the generation AI to generate an answer based on the emotion information from the emotion engine.

[0663] System configuration

[0664] 1. Anonymization of personal information

[0665] The server anonymizes questions containing personal information received from users based on certain rules, such as converting names to initials and phone numbers to "XXX-XXX-XXX."

[0666] 2. Emotion analysis using an emotion engine

[0667] The server sends the anonymized questions to an emotion engine that analyzes the user's emotions. The emotion engine recognizes emotions such as "sadness" or "joy" from the text information.

[0668] 3. Answer generation using generative AI

[0669] The server sends the emotion information from the emotion engine to the generation AI, which then generates an appropriate response based on the emotion. For example, if the generation AI recognizes the emotion "sadness," it will generate a response that comforts the user.

[0670] 4. Decrypting the Answer

[0671] The server receives the answer from the generating AI and converts the anonymized information back to the original information based on specific rules.

[0672] Hardware and software used

[0673] Hardware: Servers, user devices (smartphones, smart glasses, etc.)

[0674] Software: emotion engines (e.g., sentiment-analysis models from the transformers library), generative AI (e.g., GPT-3)

[0675] Specific examples

[0676] If a user enters "I'm very sad. My name is Taro Tanaka. My phone number is 123-456-789," the server anonymizes the name and phone number and uses an emotion engine to recognize the emotion "sad." The generative AI then generates a response saying, "You seem to be in trouble. Please don't worry," and the server provides this response to the user.

[0677] Prompt Sentence Examples

[0678] User input: "I'm so sad. My name is Taro Tanaka. My phone number is 123-456-789."

[0679] Prompt the generative AI model: "Analyze the user's sentiment and generate an appropriate response."

[0680] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0681] Step 1:

[0682] The user inputs a question including personal information using a terminal. The input question is in the form of, for example, "I am very sad. My name is Taro Tanaka. My phone number is 123-456-789." The terminal sends this input to the server.

[0683] Step 2:

[0684] The server anonymizes personal information from the received questions based on specific rules. Specifically, it converts names to initials and phone numbers to "XXX-XXX-XXX." The input is "I'm so sad. My name is Tanaka Taro. My phone number is 123-456-789." The output is "I'm so sad. My name is TT. My phone number is XXX-XXX-XXX."

[0685] Step 3:

[0686] The server sends the anonymized question to the emotion engine, which analyzes emotions from text information. The input is "I'm very sad. My name is TT. My phone number is XXX-XXX-XXX." The output is the emotion information "sadness."

[0687] Step 4:

[0688] The server sends the emotion information from the emotion engine to the generation AI. The generation AI generates an appropriate response based on the emotion information. The input is the emotion information "sadness," and the output is the response "It seems you're in trouble. Please don't worry."

[0689] Step 5:

[0690] The server receives the answer from the generative AI and converts the anonymized information back to the original information based on specific rules. The input is "I see you're having trouble. Please don't worry," and the output is the answer as is.

[0691] Step 6:

[0692] The server sends the generated answer to the user's terminal, and the user receives the answer via the terminal saying, "It seems you are having trouble. Please don't worry."

[0693] Example 3

[0694] Next, a description will be given of a third embodiment of the third embodiment. 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."

[0695] In conventional systems, to provide appropriate answers to user questions, questions containing personal information must be anonymized and sent to a generative AI model. However, these systems have the problem of not taking the user's emotions into account and being unable to generate appropriate answers based on their emotions. In addition, the decoding process required to restore anonymized information to its original form is complex, making it difficult to provide answers in a format that is easy for users to understand.

[0696] 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.

[0697] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generative AI model, means for receiving an answer from the generative AI model and decrypting the relevant anonymous information, means including an emotion engine for analyzing emotions from text information of the user's question, means for sending the emotion information analyzed by the emotion engine to the generative AI model, and means for the generative AI model to generate an answer based on the emotion information. This makes it possible to provide an appropriate answer corresponding to the user's emotion.

[0698] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0699] "Anonymization" is the process of removing or masking personal information so that a specific individual cannot be identified.

[0700] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to analyze input text and generate appropriate answers.

[0701] A "communication module" is a hardware or software component for transmitting and receiving data, and has the ability to communicate with other systems over a network.

[0702] An "emotion engine" is a software component that analyzes emotions from user input text and extracts that emotion information.

[0703] "Decryption" is the process of restoring anonymized information to its original form, and is carried out according to specific rules.

[0704] "Text information" is character string data entered by the user, and takes the form of a question, comment, or the like.

[0705] "Emotion information" is data that indicates the user's emotional state analyzed from text information, and includes, for example, "sadness" or "joy."

[0706] "Analysis" is the process of examining input data in detail to understand its meaning and structure.

[0707] An "answer" is a response message that a generative AI model generates in response to a user's question.

[0708] This invention relates to a system that anonymizes a user's question, sends it to a generative AI model, decrypts the answer from the generative AI model, and provides it to the user. It also includes a function to analyze the user's emotions and generate appropriate answers based on those emotions.

[0709] The server anonymizes the questions received from users by removing or masking personal information based on specific rules. This anonymization process can use regular expressions or data masking techniques. For example, if a user enters "I am very sad," the server will remove the "I" part and leave it as "very sad."

[0710] Next, the server sends the anonymized question to the generative AI model. This is done using a communication module, which uses Python's requests library to send an HTTP request to the generative AI model. The generative AI model analyzes the received question, and the emotion engine recognizes the user's emotion. The emotion engine uses natural language processing (NLP) techniques to extract emotion from text information. For example, it recognizes the emotion "sadness" from the text "very sad."

[0711] The generative AI model generates a response based on the recognized emotion. For example, if the emotion "sadness" is recognized, the generative AI model generates the response "Are you OK? Is there anything I can help you with?". This generation is done using a pre-trained AI model.

[0712] The server receives the answer sent by the generative AI model and uses the communication module to receive the HTTP response. The received answer is decrypted based on specific rules to restore the anonymized information to its original form. For example, the server provides the user with the answer "Are you OK? Is there anything I can help you with?"

[0713] As a concrete example, consider the case where a user types "I am very sad." This question is anonymized by the server and sent to the generative AI model through the communication module. The emotion engine recognizes the emotion "sad" and sends it to the generative AI model. Based on "sad," the generative AI model generates a response: "Are you OK? Is there anything I can help you with?" This response is sent to the server, decrypted, and provided to the user.

[0714] Example prompt sentence:

[0715] User Question: "I'm so sad"

[0716] Prompt for generative AI: "Recognize the user's emotion and generate an appropriate response. Emotion: Sadness"

[0717] Generative AI response: "Are you OK? Is there anything I can help you with?"

[0718] In this way, it becomes possible to provide an appropriate answer that corresponds to the user's feelings. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0719] Step 1:

[0720] The user enters a question.

[0721] The user enters "I am very sad" into the input field of the terminal and clicks the send button. The entered question is sent from the terminal to the server.

[0722] Step 2:

[0723] The server anonymizes the question.

[0724] The server analyzes the received question "I am very sad" and removes or masks personal information. Specifically, it uses regular expressions to remove the "I" part and writes it as "very sad." This process results in an anonymized question that does not contain personal information.

[0725] Step 3:

[0726] The server sends the anonymized questions to the generative AI model.

[0727] The server converts the anonymized question "I'm so sad" into JSON format and sends an HTTP POST request to the generative AI model using the communication module. The input is the anonymized question, and the output is the request sent to the generative AI model.

[0728] Step 4:

[0729] A generative AI model analyzes the question, and an emotion engine recognizes the emotion.

[0730] The generative AI model analyzes the received question "I'm so sad." The emotion engine uses natural language processing (NLP) techniques to extract emotions from text information. For example, it recognizes the emotion "sad" from the text "I'm so sad." The input is an anonymized question, and the output is the recognized emotion information.

[0731] Step 5:

[0732] A generative AI model generates answers based on emotions.

[0733] The generative AI model generates an answer based on the recognized emotion "sadness." For example, if the emotion "sadness" is recognized, the generative AI model generates the answer "Are you OK? Is there anything I can help you with?" The input is emotional information, and the output is the generated answer.

[0734] Step 6:

[0735] The server receives the answer from the generative AI model.

[0736] The server receives the answer sent from the generative AI model. It uses a communication module to receive the HTTP response from the generative AI model. The input is the answer from the generative AI model, and the output is the answer received by the server.

[0737] Step 7:

[0738] The server decrypts the response and provides it to the user.

[0739] The server applies certain rules in reverse to convert the answer it receives from the generative AI model back into a form the user can understand. For example, it provides the answer "Are you OK? Is there anything I can help you with?" to the user verbatim. The input is the answer from the generative AI model, and the output is the decoded answer provided to the user.

[0740] In this way, it is possible to provide an appropriate answer that corresponds to the user's feelings.

[0741] (Application example 3)

[0742] 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."

[0743] In conventional systems, when users anonymously submit questions and receive answers from generative AI, it is difficult to provide appropriate answers that take the user's emotions into consideration. In addition, there is a lack of means to properly decrypt anonymized information, making it difficult for users to receive answers in a format that is easy to understand.

[0744] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes means for receiving a question including personal information and anonymizing the corresponding personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the corresponding anonymous information, means for including an emotion engine that analyzes the user's emotions and sending the emotion information to the generation AI, and means for the generation AI to generate an answer based on the emotion information. This makes it possible to provide an appropriate answer corresponding to the user's emotions and to appropriately decrypt the anonymized information.

[0745] "Personal information" is information that can identify a specific individual.

[0746] "Anonymization" is the process of transforming information so that it can no longer identify a specific individual.

[0747] "Generative AI" is a system that uses artificial intelligence technology to generate answers based on user input.

[0748] An "emotion engine" is a system that has the ability to analyze emotions from a user's text information.

[0749] "Emotion information" is emotional data analyzed from user input.

[0750] "Decryption" is the process of restoring anonymized information to its original form.

[0751] A "question" is an inquiry that a user enters into the generated AI.

[0752] An "answer" is the response that the generation AI generates in response to a user's question.

[0753] The system for implementing the present invention uses the following hardware and software.

[0754] Hardware used

[0755] Smartphone

[0756] server

[0757] Software used

[0758] Communication Module

[0759] Emotion Engine

[0760] Generative AI models (e.g., GPT-4)

[0761] Database (storing the anonymization and decryption rules)

[0762] System configuration

[0763] 1. User Input: The user opens the smartphone app and enters their security question.

[0764] 2. Anonymization: The communication module on the smartphone anonymizes the question and converts it into a format that can be analyzed by the generative AI.

[0765] 3. Sentiment analysis: The sentiment engine analyzes the user's sentiment from the text information of the question.

[0766] 4. Send to Generative AI: Send the anonymized questions and emotion information to the Generative AI on the server.

[0767] 5. Answer generation: Generative AI generates appropriate answers based on emotional information.

[0768] 6. Decryption: The communication module on the server receives the answer from the generating AI and decrypts the anonymous information.

[0769] 7. Display to user: The decrypted answer is displayed to the user on their smartphone.

[0770] Specific examples

[0771] A user types in the question, "I've been seeing some suspicious people around my house lately. What should I do?"

[0772] Anonymization: The question is anonymized and converted to the following format: "What should you do if you see someone suspicious?"

[0773] Sentiment analysis: The emotion engine analyzes the emotion "anxiety."

[0774] Send to the generation AI: The anonymized question "What should you do if you see a suspicious person?" and the emotional information "anxiety" are sent to the generation AI.

[0775] Answer generation: The AI ​​generates the answer, "If you feel unsafe, first contact the police and ensure safety around your home."

[0776] Decoding: The communication module decodes the answer back into a format that corresponds to the original question.

[0777] What the user sees: "If you feel unsafe, please contact the police first and make sure your surroundings are safe."

[0778] Prompt Sentence Examples

[0779] User Question: I've been seeing some suspicious people around my house lately, what should I do?

[0780] Emotional information: Anxiety

[0781] Prompt for generative AI: What should you do if you see someone suspicious? (Emotion: Anxiety)

[0782] In this way, appropriate security advice can be provided that corresponds to the user's emotions.

[0783] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0784] Step 1:

[0785] The user opens the smartphone app and enters their security questions, which are then sent to the app in text format.

[0786] Step 2:

[0787] The communication module of the device anonymizes the input question. Specifically, it converts personal information based on specific rules to generate an anonymized question. The input is the user's question, and the output is the anonymized question.

[0788] Step 3:

[0789] The device's emotion engine analyzes the user's emotions from the text information of the anonymized question. The emotion engine uses natural language processing technology to extract emotional information from the text. The input is the anonymized question, and the output is emotional information.

[0790] Step 4:

[0791] The device sends anonymized questions and emotional information to the generation AI on the server. The input is the anonymized questions and emotional information, and the output is a prompt for the generation AI.

[0792] Step 5:

[0793] The server's generation AI generates an answer based on the received prompt. The generation AI takes emotional information into account and generates an appropriate answer. The input is the prompt, and the output is the generated answer.

[0794] Step 6:

[0795] The server's communication module receives the answer from the generating AI and decrypts the anonymous information. Specifically, it restores the anonymized information to its original form based on specific rules. The input is the generated answer, and the output is the decrypted answer.

[0796] Step 7:

[0797] The terminal displays the decrypted answer to the user, who can then check the answer on their smartphone. The input is the decrypted answer, and the output is the answer displayed to the user.

[0798] 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.

[0799] The data generation model 58 is a so-called generative AI (Artificial Intelligence).

[0800] An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search<URL: https: / / openai.com / blog / chatgpt> ) and other generative AIs. Data generation models 58

[0801] , 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 the data format of voice data, text data, etc. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0802] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[0803] 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.

[0804] [Second embodiment]

[0805] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0806] 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.

[0807] 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).

[0808] 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.

[0809] 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.

[0810] 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).

[0811] 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.

[0812] 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.

[0813] 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.

[0814] 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.

[0815] 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.

[0816] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0817] "Example 1"

[0818] In one embodiment of the present invention, an interface for accepting input from a user is provided as a means for receiving questions including personal information. This interface may be, for example, an input form on a web page or a text input field in a smartphone application.

[0819] "Example 2"

[0820] Next, as a means of anonymizing the received personal information, a function is provided to convert personal information based on specific rules, such as converting names to initials or addresses to area-level information.

[0821] "Example 3"

[0822] A communication module is provided as a means for transmitting the anonymized questions to the generation AI. This communication module converts the anonymized questions into a format that can be analyzed by the generation AI and transmits them to the generation AI.

[0823] "Example 4"

[0824] Finally, we will provide a mechanism to reverse the specific rules described above as a means of receiving the answers from the Generator AI and decrypting the relevant anonymous information, thereby returning the answers from the Generator AI in a form that can be understood by the user.

[0825] The processing flow of each embodiment will be described below.

[0826] "Example 1"

[0827] Step 1: A user enters a question into a form on a webpage, which may contain personal information.

[0828] Step 2: The system receives the question and anonymizes the personal information based on certain rules, for example, converting names to initials and addresses to locality level information.

[0829] Step 3: The anonymized question is sent to the generation AI, where a communication module converts the question into a format that the generation AI can parse.

[0830] Step 4: Receive the answer from the Generator AI and decrypt the corresponding anonymous information, specifically by applying the specific rules mentioned above in reverse to return the answer to a form that the user can understand.

[0831] Example 1

[0832] 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."

[0833] Conventional systems have had difficulty safely processing questions containing personal information and generating appropriate answers. Furthermore, the process of anonymizing and decrypting personal information is complex, resulting in delays in providing answers to users. Furthermore, there is a lack of efficient means for generating prompts suitable for generative AI models.

[0834] 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.

[0835] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for analyzing the anonymized question and generating a prompt sentence suitable for the generative AI model, means for sending the generated prompt sentence to the generative AI model, means for receiving an answer from the generative AI model and decrypting the relevant anonymous information, and means for displaying the decrypted answer to the user. This makes it possible to process personal information safely and provide the user with a prompt and appropriate answer.

[0836] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0837] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0838] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers based on input prompt sentences.

[0839] A "prompt sentence" is a sentence input to a generative AI model that appropriately expresses the user's question.

[0840] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[0841] A "server" is a computer system that processes and stores data over a network.

[0842] A "user" is an individual or organization that uses the system to enter questions and receive answers.

[0843] This invention is a system that allows a user to input a question that includes personal information and provides an appropriate answer to the question using a generative AI model. Specific embodiments of this system are described below.

[0844] A user enters a question using a form on a web page or a text input field in a smartphone application, for example, "My name is Taro Tanaka. My address is Shinjuku-ku, Tokyo. How can I protect my personal information?"

[0845] The device sends the entered question to the server using the HTTPS protocol, with the data encrypted.

[0846] The server analyzes the received questions and anonymizes the personal information. Anonymization involves converting personal information based on specific rules. For example, "Taro Tanaka" could be "User A" and "Shinjuku Ward, Tokyo" could be "Area X."

[0847] The server then analyzes the anonymized question and generates a prompt suitable for the generative AI model, such as:

[0848] The user's name is User A and they live in Region X. How do you protect their personal information?

[0849] The server sends the generated prompt to a generative AI model, such as OpenAI's GPT-3 or GPT-4.

[0850] The generative AI model generates an answer based on the prompt it receives, for example, the following answer:

[0851] You can help protect your personal information by:

[0852] 1. Use strong passwords.

[0853] 2. Only share personal information with people you trust.

[0854] 3. Install security software and keep it up to date.

[0855] 4. Be careful when using public Wi-Fi.

[0856] The server decrypts the answer received from the generative AI model and restores the original personal information. For example, it restores "User A" to "Taro Tanaka" and "Area X" to "Shinjuku Ward, Tokyo."

[0857] Finally, the server displays the decrypted answers to the user, who can then check them on a web page or smartphone application.

[0858] This system allows users to safely process their personal information and receive prompt and appropriate answers.

[0859] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0860] Step 1:

[0861] The user enters a question.

[0862] A user types a question into a text field on a web page or smartphone application, for example, "My name is Taro Tanaka. My address is Shinjuku-ku, Tokyo. How can I protect my personal information?"

[0863] Input: User question

[0864] Output: Question data entered

[0865] Step 2:

[0866] The terminal sends a question to the server.

[0867] The device sends the entered question to a server using the HTTPS protocol, with the data encrypted.

[0868] Input: The entered question data

[0869] Output: The query data sent to the server

[0870] Step 3:

[0871] The server analyzes the question and anonymizes personal information.

[0872] The server analyzes the received questions and anonymizes personal information based on specific rules. For example, "Taro Tanaka" is "User A" and "Shinjuku Ward, Tokyo" is "Area X."

[0873] Input: Query data sent to the server

[0874] Output: Anonymized question data

[0875] Step 4:

[0876] The server analyzes the anonymized questions and generates prompts appropriate for the generative AI model.

[0877] The server analyzes the anonymized question and generates a prompt appropriate for the generative AI model, such as: "The user's name is User A and their address is in Region X. Please tell me how to protect their personal information."

[0878] Input: Anonymized question data

[0879] Output: Generated prompt statement

[0880] Step 5:

[0881] The server sends the generated prompt to the generative AI model.

[0882] The server sends the generated prompt to a generative AI model, which uses natural language processing technology.

[0883] Input: Generated prompt text

[0884] Output: The prompt sent to the generative AI model

[0885] Step 6:

[0886] A generative AI model generates the answer.

[0887] The generative AI model generates answers based on the prompt it receives, such as: "To protect your personal information, you can: 1. Use strong passwords. 2. Only share personal information with people you trust. 3. Install security software and keep it up to date. 4. Exercise caution when using public Wi-Fi."

[0888] Input: The prompt sent to the generative AI model

[0889] Output: The generated answer

[0890] Step 7:

[0891] The server receives the answer from the generative AI model and decrypts the anonymous information.

[0892] The server decrypts the answer received from the generative AI model and restores the original personal information. For example, it restores "User A" to "Taro Tanaka" and "Area X" to "Shinjuku Ward, Tokyo."

[0893] Input: Generated Answer

[0894] Output: Decoded answer

[0895] Step 8:

[0896] The server displays the decrypted answer to the user.

[0897] The server then displays the decrypted answers to the user, who can then check them on a web page or smartphone application.

[0898] Input: Decrypted answer

[0899] Output: The answer displayed to the user

[0900] (Application example 1)

[0901] 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."

[0902] There is a need to provide an environment where questions containing personal information can be handled safely and where users can enter questions with peace of mind. However, conventional systems have a high risk of personal information leakage and unauthorized access, making it difficult to adequately protect user privacy. Furthermore, even when providing answers using generative AI models, there is a lack of means to properly anonymize and decrypt users' personal information.

[0903] 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.

[0904] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means for decrypting the encrypted data and inputting it as a prompt sentence to the generation AI model, and means for obtaining an answer from the generation AI model and providing it to the user. This makes it possible to provide an appropriate answer using the generation AI model while safely handling the user's personal information.

[0905] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0906] A "question" is a sentence or phrase entered by a user to request information.

[0907] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0908] "Generative AI" is a system that uses artificial intelligence technology to generate appropriate answers to users' questions.

[0909] "Decryption" is the process of restoring anonymized information to its original form.

[0910] "Encryption" is the process of converting data using a specific algorithm to make it unreadable to third parties.

[0911] A "prompt sentence" is a sentence input to a generative AI model, and is an instruction sentence that enables the AI ​​to generate an appropriate answer.

[0912] An "answer" is information generated by the generative AI in response to a user's question.

[0913] "User" means an individual or entity that utilizes the System to enter questions and receive answers.

[0914] As an embodiment of the present invention, the following system is constructed.

[0915] System Configuration

[0916] The system includes a user terminal, a server, and a generative AI model. The user terminal is a device such as a smartphone or PC, and provides an interface for users to input questions. The server receives questions from users, anonymizes them, encrypts them, decrypts them, generates prompts for the generative AI model, and provides answers.

[0917] Hardware and software used

[0918] Hardware: smartphones, PCs, servers

[0919] Software: Python, OpenAI API, cryptography library

[0920] Data processing and calculation

[0921] 1. User Device:

[0922] The user enters a question through an interface on the device, which is an application that includes a text entry field.

[0923] 2. Server:

[0924] When the server receives a question from a user, it first anonymizes the personal information. The anonymized question is then encrypted and sent to a generative AI model. The generative AI model generates an appropriate answer based on the prompt. The generated answer is then decrypted by the server and provided to the user.

[0925] Specific examples

[0926] If a user types the question "How do I manage my passwords securely?", the system will:

[0927] 1. Enter the question "How do I manage my passwords securely?" on the user device.

[0928] 2. The server receives this question and anonymizes your personal information.

[0929] 3. The anonymized question is encrypted and sent as a prompt to the generative AI model.

[0930] 4. The generative AI model generates an answer based on the prompt.

[0931] 5. The server decrypts the generated answer and provides it to the user.

[0932] Prompt Sentence Examples

[0933] Users ask: How do I keep my passwords safe?

[0934] Please provide security advice.

[0935] This system allows users to receive appropriate answers using generative AI models while keeping their personal information safe.

[0936] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0937] Step 1:

[0938] Enter the question on the user's device.

[0939] Users enter their questions into a text entry field on a smartphone or PC application, and the entered questions are sent to the server.

[0940] Input: A question entered by the user (e.g., "How do I keep my passwords safe?")

[0941] Output: The question data sent to the server

[0942] Step 2:

[0943] The server receives the question and anonymizes the personal information.

[0944] The server anonymizes personal information from the received questions based on specific rules, for example, masking personal information such as name and address.

[0945] Input: Question data submitted by the user

[0946] Output: Anonymized question data

[0947] Step 3:

[0948] The server encrypts the anonymized question.

[0949] The server encrypts the anonymized question data using an encryption algorithm, ensuring the security of the data.

[0950] Input: Anonymized question data

[0951] Output: Encrypted question data

[0952] Step 4:

[0953] The server sends the encrypted question to the generative AI model.

[0954] The server decrypts the encrypted question data and sends it to the generative AI model as a prompt, which is an instruction for the generative AI model to generate an appropriate answer.

[0955] Input: Encrypted question data

[0956] Output: The prompt sent to the generative AI model

[0957] Step 5:

[0958] A generative AI model generates an answer based on the prompt.

[0959] The generative AI model analyzes the prompt and generates an appropriate answer, which is then sent to the server.

[0960] Input: The prompt sent to the generative AI model

[0961] Output: Answer data from the generative AI model

[0962] Step 6:

[0963] The server receives and decodes the answer from the generative AI model.

[0964] The server decrypts the answer data received from the generative AI model and returns it to its original format.

[0965] Input: Answer data from the generative AI model

[0966] Output: Decoded answer data

[0967] Step 7:

[0968] The server provides the decrypted answer to the user.

[0969] The server sends the decrypted answer data to the user's terminal and provides it to the user, who can then check the answer on the terminal.

[0970] Input: Decrypted answer data

[0971] Output: Answer data displayed on the user's device

[0972] Example 2

[0973] 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."

[0974] When inputting data including personal information into a generative AI model, it is necessary to obtain useful responses from the generative AI model while reducing the risk of personal information leakage. However, conventional systems have not been able to efficiently anonymize personal information and input it into the generative AI model, making it difficult to achieve both security and convenience.

[0975] 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.

[0976] In this invention, the server includes means for receiving data including personal information and anonymizing the personal information, means for transmitting the anonymized data to the generative AI model, means for receiving a response from the generative AI model and generating a prompt sentence, means for inputting the generated prompt sentence into the generative AI model, and means for receiving a response from the generative AI model. This makes it possible to efficiently obtain useful responses from the generative AI model while reducing the risk of personal information leaks.

[0977] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[0978] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[0979] A "generative AI model" is an artificial intelligence model that performs natural language processing and generates a response based on an input prompt.

[0980] A "prompt" is a document that serves as input to a generative AI model and contains instructions and information that enable the model to generate an appropriate response.

[0981] A "server" is a computer system that processes and stores data over a network.

[0982] "Data" means a collection of information, including personal information and other related information.

[0983] A "response" is the output result that a generative AI model generates based on a prompt sentence.

[0984] "Transmitting" is the act of moving data from one place to another.

[0985] "Receiving" is the act of receiving transmitted data.

[0986] "Conversion" is the process of changing the format or content of data into a different format or content.

[0987] MODE FOR CARRYING OUT THE INVENTION

[0988] This invention relates to a system that anonymizes data including personal information, generates prompt sentences to be input into a generative AI model, and receives responses from the generative AI model. Specific embodiments of this system are described below.

[0989] Hardware and software used

[0990] Hardware: High-performance server (e.g. general-purpose server)

[0991] Software: Python, Pandas, Numpy, generative AI models (e.g., natural language processing models)

[0992] System configuration

[0993] 1. Receipt of personal information

[0994] The server receives personal information submitted by the user, such as name and address information entered by the user into a web form.

[0995] Specific operation: The server receives an HTTP request and extracts personal information from the request body.

[0996] 2. Application of anonymization rules

[0997] The server applies anonymization rules to the personal information it receives, for example, converting names to initials and addresses to locality level information.

[0998] Specific operation: The server runs a Python script to convert the name "Taro Tanaka" to "TT" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[0999] 3. Save the converted data

[1000] The server stores the anonymized data in a database, which allows it to be retrieved later if needed.

[1001] Specific operation: The server saves the converted data in the SQL database using an INSERT statement.

[1002] 4. Prompt generation

[1003] The server generates prompts to input to the generative AI model, for example, prompts containing anonymized names and addresses.

[1004] What happens: The server uses Python's string manipulation functions to generate a prompt like this:

[1005] Name:TT

[1006] Address: Shinjuku Ward, Tokyo

[1007] Use this information to generate a user profile.

[1008] 5. Input to the generative AI model

[1009] The server inputs the generated prompt sentence into a generative AI model and receives a response from the model. For example, it inputs a prompt sentence using a natural language processing model and obtains the generated text.

[1010] Specific operation: The server calls the API of the generative AI model, sends a prompt sentence, and receives a response.

[1011] In this way, the server can anonymize personal information, generate prompts to input into the generative AI model, and receive responses from the model. This system makes it possible to efficiently obtain useful responses from the generative AI model while reducing the risk of personal information leaks.

[1012] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1013] Step 1:

[1014] The server receives the personal information sent by the user.

[1015] Input: Personal information such as name and address that a user enters into a web form.

[1016] Specific operation: The server receives an HTTP request and extracts personal information from the request body.

[1017] Output: Extracted personal information (e.g., name "Taro Tanaka", address "Nishi-Shinjuku, Shinjuku-ku, Tokyo").

[1018] Step 2:

[1019] The server applies anonymization rules to the received personal information.

[1020] Input: Personal information extracted in step 1.

[1021] Specific operation: The server runs a Python script to convert the name "Taro Tanaka" to "TT" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[1022] Output: Anonymized data (e.g., name "TT", address "Shinjuku-ku, Tokyo").

[1023] Step 3:

[1024] The server stores the anonymized data in a database.

[1025] Input: Anonymized data from step 2.

[1026] Specific operation: The server saves the converted data in the SQL database using an INSERT statement.

[1027] Output: Anonymized data stored in a database.

[1028] Step 4:

[1029] The server generates prompt sentences to input into the generative AI model.

[1030] Input: Anonymized data from step 2.

[1031] Specific operation: The server generates a prompt sentence using Python's string manipulation functions.

[1032] Output: Generated prompt text (e.g. "Name: TT Address: Shinjuku-ku, Tokyo Please generate a user profile based on this information.").

[1033] Step 5:

[1034] The server inputs the generated prompt sentence into the generative AI model and receives a response from the model.

[1035] Input: The prompt statement generated in step 4.

[1036] Specific operation: The server calls the API of the generative AI model, sends a prompt sentence, and receives a response.

[1037] Output: The response from the generative AI model (e.g., a generated user profile).

[1038] (Application example 2)

[1039] 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."

[1040] Safely handling data that includes personal information is extremely important from the perspective of privacy protection. However, transmitting personal information as is carries the risk of information leakage. In particular, when using generative AI models, anonymization is necessary because there is a possibility that personal information may be passed on to third parties. Furthermore, the accuracy of restoring anonymized information to its original form is also important. To solve these issues, a system is needed that anonymizes personal information based on specific rules and allows the safe use of generative AI models.

[1041] 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.

[1042] In this invention, the server includes means for receiving questions containing personal information and anonymizing the relevant personal information, means for sending the anonymized questions to the generation AI, means for receiving answers from the generation AI and decrypting the relevant anonymous information, means for converting names to initials, means for converting addresses to area-level information, and means for anonymizing other personal information based on specific rules. This makes it possible to safely anonymize personal information and reduce the risk of information leakage when using a generative AI model.

[1043] "Personal information" is information that can identify a specific individual.

[1044] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1045] "Generative AI" is a system that uses artificial intelligence technology to generate new data based on input data.

[1046] "Initials" is an abbreviation that refers to the first letter of each part of a name.

[1047] "Area-level information" is information that indicates a wide area without including detailed address information.

[1048] "Decryption" is the process of restoring anonymized information to its original form.

[1049] "Specific Rules" are the set of rules used to anonymize or decrypt personal information.

[1050] A "question" is part of the data input to the generating AI and is the information used to obtain an answer.

[1051] An "answer" is the output data that the generative AI generates based on the question.

[1052] A system for implementing this invention includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to a generation AI, means for receiving a response from the generation AI and decrypting the relevant anonymous information, means for converting names to initials, means for converting addresses to area-level information, and means for anonymizing other personal information based on specific rules.

[1053] System configuration

[1054] The server receives a question containing personal information entered by the user. The received personal information is anonymized based on specific rules. For example, names are converted to initials, and addresses are converted to area-level information. The anonymized question is sent to the generation AI. The generation AI generates an answer based on the question and sends the answer back to the server. The server receives the answer from the generation AI and decrypts the anonymized information back to the original information.

[1055] Hardware and software used

[1056] The system uses mobile devices such as smartphones and tablets, and the software is implemented using Python, including functions for anonymizing personal information and an API for communicating with the generative AI model.

[1057] Specific examples

[1058] For example, if a user enters the name "Yamada Taro" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo," the server converts this to "YT" and "Tokyo." The converted information is sent to the generation AI, which generates an answer to the question. The generated answer is sent back to the server, which decrypts the anonymized information to its original form.

[1059] Prompt Sentence Examples

[1060] An example of a prompt to input to a generative AI model is as follows:

[1061] Create an application that anonymizes personal information entered by users, converting names to initials and addresses to locality level information. Implement it using Python.

[1062] In this way, it is possible to safely anonymize personal information and reduce the risk of information leakage when using generative AI models.

[1063] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1064] Step 1:

[1065] The user inputs a question containing personal information into the terminal.

[1066] Input: The name, address, and question entered by the user.

[1067] Output: Question data including personal information entered into the terminal.

[1068] Specific operation: The user enters the name "Yamada Taro," the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo," and the question into the input form on the terminal.

[1069] Step 2:

[1070] The device anonymizes your personal information.

[1071] Input: Question data, including personal information entered by the user.

[1072] Output: Anonymized question data.

[1073] Specific operations: The device will convert your name to "YT" and your address to "Tokyo." Other personal information will also be anonymized based on specific rules.

[1074] Step 3:

[1075] The device sends the anonymized question to the server.

[1076] Input: Anonymized question data.

[1077] Output: Anonymized question data sent to the server.

[1078] Specific operation: The terminal transmits data over the network to send the anonymized question data to the server.

[1079] Step 4:

[1080] The server sends the anonymized questions to the generation AI.

[1081] Input: Anonymized question data received by the server.

[1082] Output: Anonymized question data sent to the generative AI.

[1083] Specific operation: The server calls the API to send the anonymized question data to the generative AI model.

[1084] Step 5:

[1085] Generative AI generates answers to questions.

[1086] Input: Anonymized question data sent to the generative AI.

[1087] Output: Answer data generated by the generation AI.

[1088] How it works: The generative AI model analyzes anonymized question data and generates appropriate answers.

[1089] Step 6:

[1090] The server receives the answer from the generation AI.

[1091] Input: Answer data generated by the generation AI.

[1092] Output: The response data received by the server.

[1093] Specific operation: The server obtains data through an API to receive answer data from the generative AI model.

[1094] Step 7:

[1095] The server decrypts the anonymized information.

[1096] Input: The answer data received from the generation AI.

[1097] Output: The decoded answer data.

[1098] Specific operation: The server restores the anonymized information to its original form based on specific rules. For example, it decrypts "YT" into "Yamada Taro" and "Tokyo" into "Tokyo, Shinjuku Ward, Nishi-Shinjuku."

[1099] Step 8:

[1100] The server sends the decrypted response to the terminal.

[1101] Input: Decoded answer data.

[1102] Output: The decrypted answer data sent to the device.

[1103] Specific operation: The server transmits the decrypted response data to the terminal through the network.

[1104] Step 9:

[1105] The terminal displays the decrypted answer to the user.

[1106] Input: The decrypted response data received by the terminal.

[1107] Output: The decoded answer displayed to the user.

[1108] Specific operation: The terminal displays the decrypted answer data on the screen to show it to the user.

[1109] Example 3

[1110] 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."

[1111] Conventional systems have the problem of insufficient privacy protection because they send queries containing personal information directly to the generative AI model. Furthermore, there is also the issue that the responses from the generative AI model are difficult for users to understand if they remain anonymized. Furthermore, the process of converting the data into a format that the generative AI model can analyze is complex, requiring efficient data processing.

[1112] 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.

[1113] In this invention, the server includes means for receiving an inquiry including personal information and anonymizing the relevant personal information, means for sending the anonymized inquiry to the generative AI model, means for receiving a response from the generative AI model and decrypting the relevant anonymous information, a communication module for converting the response into a format that can be analyzed by the generative AI model, and means for providing the decrypted response to the user. This makes it possible to provide a response in a format that is easy for the user to understand while ensuring the protection of personal information.

[1114] "Personal information" is information that can identify a specific individual.

[1115] An "inquiry" is a question or request made by a user to the system.

[1116] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1117] A "generative AI model" is an algorithm or system that uses artificial intelligence to analyze input data and generate an appropriate response.

[1118] A "communications module" is hardware or software for transmitting and receiving data.

[1119] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1120] "User" means any individual or entity that uses the System.

[1121] A "server" is a computer system that processes and stores data on a network.

[1122] A "response" is the answer or information provided by a generative AI model as a result of its analysis.

[1123] MODE FOR CARRYING OUT THE INVENTION

[1124] This invention is a system that anonymizes inquiries that include personal information, sends them to a generative AI model, and decrypts the responses from the generative AI model and provides them to the user. Specific embodiments of this system are described below.

[1125] Hardware and software used

[1126] Hardware: Servers, communication modules, user devices (PCs, smartphones, etc.)

[1127] Software: Generative AI models (e.g., OpenAI GPT-4), programs implementing specific rules for anonymization and decryption

[1128] System Operation

[1129] 1. The user enters a question

[1130] A user uses a terminal to input a question into the system's interface, for example, "My name is Taro Tanaka. How should I learn programming?"

[1131] 2. The server receives the query

[1132] The server receives the questions sent by the user and stores them in their original format.

[1133] 3. The server anonymizes the question

[1134] The server anonymizes personal information and specific identifying information contained in the question based on specific rules, for example, converting "My name is Tanaka Taro" into "My name is [anonymous]."

[1135] 4. The server sends the anonymized questions to the generating AI via the communication module.

[1136] The server sends the anonymized questions to the Generator AI using a communication module, which converts the questions into a format that the Generator AI can parse.

[1137] 5. Generative AI analyzes the question and generates an answer

[1138] The generative AI analyzes the question it receives and generates an appropriate answer. For example, in response to the question, "How should I learn programming?", it generates the answer, "To learn programming, you should first choose a basic language and then proceed with your studies using online courses and books."

[1139] 6. The server receives the answer from the generation AI

[1140] The server receives the answers sent by the generation AI and stores them in their original format.

[1141] 7. The server decrypts the response

[1142] The server decrypts the anonymous information contained in the answer from the AI ​​generator based on specific rules. For example, it converts the answer "For [anonymous], in order to learn programming..." into "For Taro Tanaka, in order to learn programming..."

[1143] 8. The server provides the decrypted answer to the user

[1144] The server sends the decrypted answer to the user's terminal, where the user can check the answer through the terminal interface.

[1145] Specific examples

[1146] Example 1: A user submits the question "My name is Taro Tanaka. How can I learn programming?"

[1147] 1. The user uses the terminal to enter a question and clicks the submit button.

[1148] 2. The server receives the question, anonymizes it, and converts it into "My name is [Anonymous]. How can I learn programming?"

[1149] 3. Send the anonymized question to the generation AI via the communication module.

[1150] 4. In response to the question, "How should I learn programming?", the generative AI generates the answer, "To learn programming, it's a good idea to first choose a basic language and then progress through online courses and books."

[1151] 5. The server receives the answer from the generating AI and provides it to the user as is, since there is no need to decrypt the anonymous information.

[1152] Example prompt sentence:

[1153] User asks: "My name is Taro Tanaka. How can I learn programming?"

[1154] Anonymized question: "My name is [Anonymous]. How do I learn to code?"

[1155] Prompt for generative AI: "How do I learn to code?"

[1156] Generative AI's answer: "To learn programming, start by choosing a basic language and then work your way up using online courses and books."

[1157] In this way, the system anonymizes the user's question, sends it to the generation AI, and decrypts the answer from the generation AI and provides it to the user. The flow of the identification process in Example 3 will be described with reference to FIG.

[1158] Step 1:

[1159] The user enters a question

[1160] Input: The user uses the terminal to type a question, for example, "My name is Taro Tanaka. How should I learn programming?"

[1161] Specific actions: The user enters a question into the device interface and clicks the send button.

[1162] Output: The entered question is sent from the terminal to the server.

[1163] Step 2:

[1164] The server receives the query

[1165] Input: The question sent from the terminal.

[1166] Specific operation: The server receives the question sent by the user. The received question is saved in its original format.

[1167] Output: The received questions are saved on the server.

[1168] Step 3:

[1169] The server anonymizes the question

[1170] Input: A question stored on the server.

[1171] Specific operation: The server anonymizes personal information and specific identifying information included in the question based on specific rules. For example, it converts "My name is Tanaka Taro" into "My name is [anonymous]."

[1172] Output: Anonymized questions are generated.

[1173] Step 4:

[1174] The server sends the anonymized questions to the generating AI via a communication module.

[1175] Input: Anonymized question.

[1176] Specific operation: The server sends the anonymized question to the generation AI using a communication module, which converts the question into a format that the generation AI can analyze.

[1177] Output: The question is sent to the Generator AI in a format that the Generator AI can parse.

[1178] Step 5:

[1179] Generative AI analyzes the question and generates an answer

[1180] Input: Anonymized question sent to the generative AI.

[1181] How it works: Generative AI analyzes the questions it receives and generates appropriate answers. For example, in response to the question, "How should I learn programming?", it generates the answer, "To learn programming, first choose a basic language and then study using online courses and books."

[1182] Output: The generated answer is sent to the server.

[1183] Step 6:

[1184] The server receives the answer from the generation AI.

[1185] Input: The answer sent by the generation AI.

[1186] Specific operation: The server receives the answer sent from the generation AI. The received answer is saved in its original format.

[1187] Output: The received answer is saved on the server.

[1188] Step 7:

[1189] The server decrypts the response

[1190] Input: The answer sent by the generation AI.

[1191] Specific operation: The server decrypts the anonymous information included in the answer from the generation AI based on specific rules. For example, the answer "For [anonymous], in order to learn programming..." is converted into "For Taro Tanaka, in order to learn programming..."

[1192] Output: The decoded answer is generated.

[1193] Step 8:

[1194] The server provides the decrypted answer to the user

[1195] Input: The decrypted answer.

[1196] Specific operation: The server sends the decrypted answer to the user's device, where the user can check the answer through the device interface.

[1197] Output: The decrypted answer is displayed on the user's terminal.

[1198] (Application example 3)

[1199] 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."

[1200] In conventional systems, when sending questions containing personal information to the AI ​​generator, there was a high risk of personal information being leaked, and user privacy was not protected. Another issue was the difficulty of accurately decrypting anonymized information when returning the AI ​​generator's answers to a format that users could understand. Furthermore, there was a lack of efficient ways to perform these processes on smartphones.

[1201] 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.

[1202] In this invention, the server includes means for receiving questions that include personal information and anonymizing the relevant personal information, means for sending the anonymized questions to the generation AI, means for receiving answers from the generation AI and decrypting the relevant anonymous information, and means, as an application installed on a smartphone, for sending the anonymized questions to the generation AI and decrypting the answers from the generation AI and providing them to the user. This makes it possible to efficiently realize a question-answering system that uses the generation AI on a smartphone while protecting user privacy.

[1203] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1204] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1205] "Generative AI" is an artificial intelligence model that generates natural language answers based on given input data.

[1206] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1207] A "smartphone" is a type of mobile phone, a multi-function device that can connect to the Internet and install applications.

[1208] An "application" is software installed on a smartphone to provide specific functions or services.

[1209] A "question" is text data that a user enters to request information or advice from the generated AI.

[1210] An "answer" is text data generated by the generation AI in response to a question.

[1211] A "user" is an individual who uses the system to submit questions to the generating AI and receive answers.

[1212] As an embodiment of the present invention, the application "SecureQuery" installed on a smartphone is used. A specific implementation method thereof will be described below.

[1213] First, a user launches the SecureQuery application on their smartphone. The application provides an interface for entering a question, including personal information. Once the user enters the question, the application anonymizes it.

[1214] The anonymization process involves converting personal information based on specific rules. For example, personal information such as usernames and addresses are replaced with generic terms like "anonymous" or "private." These anonymized questions are then converted into a format that can be analyzed by the generative AI.

[1215] The application then sends the anonymized question to a Generative AI, which generates a natural language answer based on the given question. This Generative AI is located on a server in the cloud and accessed via the internet.

[1216] The answer from the AI ​​generator is sent back to the application, which receives the answer and performs a decryption process to restore the anonymized information to its original form based on specific rules. This allows the user to receive the answer from the AI ​​generator in a format that they can understand.

[1217] The system is implemented using the following hardware and software:

[1218] Hardware: Smartphone

[1219] Software: Python, requests library, JSON

[1220] The specific data processing and calculation flow is as follows:

[1221] 1. Anonymization processing: Personal information from questions entered by the user is anonymized within the application based on specific rules.

[1222] 2. Send to Generator: Send the anonymized question to the Generator endpoint, which analyzes the question and generates an answer.

[1223] 3. Decryption process: Receives the answer from the generation AI and restores the anonymized information to its original form based on specific rules.

[1224] For example, if a user enters the question, "What should I do if my username password is leaked?", this question is anonymized as, "What should I do if my anonymous password is leaked?" The Generative AI generates an appropriate answer to this question, and the application decrypts the answer and provides it to the user.

[1225] Example prompt sentence:

[1226] "What should I do if my username password is compromised?"

[1227] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1228] Step 1:

[1229] The user launches the "SecureQuery" application on their smartphone and enters a question.

[1230] Input: The question text entered by the user (e.g., "What should I do if my username's password is leaked?")

[1231] Output: The question text entered

[1232] Specific behavior: The user enters a question into the application's input field and presses the "Submit" button.

[1233] Step 2:

[1234] The device anonymizes the entered question.

[1235] Input: The question text entered by the user

[1236] Output: Anonymized question text (e.g., "What should I do if my anonymous password is leaked?")

[1237] What it does: An anonymization module within the application replaces personal information with generic terms based on specific rules.

[1238] Step 3:

[1239] The device sends anonymized questions to the generation AI.

[1240] Input: Anonymized question text

[1241] Output: Question text sent to the generation AI

[1242] What it does: Your application makes an HTTP request and sends an anonymized question to the Generative AI endpoint.

[1243] Step 4:

[1244] The server uses generative AI to generate answers to questions.

[1245] Input: Anonymized question text

[1246] Output: Generated answer text

[1247] How it works: The generation AI analyzes the question and generates an appropriate answer, which is then sent back to the device from the server.

[1248] Step 5:

[1249] The device receives the answer from the generating AI and decodes it.

[1250] Input: Answer text returned by the generation AI (e.g., "If your anonymous password is leaked, please change it immediately.")

[1251] Output: Decrypted answer text (e.g., "If your username password has been leaked, please change it immediately.")

[1252] What it does: A decryption module within the application converts the anonymized information back into the original information based on specific rules.

[1253] Step 6:

[1254] The terminal displays the decrypted answer to the user.

[1255] Input: Decrypted answer text

[1256] Output: The answer text that is displayed to the user

[1257] What it does: The application displays the decrypted answer on the screen for the user to review.

[1258] 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.

[1259] "Example 1"

[1260] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[1261] "Example 2"

[1262] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[1263] "Example 3"

[1264] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[1265] The processing flow of each embodiment will be described below.

[1266] "Example 1"

[1267] Step 1: The user enters a question, which is sent to the system as text information.

[1268] Step 2: The emotion engine analyzes the emotion from this text information. For example, it recognizes the emotion "sadness" from the question "I am very sad."

[1269] Step 3: The emotion engine sends the analyzed emotion information to the generation AI.

[1270] Step 4: The generative AI generates an answer based on the emotional information received from the emotion engine. For example, if the generative AI recognizes the emotion "sadness," it generates an answer that comforts the user.

[1271] Step 5: The generation AI provides the generated answer to the user, which allows it to provide a more appropriate answer that reflects the user's emotions.

[1272] Example 1

[1273] 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."

[1274] Conventional systems have had difficulty providing appropriate answers that reflect the user's emotions. Furthermore, they lacked the means to safely process questions containing personal information and to anonymize and decrypt them. This resulted in low user satisfaction and created privacy issues.

[1275] 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.

[1276] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means for providing an interface for accepting input from a user, means for including an emotion engine for analyzing emotions from text information of the user's question, means for sending the emotion information analyzed by the emotion engine to the generation AI, means for the generation AI to generate an answer based on the emotion information, and means for displaying the generated answer to the user. This makes it possible to provide an appropriate answer corresponding to the user's emotion, and to enable safe processing of personal information and privacy protection.

[1277] "Personal information" is information that can identify a specific individual.

[1278] A "question" is an inquiry entered by a user.

[1279] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1280] "Generative AI" is a system that uses artificial intelligence technology to generate answers to users' questions.

[1281] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1282] An "interface" is a means by which a user inputs information into a system, and includes input forms on web pages and text input fields in smartphone applications.

[1283] The "emotion engine" is a system that analyzes emotions from the text information of users' questions.

[1284] "Emotion information" is information about the user's emotions obtained as a result of analysis by the emotion engine.

[1285] An "answer" is the response that the generation AI generates in response to a user's question.

[1286] "Display" refers to the process of visually presenting the generated answer to the user.

[1287] MODE FOR CARRYING OUT THE INVENTION

[1288] The present invention is a system for receiving a question from a user and generating an appropriate answer based on emotions to the question. A specific embodiment of this system will be described below.

[1289] Hardware and Software Configuration

[1290] 1. User Device

[1291] Users enter questions using input forms on web pages or text input fields in smartphone applications. These interfaces are the means by which users provide input to the system.

[1292] 2. Server

[1293] The server plays a central role in processing queries received from users. It has the following functions:

[1294] Anonymization function: Personal information is converted based on specific rules so that it cannot be used to identify a specific individual.

[1295] Emotion engine: Analyzes emotions from the text information of the user's question. The emotion engine uses natural language processing (NLP) technology to analyze text data and recognize the user's emotions.

[1296] Generative AI: Generates answers to user questions based on the emotional information analyzed by the emotion engine. Generative AI uses advanced language models such as GPT-4.

[1297] Decryption function: Receives the answer from the generating AI and decrypts the corresponding anonymous information.

[1298] Display function: displays the generated answer to the user.

[1299] Data processing and calculation

[1300] 1. User Input

[1301] The user enters "I am very sad" into the text input field of the smartphone application. After completing the input, the user presses the send button.

[1302] 2. Sending input data

[1303] The terminal sends the text data entered by the user to the server using the HTTPS protocol. Since the data is encrypted, there is a low risk of it being intercepted by a third party.

[1304] 3. Emotion analysis using an emotion engine

[1305] The server passes the received text data to the emotion engine, which uses NLP technology to recognize the emotion "sadness" from the text "I am very sad." The emotion engine uses an emotion analysis library implemented in Python.

[1306] 4. Sending emotional information to the AI

[1307] The server sends the emotion information recognized by the emotion engine, "sadness," to the generation AI in JSON format, for example, {"emotion": "sadness"}.

[1308] 5. Generating emotion-aware answers

[1309] The generative AI generates an answer based on the emotional information it receives. For example, if it receives the emotional information "sadness," the generative AI will generate the answer "I understand how you feel. Let me know if there's anything I can help you with." The generative AI uses OpenAI's GPT-4 model.

[1310] 6. Submitting and Viewing Your Answers

[1311] The server sends the answer generated by the AI ​​to the device. The device displays the received answer to the user. The user sees a message on their smartphone screen saying, "We understand how you feel. Please let us know if there's anything we can do to help."

[1312] Specific examples

[1313] For example, if a user enters "I am very sad," the emotion engine will recognize the emotion "sadness" from this text, and the generation AI will generate a response that will comfort the user. An example of a specific prompt sentence is, "If a user enters 'I am very sad,' the emotion engine will recognize the emotion "sadness" from this text, and the generation AI will generate a response that will comfort the user."

[1314] In this way, a system is realized that provides appropriate answers that correspond to the user's emotions.

[1315] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1316] Step 1:

[1317] A user inputs a question into a text input field of a smartphone application. For example, the user inputs "I am very sad" and presses the send button. The input data is in text format.

[1318] Step 2:

[1319] The terminal sends the text data entered by the user to the server using the HTTPS protocol. Since the input data is encrypted, there is a low risk of it being intercepted by a third party. The output is encrypted text data.

[1320] Step 3:

[1321] The server passes the received text data to the emotion engine. The input is encrypted text data, which is decrypted to obtain text information. The emotion engine uses natural language processing (NLP) technology to analyze the text data and recognize the user's emotion. For example, it recognizes the emotion "sadness" from the text "I am very sad." The output is emotional information.

[1322] Step 4:

[1323] The server sends the emotion information recognized by the emotion engine in JSON format to the generation AI. The input is emotion information, which is converted into JSON format. For example, it is sent in the format {"emotion": "sadness"}. The output is emotion information in JSON format.

[1324] Step 5:

[1325] The generative AI generates an answer based on the emotional information it receives. The input is emotional information in JSON format, which is analyzed to generate an appropriate answer. For example, if the generative AI receives the emotional information "sadness," it generates the answer "I understand how you feel. Please let me know if there's anything I can do to help." The output is a text answer.

[1326] Step 6:

[1327] The server sends the answer generated by the generation AI to the terminal. The input is a text answer, which is encrypted and sent to the terminal. The output is an encrypted text answer.

[1328] Step 7:

[1329] The device decrypts the received answer and displays it to the user. The input is the encrypted text answer, which is decrypted and displayed to the user. The user sees a message on their smartphone screen saying, "We understand how you feel. Let us know if there's anything we can do to help." The output is the text answer displayed to the user.

[1330] (Application example 1)

[1331] 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."

[1332] Conventional systems often struggle to provide appropriate answers to user questions. In particular, they generate answers without considering the user's emotions, making it difficult to alleviate the user's anxiety or sadness. Furthermore, they lack sufficient anonymization and decryption methods to safely process questions containing personal information, which poses a privacy problem.

[1333] The identification process by the identification 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 a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing emotions from the user's input text, means for sending the analyzed emotion information to the generation AI and generating an answer based on the emotion, and means for displaying the generated answer to the user. This makes it possible to provide an appropriate answer that corresponds to the user's emotions and process the question safely while protecting privacy.

[1334] "Personal information" is information that can identify a specific individual.

[1335] A "question" is an inquiry that a user inputs to the system.

[1336] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1337] "Generative AI" is artificial intelligence that generates appropriate answers to questions from users.

[1338] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1339] An "emotion engine" is software or hardware for analyzing emotions from user input text.

[1340] "Emotion information" is data on the user's emotions analyzed by the emotion engine.

[1341] "Display" refers to the process of visually presenting the generated answer to the user.

[1342] A system for implementing this invention is configured as follows: First, a user inputs a question using a device such as a smartphone. Since the input question may contain personal information, the server anonymizes it. The anonymized question is sent to a generation AI, which then generates an appropriate answer.

[1343] Next, an emotion engine is used to analyze emotions from the user's input text. The emotion engine analyzes the user's text information and generates emotion information. This emotion information is sent to the generation AI, which then generates an answer based on the emotion. The generated answer is displayed to the user by the server.

[1344] Specifically, the following hardware and software are used:

[1345] Hardware: Smartphones, servers

[1346] Software: Python, OpenAI API, TextBlob

[1347] The server receives input from the user and analyzes the emotion using TextBlob. The analyzed emotion information is sent to the generation AI using OpenAI's API. The generation AI generates an appropriate answer based on the emotion information and returns the answer to the server. The server displays the generated answer to the user.

[1348] For example, if a user types, "There may be a suspicious person in my house," the emotion engine recognizes the emotion of "anxiety." Based on this emotional information, the generative AI generates the answer, "Please contact the police immediately. Please evacuate to a safe place and securely close the doors and windows."

[1349] An example of a prompt is as follows:

[1350] The user is feeling anxious. Please generate appropriate answers to the following questions: There might be a suspicious person in my house.

[1351] In this way, it is possible to provide appropriate answers that correspond to the user's feelings and process questions safely while protecting privacy.

[1352] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1353] Step 1:

[1354] A user inputs a question using a terminal. The input question is sent to the server in text format. The input data is the user's question text.

[1355] Step 2:

[1356] The server anonymizes personal information from the received question text. For anonymization, the server converts the personal information based on specific rules so that a specific individual cannot be identified. The input data is the user's question text, and the output data is the anonymized question text.

[1357] Step 3:

[1358] The server sends the anonymized question text to the generation AI, which generates an appropriate answer based on the received question text. The input data is the anonymized question text, and the output data is the generated answer text.

[1359] Step 4:

[1360] The server uses an emotion engine to analyze emotions from the user's input text. The emotion engine uses software such as TextBlob to analyze the text information and generate emotion information. The input data is the user's question text, and the output data is the analyzed emotion information.

[1361] Step 5:

[1362] The server sends the analyzed emotional information to the generation AI, which then generates an answer based on the emotion. The generation AI then generates an appropriate answer taking the emotional information into account. The input data is the analyzed emotional information and anonymized question text, and the output data is an answer text based on the emotion.

[1363] Step 6:

[1364] The server displays the generated answer text to the user, who can visually check the generated answer through the terminal. The input data is the generated answer text, and the output data is the answer displayed to the user.

[1365] In this way, it is possible to provide an appropriate answer that corresponds to the user's feelings and process questions safely while protecting privacy.

[1366] Example 2

[1367] 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."

[1368] Conventional systems do not adequately anonymize and decrypt data, including personal information, resulting in incomplete privacy protection. They also have difficulty generating appropriate answers that take user sentiment into account. This leads to problems such as a poor user experience and a loss of system reliability.

[1369] 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.

[1370] In this invention, the server includes means for receiving data including personal information and anonymizing the personal information, means for transmitting the anonymized data to the generative AI model, means for receiving a question from a user and analyzing the corresponding emotion, means for transmitting the analyzed emotion information to the generative AI model, means for receiving an answer from the generative AI model and decrypting the corresponding anonymous information, and means for transmitting the generated answer to the user. This makes it possible to provide an appropriate answer corresponding to the user's emotion while ensuring the protection of personal information.

[1371] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1372] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1373] A "generative AI model" is an algorithm or system that uses artificial intelligence technology to analyze data and generate appropriate answers or information.

[1374] "Emotion analysis" is a process of identifying emotions from text information entered by a user and extracting that emotion information.

[1375] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1376] A "user" is an individual or group that uses the system and enters questions and personal information.

[1377] A "server" is a computer that is the center of a system and is a device that receives, processes, and transmits data.

[1378] "Data" is a collection of information, expressed in the form of text, numbers, images, etc.

[1379] A "question" is an inquiry or request that a user inputs to the system.

[1380] An "answer" is a response or piece of information that a generative AI model generates in response to a user's question.

[1381] The present invention relates to a system that anonymizes data including personal information, analyzes user emotions, and generates appropriate answers. A specific embodiment of this system is described below.

[1382] The server receives data including personal information. For example, a user inputs personal information such as "Yamada Taro, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo" using a terminal. The terminal then transmits this information to the server.

[1383] The server anonymizes the received personal information. Specifically, it converts names into initials and addresses into area-level information. For example, it converts "Yamada Taro" into "YT" and "1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo" into "Shinjuku-ku, Tokyo." This anonymization process is performed based on specific rules.

[1384] Next, the user uses the terminal to input a question, for example, "I am very sad," which is then sent to the server.

[1385] The server uses an emotion engine to analyze emotions from the text information of the user's question. The emotion engine recognizes the emotion "sadness" from the input text. This emotion information is sent to the generative AI model.

[1386] The generative AI model generates a response based on the emotional information it receives. For example, if it recognizes the emotion "sadness," it will generate the response "That's very painful. Please let me know if there's anything I can do to help."

[1387] The server generates a response and sends it to the user, who receives it at his terminal.

[1388] As a concrete example, the following prompt sentence is shown.

[1389] Input: The user types "Taro Yamada, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo."

[1390] Anonymization: The server converts it to "YT, Shinjuku-ku, Tokyo."

[1391] Question input: User types "I am very sad."

[1392] Emotion analysis: The server recognizes it as "sadness."

[1393] Emotion information transmission: The server sends the emotion information of "sadness" to the generative AI model.

[1394] Answer generation: The generative AI model generates the answer, "That's very painful. Please let me know if there's anything I can do to help."

[1395] Send Answer: The server sends the generated answer to the user.

[1396] In this way, the server can provide appropriate responses that correspond to the user's emotions while protecting the user's personal information.

[1397] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1398] Step 1:

[1399] The user enters personal information.

[1400] The user uses the terminal to input personal information such as name and address. For example, they input "Yamada Taro, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo." The terminal then sends this information to the server.

[1401] Input: "Yamada Taro, 1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo"

[1402] Output: Personal information sent to the server

[1403] Step 2:

[1404] The server receives the personal information.

[1405] The server receives the personal information sent by the device and temporarily stores it for further processing.

[1406] Input: Personal information sent from the device

[1407] Output: Temporarily stored personal information

[1408] Step 3:

[1409] The server anonymizes your personal information.

[1410] The server anonymizes personal information based on specific rules. Specifically, it converts names to initials and addresses to area-level information. For example, it converts "Yamada Taro" to "YT" and "1-1-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[1411] Input: Temporarily stored personal information

[1412] Output: Anonymized personal information (e.g., "YT, Shinjuku-ku, Tokyo")

[1413] Step 4:

[1414] The user enters a question.

[1415] The user uses the terminal to input a question containing an emotion, for example, "I am very sad." The terminal then sends this question to the server.

[1416] Enter: "I'm so sad."

[1417] Output: The question sent to the server

[1418] Step 5:

[1419] The server analyzes emotions using an emotion engine.

[1420] The server uses an emotion engine to analyze emotions from the text information of the user's question. For example, it recognizes the emotion "sadness" from the text "I am very sad."

[1421] Input: The question sent to the server

[1422] Output: Parsed emotion information (e.g. "sadness")

[1423] Step 6:

[1424] The server sends the emotion information to the generative AI model.

[1425] The server sends the analyzed emotion information to the generative AI model, for example, emotion information of "sadness."

[1426] Input: Parsed emotion information

[1427] Output: Emotion information sent to the generative AI model

[1428] Step 7:

[1429] A generative AI model generates answers based on emotional information.

[1430] The generative AI model generates a response based on the emotional information it receives. For example, if it recognizes the emotion "sadness," it will generate the response "That's very painful. Please let me know if there's anything I can do to help."

[1431] Input: Emotion information sent to the generative AI model

[1432] Output: The generated answer (e.g., "That's very painful. Let me know if there's anything I can do to help you.")

[1433] Step 8:

[1434] The server sends the generated answer to the user.

[1435] The server sends the answer generated by the generative AI model to the user, who receives the answer on their device.

[1436] Input: Generated answer

[1437] Output: The answer sent to the user

[1438] In this way, the server can provide appropriate responses that correspond to the user's emotions while protecting the user's personal information.

[1439] (Application example 2)

[1440] 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."

[1441] Conventional systems have difficulty not only anonymizing and decrypting personal information, but also responding appropriately based on the user's emotions. Furthermore, responses that do not take the user's emotions into consideration can potentially reduce user satisfaction. Furthermore, there was no system that could simultaneously anonymize personal information and analyze emotions, making it difficult to balance security and user experience.

[1442] 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.

[1443] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing the user's emotions, and means for the generation AI to generate an answer based on the emotion information from the emotion engine. This makes it possible to protect personal information and respond appropriately according to the user's emotions.

[1444] "Personal information" is information that can identify a specific individual, including name, address, telephone number, etc.

[1445] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1446] "Generative AI" is an artificial intelligence system that generates appropriate answers or information based on input data.

[1447] The "emotion engine" is a system that analyzes emotions from the user's text information and sends that emotional information to the generation AI.

[1448] "Emotion information" is data that represents the user's emotions, and indicates an emotional state such as "sadness" or "happiness," for example.

[1449] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1450] A "question" is an inquiry or request that a user enters into the system.

[1451] An "answer" is a response or information that a generative AI generates in response to a user's question.

[1452] A system for implementing this invention includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to a generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing the user's emotions, and means for the generation AI to generate an answer based on the emotion information from the emotion engine.

[1453] System configuration

[1454] 1. Anonymization of personal information

[1455] The server anonymizes questions containing personal information received from users based on certain rules, such as converting names to initials and phone numbers to "XXX-XXX-XXX."

[1456] 2. Emotion analysis using an emotion engine

[1457] The server sends the anonymized questions to an emotion engine that analyzes the user's emotions. The emotion engine recognizes emotions such as "sadness" or "joy" from the text information.

[1458] 3. Answer generation using generative AI

[1459] The server sends the emotion information from the emotion engine to the generation AI, which then generates an appropriate response based on the emotion. For example, if the generation AI recognizes the emotion "sadness," it will generate a response that comforts the user.

[1460] 4. Decrypting the Answer

[1461] The server receives the answer from the generating AI and converts the anonymized information back to the original information based on specific rules.

[1462] Hardware and software used

[1463] Hardware: Servers, user devices (smartphones, smart glasses, etc.)

[1464] Software: emotion engines (e.g., sentiment-analysis models from the transformers library), generative AI (e.g., GPT-3)

[1465] Specific examples

[1466] If a user enters "I'm very sad. My name is Taro Tanaka. My phone number is 123-456-789," the server anonymizes the name and phone number and uses an emotion engine to recognize the emotion "sad." The generative AI then generates a response saying, "You seem to be in trouble. Please don't worry," and the server provides this response to the user.

[1467] Prompt Sentence Examples

[1468] User input: "I'm so sad. My name is Taro Tanaka. My phone number is 123-456-789."

[1469] Prompt the generative AI model: "Analyze the user's sentiment and generate an appropriate response."

[1470] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1471] Step 1:

[1472] The user inputs a question including personal information using a terminal. The input question is in the form of, for example, "I am very sad. My name is Taro Tanaka. My phone number is 123-456-789." The terminal sends this input to the server.

[1473] Step 2:

[1474] The server anonymizes personal information from the received questions based on specific rules. Specifically, it converts names to initials and phone numbers to "XXX-XXX-XXX." The input is "I'm so sad. My name is Tanaka Taro. My phone number is 123-456-789." The output is "I'm so sad. My name is TT. My phone number is XXX-XXX-XXX."

[1475] Step 3:

[1476] The server sends the anonymized question to the emotion engine, which analyzes emotions from text information. The input is "I'm very sad. My name is TT. My phone number is XXX-XXX-XXX." The output is the emotion information "sadness."

[1477] Step 4:

[1478] The server sends the emotion information from the emotion engine to the generation AI. The generation AI generates an appropriate response based on the emotion information. The input is the emotion information "sadness," and the output is the response "It seems you're in trouble. Please don't worry."

[1479] Step 5:

[1480] The server receives the answer from the generative AI and converts the anonymized information back to the original information based on specific rules. The input is "I see you're having trouble. Please don't worry," and the output is the answer as is.

[1481] Step 6:

[1482] The server sends the generated answer to the user's terminal, and the user receives the answer via the terminal saying, "It seems you are having trouble. Please don't worry."

[1483] Example 3

[1484] 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."

[1485] In conventional systems, to provide appropriate answers to user questions, questions containing personal information must be anonymized and sent to a generative AI model. However, these systems have the problem of not taking the user's emotions into account and being unable to generate appropriate answers based on their emotions. In addition, the decoding process required to restore anonymized information to its original form is complex, making it difficult to provide answers in a format that is easy for users to understand.

[1486] 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.

[1487] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generative AI model, means for receiving an answer from the generative AI model and decrypting the relevant anonymous information, means including an emotion engine for analyzing emotions from text information of the user's question, means for sending the emotion information analyzed by the emotion engine to the generative AI model, and means for the generative AI model to generate an answer based on the emotion information. This makes it possible to provide an appropriate answer corresponding to the user's emotion.

[1488] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1489] "Anonymization" is the process of removing or masking personal information so that a specific individual cannot be identified.

[1490] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to analyze input text and generate appropriate answers.

[1491] A "communication module" is a hardware or software component for transmitting and receiving data, and has the ability to communicate with other systems over a network.

[1492] An "emotion engine" is a software component that analyzes emotions from user input text and extracts that emotion information.

[1493] "Decryption" is the process of restoring anonymized information to its original form, and is carried out according to specific rules.

[1494] "Text information" is character string data entered by the user, and takes the form of a question, comment, or the like.

[1495] "Emotion information" is data that indicates the user's emotional state analyzed from text information, and includes, for example, "sadness" or "joy."

[1496] "Analysis" is the process of examining input data in detail to understand its meaning and structure.

[1497] An "answer" is a response message that a generative AI model generates in response to a user's question.

[1498] This invention relates to a system that anonymizes a user's question, sends it to a generative AI model, decrypts the answer from the generative AI model, and provides it to the user. It also includes a function to analyze the user's emotions and generate appropriate answers based on those emotions.

[1499] The server anonymizes the questions received from users by removing or masking personal information based on specific rules. This anonymization process can use regular expressions or data masking techniques. For example, if a user enters "I am very sad," the server will remove the "I" part and leave it as "very sad."

[1500] Next, the server sends the anonymized question to the generative AI model. This is done using a communication module, which uses Python's requests library to send an HTTP request to the generative AI model. The generative AI model analyzes the received question, and the emotion engine recognizes the user's emotion. The emotion engine uses natural language processing (NLP) techniques to extract emotion from text information. For example, it recognizes the emotion "sadness" from the text "very sad."

[1501] The generative AI model generates a response based on the recognized emotion. For example, if the emotion "sadness" is recognized, the generative AI model generates the response "Are you OK? Is there anything I can help you with?". This generation is done using a pre-trained AI model.

[1502] The server receives the answer sent by the generative AI model and uses the communication module to receive the HTTP response. The received answer is decrypted based on specific rules to restore the anonymized information to its original form. For example, the server provides the user with the answer "Are you OK? Is there anything I can help you with?"

[1503] As a concrete example, consider the case where a user types "I am very sad." This question is anonymized by the server and sent to the generative AI model through the communication module. The emotion engine recognizes the emotion "sad" and sends it to the generative AI model. Based on "sad," the generative AI model generates a response: "Are you OK? Is there anything I can help you with?" This response is sent to the server, decrypted, and provided to the user.

[1504] Example prompt sentence:

[1505] User Question: "I'm so sad"

[1506] Prompt for generative AI: "Recognize the user's emotion and generate an appropriate response. Emotion: Sadness"

[1507] Generative AI response: "Are you OK? Is there anything I can help you with?"

[1508] In this way, it becomes possible to provide an appropriate answer that corresponds to the user's feelings. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1509] Step 1:

[1510] The user enters a question.

[1511] The user enters "I am very sad" into the input field of the terminal and clicks the send button. The entered question is sent from the terminal to the server.

[1512] Step 2:

[1513] The server anonymizes the question.

[1514] The server analyzes the received question "I am very sad" and removes or masks personal information. Specifically, it uses regular expressions to remove the "I" part and writes it as "very sad." This process results in an anonymized question that does not contain personal information.

[1515] Step 3:

[1516] The server sends the anonymized questions to the generative AI model.

[1517] The server converts the anonymized question "I'm so sad" into JSON format and sends an HTTP POST request to the generative AI model using the communication module. The input is the anonymized question, and the output is the request sent to the generative AI model.

[1518] Step 4:

[1519] A generative AI model analyzes the question, and an emotion engine recognizes the emotion.

[1520] The generative AI model analyzes the received question "I'm so sad." The emotion engine uses natural language processing (NLP) techniques to extract emotions from text information. For example, it recognizes the emotion "sad" from the text "I'm so sad." The input is an anonymized question, and the output is the recognized emotion information.

[1521] Step 5:

[1522] A generative AI model generates answers based on emotions.

[1523] The generative AI model generates an answer based on the recognized emotion "sadness." For example, if the emotion "sadness" is recognized, the generative AI model generates the answer "Are you OK? Is there anything I can help you with?" The input is emotional information, and the output is the generated answer.

[1524] Step 6:

[1525] The server receives the answer from the generative AI model.

[1526] The server receives the answer sent from the generative AI model. It uses a communication module to receive the HTTP response from the generative AI model. The input is the answer from the generative AI model, and the output is the answer received by the server.

[1527] Step 7:

[1528] The server decrypts the response and provides it to the user.

[1529] The server applies certain rules in reverse to convert the answer it receives from the generative AI model back into a form the user can understand. For example, it provides the answer "Are you OK? Is there anything I can help you with?" to the user verbatim. The input is the answer from the generative AI model, and the output is the decoded answer provided to the user.

[1530] In this way, it is possible to provide an appropriate answer that corresponds to the user's feelings.

[1531] (Application example 3)

[1532] 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."

[1533] In conventional systems, when users anonymously submit questions and receive answers from generative AI, it is difficult to provide appropriate answers that take the user's emotions into consideration. In addition, there is a lack of means to properly decrypt anonymized information, making it difficult for users to receive answers in a format that is easy to understand.

[1534] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes means for receiving a question including personal information and anonymizing the corresponding personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the corresponding anonymous information, means for including an emotion engine that analyzes the user's emotions and sending the emotion information to the generation AI, and means for the generation AI to generate an answer based on the emotion information. This makes it possible to provide an appropriate answer corresponding to the user's emotions and to appropriately decrypt the anonymized information.

[1535] "Personal information" is information that can identify a specific individual.

[1536] "Anonymization" is the process of transforming information so that it can no longer identify a specific individual.

[1537] "Generative AI" is a system that uses artificial intelligence technology to generate answers based on user input.

[1538] An "emotion engine" is a system that has the ability to analyze emotions from a user's text information.

[1539] "Emotion information" is emotional data analyzed from user input.

[1540] "Decryption" is the process of restoring anonymized information to its original form.

[1541] A "question" is an inquiry that a user enters into the generated AI.

[1542] An "answer" is the response that the generation AI generates in response to a user's question.

[1543] The system for implementing the present invention uses the following hardware and software.

[1544] Hardware used

[1545] Smartphone

[1546] server

[1547] Software used

[1548] Communication Module

[1549] Emotion Engine

[1550] Generative AI models (e.g., GPT-4)

[1551] Database (storing the anonymization and decryption rules)

[1552] System configuration

[1553] 1. User Input: The user opens the smartphone app and enters their security question.

[1554] 2. Anonymization: The communication module on the smartphone anonymizes the question and converts it into a format that can be analyzed by the generative AI.

[1555] 3. Sentiment analysis: The sentiment engine analyzes the user's sentiment from the text information of the question.

[1556] 4. Send to Generative AI: Send the anonymized questions and emotion information to the Generative AI on the server.

[1557] 5. Answer generation: Generative AI generates appropriate answers based on emotional information.

[1558] 6. Decryption: The communication module on the server receives the answer from the generating AI and decrypts the anonymous information.

[1559] 7. Display to user: The decrypted answer is displayed to the user on their smartphone.

[1560] Specific examples

[1561] A user types in the question, "I've been seeing some suspicious people around my house lately. What should I do?"

[1562] Anonymization: The question is anonymized and converted to the following format: "What should you do if you see someone suspicious?"

[1563] Sentiment analysis: The emotion engine analyzes the emotion "anxiety."

[1564] Send to the generation AI: The anonymized question "What should you do if you see a suspicious person?" and the emotional information "anxiety" are sent to the generation AI.

[1565] Answer generation: The AI ​​generates the answer, "If you feel unsafe, first contact the police and ensure safety around your home."

[1566] Decoding: The communication module decodes the answer back into a format that corresponds to the original question.

[1567] What the user sees: "If you feel unsafe, please contact the police first and make sure your surroundings are safe."

[1568] Prompt Sentence Examples

[1569] User Question: I've been seeing some suspicious people around my house lately, what should I do?

[1570] Emotional information: Anxiety

[1571] Prompt for generative AI: What should you do if you see someone suspicious? (Emotion: Anxiety)

[1572] In this way, appropriate security advice can be provided that corresponds to the user's emotions.

[1573] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1574] Step 1:

[1575] The user opens the smartphone app and enters their security questions, which are then sent to the app in text format.

[1576] Step 2:

[1577] The communication module of the device anonymizes the input question. Specifically, it converts personal information based on specific rules to generate an anonymized question. The input is the user's question, and the output is the anonymized question.

[1578] Step 3:

[1579] The device's emotion engine analyzes the user's emotions from the text information of the anonymized question. The emotion engine uses natural language processing technology to extract emotional information from the text. The input is the anonymized question, and the output is emotional information.

[1580] Step 4:

[1581] The device sends anonymized questions and emotional information to the generation AI on the server. The input is the anonymized questions and emotional information, and the output is a prompt for the generation AI.

[1582] Step 5:

[1583] The server's generation AI generates an answer based on the received prompt. The generation AI takes emotional information into account and generates an appropriate answer. The input is the prompt, and the output is the generated answer.

[1584] Step 6:

[1585] The server's communication module receives the answer from the generating AI and decrypts the anonymous information. Specifically, it restores the anonymized information to its original form based on specific rules. The input is the generated answer, and the output is the decrypted answer.

[1586] Step 7:

[1587] The terminal displays the decrypted answer to the user, who can then check the answer on their smartphone. The input is the decrypted answer, and the output is the answer displayed to the user.

[1588] 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.

[1589] 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.

[1590] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[1591] 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.

[1592] [Third embodiment]

[1593] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1594] 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.

[1595] 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).

[1596] 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.

[1597] 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.

[1598] 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).

[1599] 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.

[1600] 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.

[1601] 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.

[1602] 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.

[1603] 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.

[1604] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1605] "Example 1"

[1606] In one embodiment of the present invention, an interface for accepting input from a user is provided as a means for receiving questions including personal information. This interface may be, for example, an input form on a web page or a text input field in a smartphone application.

[1607] "Example 2"

[1608] Next, as a means of anonymizing the received personal information, a function is provided to convert personal information based on specific rules, such as converting names to initials or addresses to area-level information.

[1609] "Example 3"

[1610] A communication module is provided as a means for transmitting the anonymized questions to the generation AI. This communication module converts the anonymized questions into a format that can be analyzed by the generation AI and transmits them to the generation AI.

[1611] "Example 4"

[1612] Finally, we will provide a mechanism to reverse the specific rules described above as a means of receiving the answers from the Generator AI and decrypting the relevant anonymous information, thereby returning the answers from the Generator AI in a form that can be understood by the user.

[1613] The processing flow of each embodiment will be described below.

[1614] "Example 1"

[1615] Step 1: A user enters a question into a form on a webpage, which may contain personal information.

[1616] Step 2: The system receives the question and anonymizes the personal information based on certain rules, for example, converting names to initials and addresses to locality level information.

[1617] Step 3: The anonymized question is sent to the generation AI, where a communication module converts the question into a format that the generation AI can parse.

[1618] Step 4: Receive the answer from the Generator AI and decrypt the corresponding anonymous information, specifically by applying the specific rules mentioned above in reverse to return the answer to a form that the user can understand.

[1619] Example 1

[1620] 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."

[1621] Conventional systems have had difficulty safely processing questions containing personal information and generating appropriate answers. Furthermore, the process of anonymizing and decrypting personal information is complex, resulting in delays in providing answers to users. Furthermore, there is a lack of efficient means for generating prompts suitable for generative AI models.

[1622] 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.

[1623] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for analyzing the anonymized question and generating a prompt sentence suitable for the generative AI model, means for sending the generated prompt sentence to the generative AI model, means for receiving an answer from the generative AI model and decrypting the relevant anonymous information, and means for displaying the decrypted answer to the user. This makes it possible to process personal information safely and provide the user with a prompt and appropriate answer.

[1624] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1625] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1626] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate appropriate answers based on input prompt sentences.

[1627] A "prompt sentence" is a sentence input to a generative AI model that appropriately expresses the user's question.

[1628] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1629] A "server" is a computer system that processes and stores data over a network.

[1630] A "user" is an individual or organization that uses the system to enter questions and receive answers.

[1631] This invention is a system that allows a user to input a question that includes personal information and provides an appropriate answer to the question using a generative AI model. Specific embodiments of this system are described below.

[1632] A user enters a question using a form on a web page or a text input field in a smartphone application, for example, "My name is Taro Tanaka. My address is Shinjuku-ku, Tokyo. How can I protect my personal information?"

[1633] The device sends the entered question to the server using the HTTPS protocol, with the data encrypted.

[1634] The server analyzes the received questions and anonymizes the personal information. Anonymization involves converting personal information based on specific rules. For example, "Taro Tanaka" could be "User A" and "Shinjuku Ward, Tokyo" could be "Area X."

[1635] The server then analyzes the anonymized question and generates a prompt suitable for the generative AI model, such as:

[1636] The user's name is User A and they live in Region X. How do you protect their personal information?

[1637] The server sends the generated prompt to a generative AI model, such as OpenAI's GPT-3 or GPT-4.

[1638] The generative AI model generates an answer based on the prompt it receives, for example, the following answer:

[1639] You can help protect your personal information by:

[1640] 1. Use strong passwords.

[1641] 2. Only share personal information with people you trust.

[1642] 3. Install security software and keep it up to date.

[1643] 4. Be careful when using public Wi-Fi.

[1644] The server decrypts the answer received from the generative AI model and restores the original personal information. For example, it restores "User A" to "Taro Tanaka" and "Area X" to "Shinjuku Ward, Tokyo."

[1645] Finally, the server displays the decrypted answers to the user, who can then check them on a web page or smartphone application.

[1646] This system allows users to safely process their personal information and receive prompt and appropriate answers.

[1647] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1648] Step 1:

[1649] The user enters a question.

[1650] A user types a question into a text field on a web page or smartphone application, for example, "My name is Taro Tanaka. My address is Shinjuku-ku, Tokyo. How can I protect my personal information?"

[1651] Input: User question

[1652] Output: Question data entered

[1653] Step 2:

[1654] The terminal sends a question to the server.

[1655] The device sends the entered question to a server using the HTTPS protocol, with the data encrypted.

[1656] Input: The entered question data

[1657] Output: The query data sent to the server

[1658] Step 3:

[1659] The server analyzes the question and anonymizes personal information.

[1660] The server analyzes the received questions and anonymizes personal information based on specific rules. For example, "Taro Tanaka" is "User A" and "Shinjuku Ward, Tokyo" is "Area X."

[1661] Input: Query data sent to the server

[1662] Output: Anonymized question data

[1663] Step 4:

[1664] The server analyzes the anonymized questions and generates prompts appropriate for the generative AI model.

[1665] The server analyzes the anonymized question and generates a prompt appropriate for the generative AI model, such as: "The user's name is User A and their address is in Region X. Please tell me how to protect their personal information."

[1666] Input: Anonymized question data

[1667] Output: Generated prompt statement

[1668] Step 5:

[1669] The server sends the generated prompt to the generative AI model.

[1670] The server sends the generated prompt to a generative AI model, which uses natural language processing technology.

[1671] Input: Generated prompt text

[1672] Output: The prompt sent to the generative AI model

[1673] Step 6:

[1674] A generative AI model generates the answer.

[1675] The generative AI model generates answers based on the prompt it receives, such as: "To protect your personal information, you can: 1. Use strong passwords. 2. Only share personal information with people you trust. 3. Install security software and keep it up to date. 4. Exercise caution when using public Wi-Fi."

[1676] Input: The prompt sent to the generative AI model

[1677] Output: The generated answer

[1678] Step 7:

[1679] The server receives the answer from the generative AI model and decrypts the anonymous information.

[1680] The server decrypts the answer received from the generative AI model and restores the original personal information. For example, it restores "User A" to "Taro Tanaka" and "Area X" to "Shinjuku Ward, Tokyo."

[1681] Input: Generated Answer

[1682] Output: Decoded answer

[1683] Step 8:

[1684] The server displays the decrypted answer to the user.

[1685] The server then displays the decrypted answers to the user, who can then check them on a web page or smartphone application.

[1686] Input: Decrypted answer

[1687] Output: The answer displayed to the user

[1688] (Application example 1)

[1689] 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."

[1690] There is a need to provide an environment where questions containing personal information can be handled safely and where users can enter questions with peace of mind. However, conventional systems have a high risk of personal information leakage and unauthorized access, making it difficult to adequately protect user privacy. Furthermore, even when providing answers using generative AI models, there is a lack of means to properly anonymize and decrypt users' personal information.

[1691] 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.

[1692] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means for decrypting the encrypted data and inputting it as a prompt sentence to the generation AI model, and means for obtaining an answer from the generation AI model and providing it to the user. This makes it possible to provide an appropriate answer using the generation AI model while safely handling the user's personal information.

[1693] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1694] A "question" is a sentence or phrase entered by a user to request information.

[1695] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1696] "Generative AI" is a system that uses artificial intelligence technology to generate appropriate answers to users' questions.

[1697] "Decryption" is the process of restoring anonymized information to its original form.

[1698] "Encryption" is the process of converting data using a specific algorithm to make it unreadable to third parties.

[1699] A "prompt sentence" is a sentence input to a generative AI model, and is an instruction sentence that enables the AI ​​to generate an appropriate answer.

[1700] An "answer" is information generated by the generative AI in response to a user's question.

[1701] "User" means an individual or entity that utilizes the System to enter questions and receive answers.

[1702] As an embodiment of the present invention, the following system is constructed.

[1703] System Configuration

[1704] The system includes a user terminal, a server, and a generative AI model. The user terminal is a device such as a smartphone or PC, and provides an interface for users to input questions. The server receives questions from users, anonymizes them, encrypts them, decrypts them, generates prompts for the generative AI model, and provides answers.

[1705] Hardware and software used

[1706] Hardware: smartphones, PCs, servers

[1707] Software: Python, OpenAI API, cryptography library

[1708] Data processing and calculation

[1709] 1. User Device:

[1710] The user enters a question through an interface on the device, which is an application that includes a text entry field.

[1711] 2. Server:

[1712] When the server receives a question from a user, it first anonymizes the personal information. The anonymized question is then encrypted and sent to a generative AI model. The generative AI model generates an appropriate answer based on the prompt. The generated answer is then decrypted by the server and provided to the user.

[1713] Specific examples

[1714] If a user types the question "How do I manage my passwords securely?", the system will:

[1715] 1. Enter the question "How do I manage my passwords securely?" on the user device.

[1716] 2. The server receives this question and anonymizes your personal information.

[1717] 3. The anonymized question is encrypted and sent as a prompt to the generative AI model.

[1718] 4. The generative AI model generates an answer based on the prompt.

[1719] 5. The server decrypts the generated answer and provides it to the user.

[1720] Prompt Sentence Examples

[1721] Users ask: How do I keep my passwords safe?

[1722] Please provide security advice.

[1723] This system allows users to receive appropriate answers using generative AI models while keeping their personal information safe.

[1724] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1725] Step 1:

[1726] Enter the question on the user's device.

[1727] Users enter their questions into a text entry field on a smartphone or PC application, and the entered questions are sent to the server.

[1728] Input: A question entered by the user (e.g., "How do I keep my passwords safe?")

[1729] Output: The question data sent to the server

[1730] Step 2:

[1731] The server receives the question and anonymizes the personal information.

[1732] The server anonymizes personal information from the received questions based on specific rules, for example, masking personal information such as name and address.

[1733] Input: Question data submitted by the user

[1734] Output: Anonymized question data

[1735] Step 3:

[1736] The server encrypts the anonymized question.

[1737] The server encrypts the anonymized question data using an encryption algorithm, ensuring the security of the data.

[1738] Input: Anonymized question data

[1739] Output: Encrypted question data

[1740] Step 4:

[1741] The server sends the encrypted question to the generative AI model.

[1742] The server decrypts the encrypted question data and sends it to the generative AI model as a prompt, which is an instruction for the generative AI model to generate an appropriate answer.

[1743] Input: Encrypted question data

[1744] Output: The prompt sent to the generative AI model

[1745] Step 5:

[1746] A generative AI model generates an answer based on the prompt.

[1747] The generative AI model analyzes the prompt and generates an appropriate answer, which is then sent to the server.

[1748] Input: The prompt sent to the generative AI model

[1749] Output: Answer data from the generative AI model

[1750] Step 6:

[1751] The server receives and decodes the answer from the generative AI model.

[1752] The server decrypts the answer data received from the generative AI model and returns it to its original format.

[1753] Input: Answer data from the generative AI model

[1754] Output: Decoded answer data

[1755] Step 7:

[1756] The server provides the decrypted answer to the user.

[1757] The server sends the decrypted answer data to the user's terminal and provides it to the user, who can then check the answer on the terminal.

[1758] Input: Decrypted answer data

[1759] Output: Answer data displayed on the user's device

[1760] Example 2

[1761] 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."

[1762] When inputting data including personal information into a generative AI model, it is necessary to obtain useful responses from the generative AI model while reducing the risk of personal information leakage. However, conventional systems have not been able to efficiently anonymize personal information and input it into the generative AI model, making it difficult to achieve both security and convenience.

[1763] 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.

[1764] In this invention, the server includes means for receiving data including personal information and anonymizing the personal information, means for transmitting the anonymized data to the generative AI model, means for receiving a response from the generative AI model and generating a prompt sentence, means for inputting the generated prompt sentence into the generative AI model, and means for receiving a response from the generative AI model. This makes it possible to efficiently obtain useful responses from the generative AI model while reducing the risk of personal information leaks.

[1765] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1766] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1767] A "generative AI model" is an artificial intelligence model that performs natural language processing and generates a response based on an input prompt.

[1768] A "prompt" is a document that serves as input to a generative AI model and contains instructions and information that enable the model to generate an appropriate response.

[1769] A "server" is a computer system that processes and stores data over a network.

[1770] "Data" means a collection of information, including personal information and other related information.

[1771] A "response" is the output result that a generative AI model generates based on a prompt sentence.

[1772] "Transmitting" is the act of moving data from one place to another.

[1773] "Receiving" is the act of receiving transmitted data.

[1774] "Conversion" is the process of changing the format or content of data into a different format or content.

[1775] MODE FOR CARRYING OUT THE INVENTION

[1776] This invention relates to a system that anonymizes data including personal information, generates prompt sentences to be input into a generative AI model, and receives responses from the generative AI model. Specific embodiments of this system are described below.

[1777] Hardware and software used

[1778] Hardware: High-performance server (e.g. general-purpose server)

[1779] Software: Python, Pandas, Numpy, generative AI models (e.g., natural language processing models)

[1780] System configuration

[1781] 1. Receipt of personal information

[1782] The server receives personal information submitted by the user, such as name and address information entered by the user into a web form.

[1783] Specific operation: The server receives an HTTP request and extracts personal information from the request body.

[1784] 2. Application of anonymization rules

[1785] The server applies anonymization rules to the personal information it receives, for example, converting names to initials and addresses to locality level information.

[1786] Specific operation: The server runs a Python script to convert the name "Taro Tanaka" to "TT" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[1787] 3. Save the converted data

[1788] The server stores the anonymized data in a database, which allows it to be retrieved later if needed.

[1789] Specific operation: The server saves the converted data in the SQL database using an INSERT statement.

[1790] 4. Prompt generation

[1791] The server generates prompts to input to the generative AI model, for example, prompts containing anonymized names and addresses.

[1792] What happens: The server uses Python's string manipulation functions to generate a prompt like this:

[1793] Name:TT

[1794] Address: Shinjuku Ward, Tokyo

[1795] Use this information to generate a user profile.

[1796] 5. Input to the generative AI model

[1797] The server inputs the generated prompt sentence into a generative AI model and receives a response from the model. For example, it inputs a prompt sentence using a natural language processing model and obtains the generated text.

[1798] Specific operation: The server calls the API of the generative AI model, sends a prompt sentence, and receives a response.

[1799] In this way, the server can anonymize personal information, generate prompts to input into the generative AI model, and receive responses from the model. This system makes it possible to efficiently obtain useful responses from the generative AI model while reducing the risk of personal information leaks.

[1800] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1801] Step 1:

[1802] The server receives the personal information sent by the user.

[1803] Input: Personal information such as name and address that a user enters into a web form.

[1804] Specific operation: The server receives an HTTP request and extracts personal information from the request body.

[1805] Output: Extracted personal information (e.g., name "Taro Tanaka", address "Nishi-Shinjuku, Shinjuku-ku, Tokyo").

[1806] Step 2:

[1807] The server applies anonymization rules to the received personal information.

[1808] Input: Personal information extracted in step 1.

[1809] Specific operation: The server runs a Python script to convert the name "Taro Tanaka" to "TT" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo" to "Shinjuku-ku, Tokyo."

[1810] Output: Anonymized data (e.g., name "TT", address "Shinjuku-ku, Tokyo").

[1811] Step 3:

[1812] The server stores the anonymized data in a database.

[1813] Input: Anonymized data from step 2.

[1814] Specific operation: The server saves the converted data in the SQL database using an INSERT statement.

[1815] Output: Anonymized data stored in a database.

[1816] Step 4:

[1817] The server generates prompt sentences to input into the generative AI model.

[1818] Input: Anonymized data from step 2.

[1819] Specific operation: The server generates a prompt sentence using Python's string manipulation functions.

[1820] Output: Generated prompt text (e.g. "Name: TT Address: Shinjuku-ku, Tokyo Please generate a user profile based on this information.").

[1821] Step 5:

[1822] The server inputs the generated prompt sentence into the generative AI model and receives a response from the model.

[1823] Input: The prompt statement generated in step 4.

[1824] Specific operation: The server calls the API of the generative AI model, sends a prompt sentence, and receives a response.

[1825] Output: The response from the generative AI model (e.g., a generated user profile).

[1826] (Application example 2)

[1827] 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."

[1828] Safely handling data that includes personal information is extremely important from the perspective of privacy protection. However, transmitting personal information as is carries the risk of information leakage. In particular, when using generative AI models, anonymization is necessary because there is a possibility that personal information may be passed on to third parties. Furthermore, the accuracy of restoring anonymized information to its original form is also important. To solve these issues, a system is needed that anonymizes personal information based on specific rules and allows the safe use of generative AI models.

[1829] 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.

[1830] In this invention, the server includes means for receiving questions containing personal information and anonymizing the relevant personal information, means for sending the anonymized questions to the generation AI, means for receiving answers from the generation AI and decrypting the relevant anonymous information, means for converting names to initials, means for converting addresses to area-level information, and means for anonymizing other personal information based on specific rules. This makes it possible to safely anonymize personal information and reduce the risk of information leakage when using a generative AI model.

[1831] "Personal information" is information that can identify a specific individual.

[1832] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1833] "Generative AI" is a system that uses artificial intelligence technology to generate new data based on input data.

[1834] "Initials" is an abbreviation that refers to the first letter of each part of a name.

[1835] "Area-level information" is information that indicates a wide area without including detailed address information.

[1836] "Decryption" is the process of restoring anonymized information to its original form.

[1837] "Specific Rules" are the set of rules used to anonymize or decrypt personal information.

[1838] A "question" is part of the data input to the generating AI and is the information used to obtain an answer.

[1839] An "answer" is the output data that the generative AI generates based on the question.

[1840] A system for implementing this invention includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to a generation AI, means for receiving a response from the generation AI and decrypting the relevant anonymous information, means for converting names to initials, means for converting addresses to area-level information, and means for anonymizing other personal information based on specific rules.

[1841] System configuration

[1842] The server receives a question containing personal information entered by the user. The received personal information is anonymized based on specific rules. For example, names are converted to initials, and addresses are converted to area-level information. The anonymized question is sent to the generation AI. The generation AI generates an answer based on the question and sends the answer back to the server. The server receives the answer from the generation AI and decrypts the anonymized information back to the original information.

[1843] Hardware and software used

[1844] The system uses mobile devices such as smartphones and tablets, and the software is implemented using Python, including functions for anonymizing personal information and an API for communicating with the generative AI model.

[1845] Specific examples

[1846] For example, if a user enters the name "Yamada Taro" and the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo," the server converts this to "YT" and "Tokyo." The converted information is sent to the generation AI, which generates an answer to the question. The generated answer is sent back to the server, which decrypts the anonymized information to its original form.

[1847] Prompt Sentence Examples

[1848] An example of a prompt to input to a generative AI model is as follows:

[1849] Create an application that anonymizes personal information entered by users, converting names to initials and addresses to locality level information. Implement it using Python.

[1850] In this way, it is possible to safely anonymize personal information and reduce the risk of information leakage when using generative AI models.

[1851] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1852] Step 1:

[1853] The user inputs a question containing personal information into the terminal.

[1854] Input: The name, address, and question entered by the user.

[1855] Output: Question data including personal information entered into the terminal.

[1856] Specific operation: The user enters the name "Yamada Taro," the address "Nishi-Shinjuku, Shinjuku-ku, Tokyo," and the question into the input form on the terminal.

[1857] Step 2:

[1858] The device anonymizes your personal information.

[1859] Input: Question data, including personal information entered by the user.

[1860] Output: Anonymized question data.

[1861] Specific operations: The device will convert your name to "YT" and your address to "Tokyo." Other personal information will also be anonymized based on specific rules.

[1862] Step 3:

[1863] The device sends the anonymized question to the server.

[1864] Input: Anonymized question data.

[1865] Output: Anonymized question data sent to the server.

[1866] Specific operation: The terminal transmits data over the network to send the anonymized question data to the server.

[1867] Step 4:

[1868] The server sends the anonymized questions to the generation AI.

[1869] Input: Anonymized question data received by the server.

[1870] Output: Anonymized question data sent to the generative AI.

[1871] Specific operation: The server calls the API to send the anonymized question data to the generative AI model.

[1872] Step 5:

[1873] Generative AI generates answers to questions.

[1874] Input: Anonymized question data sent to the generative AI.

[1875] Output: Answer data generated by the generation AI.

[1876] How it works: The generative AI model analyzes anonymized question data and generates appropriate answers.

[1877] Step 6:

[1878] The server receives the answer from the generation AI.

[1879] Input: Answer data generated by the generation AI.

[1880] Output: The response data received by the server.

[1881] Specific operation: The server obtains data through an API to receive answer data from the generative AI model.

[1882] Step 7:

[1883] The server decrypts the anonymized information.

[1884] Input: The answer data received from the generation AI.

[1885] Output: The decoded answer data.

[1886] Specific operation: The server restores the anonymized information to its original form based on specific rules. For example, it decrypts "YT" into "Yamada Taro" and "Tokyo" into "Tokyo, Shinjuku Ward, Nishi-Shinjuku."

[1887] Step 8:

[1888] The server sends the decrypted response to the terminal.

[1889] Input: Decoded answer data.

[1890] Output: The decrypted answer data sent to the device.

[1891] Specific operation: The server transmits the decrypted response data to the terminal through the network.

[1892] Step 9:

[1893] The terminal displays the decrypted answer to the user.

[1894] Input: The decrypted response data received by the terminal.

[1895] Output: The decoded answer displayed to the user.

[1896] Specific operation: The terminal displays the decrypted answer data on the screen to show it to the user.

[1897] Example 3

[1898] 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."

[1899] Conventional systems have the problem of insufficient privacy protection because they send queries containing personal information directly to the generative AI model. Furthermore, there is also the issue that the responses from the generative AI model are difficult for users to understand if they remain anonymized. Furthermore, the process of converting the data into a format that the generative AI model can analyze is complex, requiring efficient data processing.

[1900] 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.

[1901] In this invention, the server includes means for receiving an inquiry including personal information and anonymizing the relevant personal information, means for sending the anonymized inquiry to the generative AI model, means for receiving a response from the generative AI model and decrypting the relevant anonymous information, a communication module for converting the response into a format that can be analyzed by the generative AI model, and means for providing the decrypted response to the user. This makes it possible to provide a response in a format that is easy for the user to understand while ensuring the protection of personal information.

[1902] "Personal information" is information that can identify a specific individual.

[1903] An "inquiry" is a question or request made by a user to the system.

[1904] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1905] A "generative AI model" is an algorithm or system that uses artificial intelligence to analyze input data and generate an appropriate response.

[1906] A "communications module" is hardware or software for transmitting and receiving data.

[1907] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1908] "User" means any individual or entity that uses the System.

[1909] A "server" is a computer system that processes and stores data on a network.

[1910] A "response" is the answer or information provided by a generative AI model as a result of its analysis.

[1911] MODE FOR CARRYING OUT THE INVENTION

[1912] This invention is a system that anonymizes inquiries that include personal information, sends them to a generative AI model, and decrypts the responses from the generative AI model and provides them to the user. Specific embodiments of this system are described below.

[1913] Hardware and software used

[1914] Hardware: Servers, communication modules, user devices (PCs, smartphones, etc.)

[1915] Software: Generative AI models (e.g., OpenAI GPT-4), programs implementing specific rules for anonymization and decryption

[1916] System Operation

[1917] 1. The user enters a question

[1918] A user uses a terminal to input a question into the system's interface, for example, "My name is Taro Tanaka. How should I learn programming?"

[1919] 2. The server receives the query

[1920] The server receives the questions sent by the user and stores them in their original format.

[1921] 3. The server anonymizes the question

[1922] The server anonymizes personal information and specific identifying information contained in the question based on specific rules, for example, converting "My name is Tanaka Taro" into "My name is [anonymous]."

[1923] 4. The server sends the anonymized questions to the generating AI via the communication module.

[1924] The server sends the anonymized questions to the Generator AI using a communication module, which converts the questions into a format that the Generator AI can parse.

[1925] 5. Generative AI analyzes the question and generates an answer

[1926] The generative AI analyzes the question it receives and generates an appropriate answer. For example, in response to the question, "How should I learn programming?", it generates the answer, "To learn programming, you should first choose a basic language and then proceed with your studies using online courses and books."

[1927] 6. The server receives the answer from the generation AI

[1928] The server receives the answers sent by the generation AI and stores them in their original format.

[1929] 7. The server decrypts the response

[1930] The server decrypts the anonymous information contained in the answer from the AI ​​generator based on specific rules. For example, it converts the answer "For [anonymous], in order to learn programming..." into "For Taro Tanaka, in order to learn programming..."

[1931] 8. The server provides the decrypted answer to the user

[1932] The server sends the decrypted answer to the user's terminal, where the user can check the answer through the terminal interface.

[1933] Specific examples

[1934] Example 1: A user submits the question "My name is Taro Tanaka. How can I learn programming?"

[1935] 1. The user uses the terminal to enter a question and clicks the submit button.

[1936] 2. The server receives the question, anonymizes it, and converts it into "My name is [Anonymous]. How can I learn programming?"

[1937] 3. Send the anonymized question to the generation AI via the communication module.

[1938] 4. In response to the question, "How should I learn programming?", the generative AI generates the answer, "To learn programming, it's a good idea to first choose a basic language and then progress through online courses and books."

[1939] 5. The server receives the answer from the generating AI and provides it to the user as is, since there is no need to decrypt the anonymous information.

[1940] Example prompt sentence:

[1941] User asks: "My name is Taro Tanaka. How can I learn programming?"

[1942] Anonymized question: "My name is [Anonymous]. How do I learn to code?"

[1943] Prompt for generative AI: "How do I learn to code?"

[1944] Generative AI's answer: "To learn programming, start by choosing a basic language and then work your way up using online courses and books."

[1945] In this way, the system anonymizes the user's question, sends it to the generation AI, and decrypts the answer from the generation AI and provides it to the user. The flow of the identification process in Example 3 will be described with reference to FIG.

[1946] Step 1:

[1947] The user enters a question

[1948] Input: The user uses the terminal to type a question, for example, "My name is Taro Tanaka. How should I learn programming?"

[1949] Specific actions: The user enters a question into the device interface and clicks the send button.

[1950] Output: The entered question is sent from the terminal to the server.

[1951] Step 2:

[1952] The server receives the query

[1953] Input: The question sent from the terminal.

[1954] Specific operation: The server receives the question sent by the user. The received question is saved in its original format.

[1955] Output: The received questions are saved on the server.

[1956] Step 3:

[1957] The server anonymizes the question

[1958] Input: A question stored on the server.

[1959] Specific operation: The server anonymizes personal information and specific identifying information included in the question based on specific rules. For example, it converts "My name is Tanaka Taro" into "My name is [anonymous]."

[1960] Output: Anonymized questions are generated.

[1961] Step 4:

[1962] The server sends the anonymized questions to the generating AI via a communication module.

[1963] Input: Anonymized question.

[1964] Specific operation: The server sends the anonymized question to the generation AI using a communication module, which converts the question into a format that the generation AI can analyze.

[1965] Output: The question is sent to the Generator AI in a format that the Generator AI can parse.

[1966] Step 5:

[1967] Generative AI analyzes the question and generates an answer

[1968] Input: Anonymized question sent to the generative AI.

[1969] How it works: Generative AI analyzes the questions it receives and generates appropriate answers. For example, in response to the question, "How should I learn programming?", it generates the answer, "To learn programming, first choose a basic language and then study using online courses and books."

[1970] Output: The generated answer is sent to the server.

[1971] Step 6:

[1972] The server receives the answer from the generation AI.

[1973] Input: The answer sent by the generation AI.

[1974] Specific operation: The server receives the answer sent from the generation AI. The received answer is saved in its original format.

[1975] Output: The received answer is saved on the server.

[1976] Step 7:

[1977] The server decrypts the response

[1978] Input: The answer sent by the generation AI.

[1979] Specific operation: The server decrypts the anonymous information included in the answer from the generation AI based on specific rules. For example, the answer "For [anonymous], in order to learn programming..." is converted into "For Taro Tanaka, in order to learn programming..."

[1980] Output: The decoded answer is generated.

[1981] Step 8:

[1982] The server provides the decrypted answer to the user

[1983] Input: The decrypted answer.

[1984] Specific operation: The server sends the decrypted answer to the user's device, where the user can check the answer through the device interface.

[1985] Output: The decrypted answer is displayed on the user's terminal.

[1986] (Application example 3)

[1987] 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 headset type terminal 314 will be referred to as a "terminal."

[1988] In conventional systems, when sending questions containing personal information to the AI ​​generator, there was a high risk of personal information being leaked, and user privacy was not protected. Another issue was the difficulty of accurately decrypting anonymized information when returning the AI ​​generator's answers to a format that users could understand. Furthermore, there was a lack of efficient ways to perform these processes on smartphones.

[1989] 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.

[1990] In this invention, the server includes means for receiving questions that include personal information and anonymizing the relevant personal information, means for sending the anonymized questions to the generation AI, means for receiving answers from the generation AI and decrypting the relevant anonymous information, and means, as an application installed on a smartphone, for sending the anonymized questions to the generation AI and decrypting the answers from the generation AI and providing them to the user. This makes it possible to efficiently realize a question-answering system that uses the generation AI on a smartphone while protecting user privacy.

[1991] "Personal information" is information that can identify a specific individual, including name, address, telephone number, email address, etc.

[1992] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[1993] "Generative AI" is an artificial intelligence model that generates natural language answers based on given input data.

[1994] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[1995] A "smartphone" is a type of mobile phone, a multi-function device that can connect to the Internet and install applications.

[1996] An "application" is software installed on a smartphone to provide specific functions or services.

[1997] A "question" is text data that a user enters to request information or advice from the generated AI.

[1998] An "answer" is text data generated by the generation AI in response to a question.

[1999] A "user" is an individual who uses the system to submit questions to the generating AI and receive answers.

[2000] As an embodiment of the present invention, the application "SecureQuery" installed on a smartphone is used. A specific implementation method thereof will be described below.

[2001] First, a user launches the SecureQuery application on their smartphone. The application provides an interface for entering a question, including personal information. Once the user enters the question, the application anonymizes it.

[2002] The anonymization process involves converting personal information based on specific rules. For example, personal information such as usernames and addresses are replaced with generic terms like "anonymous" or "private." These anonymized questions are then converted into a format that can be analyzed by the generative AI.

[2003] The application then sends the anonymized question to a Generative AI, which generates a natural language answer based on the given question. This Generative AI is located on a server in the cloud and accessed via the internet.

[2004] The answer from the AI ​​generator is sent back to the application, which receives the answer and performs a decryption process to restore the anonymized information to its original form based on specific rules. This allows the user to receive the answer from the AI ​​generator in a format that they can understand.

[2005] The system is implemented using the following hardware and software:

[2006] Hardware: Smartphone

[2007] Software: Python, requests library, JSON

[2008] The specific data processing and calculation flow is as follows:

[2009] 1. Anonymization processing: Personal information from questions entered by the user is anonymized within the application based on specific rules.

[2010] 2. Send to Generator: Send the anonymized question to the Generator endpoint, which analyzes the question and generates an answer.

[2011] 3. Decryption process: Receives the answer from the generation AI and restores the anonymized information to its original form based on specific rules.

[2012] For example, if a user enters the question, "What should I do if my username password is leaked?", this question is anonymized as, "What should I do if my anonymous password is leaked?" The Generative AI generates an appropriate answer to this question, and the application decrypts the answer and provides it to the user.

[2013] Example prompt sentence:

[2014] "What should I do if my username password is compromised?"

[2015] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[2016] Step 1:

[2017] The user launches the "SecureQuery" application on their smartphone and enters a question.

[2018] Input: The question text entered by the user (e.g., "What should I do if my username's password is leaked?")

[2019] Output: The question text entered

[2020] Specific behavior: The user enters a question into the application's input field and presses the "Submit" button.

[2021] Step 2:

[2022] The device anonymizes the entered question.

[2023] Input: The question text entered by the user

[2024] Output: Anonymized question text (e.g., "What should I do if my anonymous password is leaked?")

[2025] What it does: An anonymization module within the application replaces personal information with generic terms based on specific rules.

[2026] Step 3:

[2027] The device sends anonymized questions to the generation AI.

[2028] Input: Anonymized question text

[2029] Output: Question text sent to the generation AI

[2030] What it does: Your application makes an HTTP request and sends an anonymized question to the Generative AI endpoint.

[2031] Step 4:

[2032] The server uses generative AI to generate answers to questions.

[2033] Input: Anonymized question text

[2034] Output: Generated answer text

[2035] How it works: The generation AI analyzes the question and generates an appropriate answer, which is then sent back to the device from the server.

[2036] Step 5:

[2037] The device receives the answer from the generating AI and decodes it.

[2038] Input: Answer text returned by the generation AI (e.g., "If your anonymous password is leaked, please change it immediately.")

[2039] Output: Decrypted answer text (e.g., "If your username password has been leaked, please change it immediately.")

[2040] What it does: A decryption module within the application converts the anonymized information back into the original information based on specific rules.

[2041] Step 6:

[2042] The terminal displays the decrypted answer to the user.

[2043] Input: Decrypted answer text

[2044] Output: The answer text that is displayed to the user

[2045] What it does: The application displays the decrypted answer on the screen for the user to review.

[2046] 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.

[2047] "Example 1"

[2048] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[2049] "Example 2"

[2050] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[2051] "Example 3"

[2052] In one embodiment of the present invention, the generation AI includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from the text information of the user's question. For example, if the user enters the question "I am very sad," the emotion engine recognizes the emotion "sadness" from this text. This emotion information is then sent to the generation AI. The generation AI generates an answer based on this emotion information. For example, if the emotion "sadness" is recognized, the generation AI generates an answer that comforts the user. This makes it possible to provide a more appropriate answer that corresponds to the user's emotions.

[2053] The processing flow of each embodiment will be described below.

[2054] "Example 1"

[2055] Step 1: The user enters a question, which is sent to the system as text information.

[2056] Step 2: The emotion engine analyzes the emotion from this text information. For example, it recognizes the emotion "sadness" from the question "I am very sad."

[2057] Step 3: The emotion engine sends the analyzed emotion information to the generation AI.

[2058] Step 4: The generative AI generates an answer based on the emotional information received from the emotion engine. For example, if the generative AI recognizes the emotion "sadness," it generates an answer that comforts the user.

[2059] Step 5: The generation AI provides the generated answer to the user, which allows it to provide a more appropriate answer that reflects the user's emotions.

[2060] Example 1

[2061] 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."

[2062] Conventional systems have had difficulty providing appropriate answers that reflect the user's emotions. Furthermore, they lacked the means to safely process questions containing personal information and to anonymize and decrypt them. This resulted in low user satisfaction and created privacy issues.

[2063] 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.

[2064] In this invention, the server includes means for receiving a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means for providing an interface for accepting input from a user, means for including an emotion engine for analyzing emotions from text information of the user's question, means for sending the emotion information analyzed by the emotion engine to the generation AI, means for the generation AI to generate an answer based on the emotion information, and means for displaying the generated answer to the user. This makes it possible to provide an appropriate answer corresponding to the user's emotion, and to enable safe processing of personal information and privacy protection.

[2065] "Personal information" is information that can identify a specific individual.

[2066] A "question" is an inquiry entered by a user.

[2067] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[2068] "Generative AI" is a system that uses artificial intelligence technology to generate answers to users' questions.

[2069] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[2070] An "interface" is a means by which a user inputs information into a system, and includes input forms on web pages and text input fields in smartphone applications.

[2071] The "emotion engine" is a system that analyzes emotions from the text information of users' questions.

[2072] "Emotion information" is information about the user's emotions obtained as a result of analysis by the emotion engine.

[2073] An "answer" is the response that the generation AI generates in response to a user's question.

[2074] "Display" refers to the process of visually presenting the generated answer to the user.

[2075] MODE FOR CARRYING OUT THE INVENTION

[2076] The present invention is a system for receiving a question from a user and generating an appropriate answer based on emotions to the question. A specific embodiment of this system will be described below.

[2077] Hardware and Software Configuration

[2078] 1. User Device

[2079] Users enter questions using input forms on web pages or text input fields in smartphone applications. These interfaces are the means by which users provide input to the system.

[2080] 2. Server

[2081] The server plays a central role in processing queries received from users. It has the following functions:

[2082] Anonymization function: Personal information is converted based on specific rules so that it cannot be used to identify a specific individual.

[2083] Emotion engine: Analyzes emotions from the text information of the user's question. The emotion engine uses natural language processing (NLP) technology to analyze text data and recognize the user's emotions.

[2084] Generative AI: Generates answers to user questions based on the emotional information analyzed by the emotion engine. Generative AI uses advanced language models such as GPT-4.

[2085] Decryption function: Receives the answer from the generating AI and decrypts the corresponding anonymous information.

[2086] Display function: displays the generated answer to the user.

[2087] Data processing and calculation

[2088] 1. User Input

[2089] The user enters "I am very sad" into the text input field of the smartphone application. After completing the input, the user presses the send button.

[2090] 2. Sending input data

[2091] The terminal sends the text data entered by the user to the server using the HTTPS protocol. Since the data is encrypted, there is a low risk of it being intercepted by a third party.

[2092] 3. Emotion analysis using an emotion engine

[2093] The server passes the received text data to the emotion engine, which uses NLP technology to recognize the emotion "sadness" from the text "I am very sad." The emotion engine uses an emotion analysis library implemented in Python.

[2094] 4. Sending emotional information to the AI

[2095] The server sends the emotion information recognized by the emotion engine, "sadness," to the generation AI in JSON format, for example, {"emotion": "sadness"}.

[2096] 5. Generating emotion-aware answers

[2097] The generative AI generates an answer based on the emotional information it receives. For example, if it receives the emotional information "sadness," the generative AI will generate the answer "I understand how you feel. Let me know if there's anything I can help you with." The generative AI uses OpenAI's GPT-4 model.

[2098] 6. Submitting and Viewing Your Answers

[2099] The server sends the answer generated by the AI ​​to the device. The device displays the received answer to the user. The user sees a message on their smartphone screen saying, "We understand how you feel. Please let us know if there's anything we can do to help."

[2100] Specific examples

[2101] For example, if a user enters "I am very sad," the emotion engine will recognize the emotion "sadness" from this text, and the generation AI will generate a response that will comfort the user. An example of a specific prompt sentence is, "If a user enters 'I am very sad,' the emotion engine will recognize the emotion "sadness" from this text, and the generation AI will generate a response that will comfort the user."

[2102] In this way, a system is realized that provides appropriate answers that correspond to the user's emotions.

[2103] The flow of the identification process in the first embodiment will be described with reference to FIG.

[2104] Step 1:

[2105] A user inputs a question into a text input field of a smartphone application. For example, the user inputs "I am very sad" and presses the send button. The input data is in text format.

[2106] Step 2:

[2107] The terminal sends the text data entered by the user to the server using the HTTPS protocol. Since the input data is encrypted, there is a low risk of it being intercepted by a third party. The output is encrypted text data.

[2108] Step 3:

[2109] The server passes the received text data to the emotion engine. The input is encrypted text data, which is decrypted to obtain text information. The emotion engine uses natural language processing (NLP) technology to analyze the text data and recognize the user's emotion. For example, it recognizes the emotion "sadness" from the text "I am very sad." The output is emotional information.

[2110] Step 4:

[2111] The server sends the emotion information recognized by the emotion engine in JSON format to the generation AI. The input is emotion information, which is converted into JSON format. For example, it is sent in the format {"emotion": "sadness"}. The output is emotion information in JSON format.

[2112] Step 5:

[2113] The generative AI generates an answer based on the emotional information it receives. The input is emotional information in JSON format, which is analyzed to generate an appropriate answer. For example, if the generative AI receives the emotional information "sadness," it generates the answer "I understand how you feel. Please let me know if there's anything I can do to help." The output is a text answer.

[2114] Step 6:

[2115] The server sends the answer generated by the generation AI to the terminal. The input is a text answer, which is encrypted and sent to the terminal. The output is an encrypted text answer.

[2116] Step 7:

[2117] The device decrypts the received answer and displays it to the user. The input is the encrypted text answer, which is decrypted and displayed to the user. The user sees a message on their smartphone screen saying, "We understand how you feel. Let us know if there's anything we can do to help." The output is the text answer displayed to the user.

[2118] (Application example 1)

[2119] 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."

[2120] Conventional systems often struggle to provide appropriate answers to user questions. In particular, they generate answers without considering the user's emotions, making it difficult to alleviate the user's anxiety or sadness. Furthermore, they lack sufficient anonymization and decryption methods to safely process questions containing personal information, which poses a privacy problem.

[2121] The identification process by the identification 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 a question including personal information and anonymizing the relevant personal information, means for sending the anonymized question to the generation AI, means for receiving an answer from the generation AI and decrypting the relevant anonymous information, means including an emotion engine for analyzing emotions from the user's input text, means for sending the analyzed emotion information to the generation AI and generating an answer based on the emotion, and means for displaying the generated answer to the user. This makes it possible to provide an appropriate answer that corresponds to the user's emotions and process the question safely while protecting privacy.

[2122] "Personal information" is information that can identify a specific individual.

[2123] A "question" is an inquiry that a user inputs to the system.

[2124] "Anonymization" is the process of converting personal information based on specific rules so that a specific individual cannot be identified.

[2125] "Generative AI" is artificial intelligence that generates appropriate answers to questions from users.

[2126] "Decryption" is the process of restoring anonymized information to its original form based on specific rules.

[2127] An "emotion engine" is software or hardware for analyzing emotions from user input text.

[2128] "Emotion information" is data on the user's emotions analyzed by the emotion engine.

[2129] "Display" refers to the process of visually presenting the generated answer to the user.

[2130] A system for implementing this invention is configured as follows: First, a user inputs a question using a device such as a smartphone. Since the input question may contain personal information, the server anonymizes it. The anonymized question is sent to a generation AI, which then generates an appropriate answer.

[2131] Next, an emotion engine is used to analyze emotions from the user's input text. The emotion engine analyzes the user's text information and generates emotion information. This emotion information is sent to the generation AI, which then generates an answer based on the emotion. The generated answer is displayed to the user by the server.

[2132] Specifically, the following hardware and software are used:

[2133] Hardware: Smartphones, servers

[2134] Software: Python, OpenAI API, TextBlob

[2135] The server receives input from the user and analyzes the emotion using TextBlob. The analyzed emotion information is sent to the generation AI using OpenAI's API. The generation AI generates an appropriate answer based on the emotion information and returns the answer to the server. The server displays the generated answer to the user.

[2136] For example, if a user types, "There may be a suspicious person in my house," the emotion engine recognizes the emotion of "anxiety." Based on this emotional information, the generative AI generates the answer, "Please contact the police immediately. Please evacuate to a safe place and securely close the doors and windows."

[2137] An example of a prompt is as follows:

[2138] The user is feeling anxious. Please generate appropriate answers to the following questions: There might be a suspicious person in my house.

[2139] In this way, it is possible to provide appropriate answers that correspond to the user's feelings and process questions safely while protecting privacy.

[2140] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2141] Step 1:

[2142] A user inputs a question using a terminal. The input question is sent to the server in text format. The input data is the user's question text.

[2143] Step 2:

[2144] The server anonymizes personal information from the received question text. For anonymization, the server converts the personal information based on specific rules so that a specific individual cannot be identified. The input data is the user's question text, and the output data is the anonymized question text.

[2145] Step 3:

[2146] The server sends the anonymized question text to the generation AI, which generates an appropriate answer based on the received question text. The input data is the anonymized question text, and the output data is the generated answer text.

[2147] Step 4:

[2148] The server uses an emotion engine to analyze emotions from the user's input text. The emotion engine uses software such as TextBlob to analyze the text information and generate emotion information. The input data is the user's question text, and the output data is the analyzed emotion information.

[2149] Step 5:

[2150] The server sends the analyzed emotional information to the generation AI, which then generates an answer based on the emotion. The generation AI then generates an appropriate answer taking the emotional information into account. The input data is the analyzed emotional information and anonymized question text, and the output data is an answer text based on the emotion.

[2151] Step 6:

[2152] The server displays the generated answer text to the user, who can visually check the generated answer through the terminal. The input data is the generated answer text, and the output data is the answer displayed to the user.

[2153] In this way, it is possible to provide an appropriate answer that corresponds to the user's feelings and process questions safely while protecting privacy.

[2154] Example 2

[2155] 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."

[2156] Conventional systems do not adequately anonymize and decrypt data, including personal information, resulting in incomplete privacy protection. They also have difficulty generating appropriate answers that take user sentiment into account. This leads to problems such as a poor user experience and a loss of system reliability.

[2157] 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.

[2158] In this invention, the server includes mean...

Claims

1. A means to receive questions containing personal information and to anonymize such personal information; and A means for sending anonymized questions to the generation AI; and A means for receiving a response from the generating AI and decrypting the corresponding anonymous information; means for providing an interface for accepting input from a user; means for analyzing emotions from text information of the user's question, the emotion engine; A means for transmitting emotion information analyzed by the emotion engine to the generation AI; A means for the generation AI to generate an answer based on emotion information; means for displaying the generated answer to the user; A system including:

2. The system according to claim 1 , wherein the anonymizing means performs anonymization by converting the personal information based on a specific rule.

3. The system according to claim 2 , wherein the decrypting means decrypts the anonymized information by restoring the anonymized information to original information based on the specific rule.

Images